WO2015168909A1 - 数据传输控制节点、通信***及数据传输管理方法 - Google Patents

数据传输控制节点、通信***及数据传输管理方法 Download PDF

Info

Publication number
WO2015168909A1
WO2015168909A1 PCT/CN2014/077061 CN2014077061W WO2015168909A1 WO 2015168909 A1 WO2015168909 A1 WO 2015168909A1 CN 2014077061 W CN2014077061 W CN 2014077061W WO 2015168909 A1 WO2015168909 A1 WO 2015168909A1
Authority
WO
WIPO (PCT)
Prior art keywords
tcp
data transmission
mptcp
substream
network node
Prior art date
Application number
PCT/CN2014/077061
Other languages
English (en)
French (fr)
Other versions
WO2015168909A9 (zh
Inventor
黄敏
张姝
Original Assignee
华为技术有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 华为技术有限公司 filed Critical 华为技术有限公司
Priority to CN201480029919.1A priority Critical patent/CN105264845B/zh
Priority to EP14891549.9A priority patent/EP3133784B1/en
Priority to PCT/CN2014/077061 priority patent/WO2015168909A1/zh
Publication of WO2015168909A1 publication Critical patent/WO2015168909A1/zh
Publication of WO2015168909A9 publication Critical patent/WO2015168909A9/zh
Priority to US15/345,022 priority patent/US10511535B2/en

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00—Traffic control in data switching networks
    • H04L47/10—Flow control; Congestion control
    • H04L47/19—Flow control; Congestion control at layers above the network layer
    • H04L47/193—Flow control; Congestion control at layers above the network layer at the transport layer, e.g. TCP related
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00—Arrangements for monitoring or testing data switching networks
    • H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
    • H04L43/0823—Errors, e.g. transmission errors
    • H04L43/0829—Packet loss
    • H04L43/0841—Round trip packet loss
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00—Traffic control in data switching networks
    • H04L47/10—Flow control; Congestion control
    • H04L47/12—Avoiding congestion; Recovering from congestion
    • H04L47/122—Avoiding congestion; Recovering from congestion by diverting traffic away from congested entities
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00—Traffic control in data switching networks
    • H04L47/10—Flow control; Congestion control
    • H04L47/32—Flow control; Congestion control by discarding or delaying data units, e.g. packets or frames
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L69/00—Network arrangements, protocols or services independent of the application payload and not provided for in the other groups of this subclass
    • H04L69/14—Multichannel or multilink protocols
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/24—Multipath

Definitions

  • Embodiments of the present invention relate to the field of communications technologies, and in particular, to a data transmission control section. Point, communication system and data transmission management method.
  • UE User Equipment
  • MPTCP Multi-Path Transmission Control Protocol
  • TCP Transmission Control Protocol
  • Congestion Control Coupled Congestion Control
  • One of the criteria is to balance congestion, that is, to divert data from the most congested links to others as much as possible.
  • End-to-end congestion control algorithm specified by existing protocols although not Data is offloaded with the communication link, but the data splitting result cannot be fully adapted to the system operation.
  • Various expected requirements which in turn cannot optimize system performance, congestion control algorithm The practicability is poor.
  • Embodiments of the present invention provide a data transmission control node, a system, and a data transmission tube.
  • the method can adapt the MPTCP data offloading result to the expected demand of the system operation.
  • an embodiment of the present invention provides a data transmission control node for controlling Data transmission between the two ends of the communication of the multi-path transmission control protocol MPTCP connection
  • the MPTCP connection includes a plurality of Transmission Control Protocol TCP substreams, and the data transmission control section
  • the point includes a processing unit, configured to select, in the plurality of TCP substreams, data transmission Transmitting at least one TCP substream; the processing unit is further configured to determine MPTCP congestion An adjustment policy of an input of the control algorithm; a sending unit, configured to send the adjustment policy Giving the network node corresponding to the at least one TCP substream, the adjustment policy is used to indicate the location
  • the network node adjusts a data transmission process on the at least one TCP substream, triggering the Both ends of the communication perform the MPTCP congestion control algorithm, and the MPTCP congestion control algorithm
  • the output of the data includes determining a portion of the data transmitted on the at least one TCP substream or All migrated to other TCP substreams belonging to the MPTCP
  • the processing unit is configured to determine An adjustment policy of an input item of the MPTCP congestion control algorithm, including: determining to reduce the The data transmission rate of one less TCP substream and the extent to which the data transmission rate is reduced.
  • the MPTCP An input of the plug control algorithm includes a data round trip delay of the at least one TCP substream RTT and/or packet loss rate.
  • the sending unit is used by the sending unit Instructing the network node to adjust a data transmission process on the at least one TCP substream Obtaining an adjusted input, comprising: instructing the network node to delay sending the at least a packet on a TCP substream to increase the RTT, delay time and the data transmission Corresponding to the magnitude of the rate reduction; and/or indicating that the network node discards the at least one Traffic packets on the TCP substream to increase the packet loss rate, discarding the service data packet The number corresponds to the magnitude of the decrease in the data transmission rate.
  • the processing unit is used by the processing unit Selecting at least one of the plurality of TCP substreams that requires data transmission control
  • the TCP substream includes: selecting the at least one TCP substream according to an operator management policy.
  • the processing unit is used by the processing unit Selecting at least one of the plurality of TCP substreams that requires data transmission control a TCP substream, including: radio resources of network nodes corresponding to the plurality of TCP substreams The status determines a network node that needs to perform data transmission control, and selects the network node corresponding to At least one TCP substream that is undergoing data transfer.
  • the processing unit is used by the processing unit Selecting at least one of the plurality of TCP substreams that requires data transmission control a TCP substream, including: determining a link status of a communication link where the plurality of TCP substreams are located Whether the preset condition is met, and in the TCP substream that does not satisfy the preset condition, select At least one TCP substream for data transmission.
  • the method further includes receiving a single a unit, configured to receive a measurement report reported by the network node, where the measurement report is used to indicate Link status of the communication link of the plurality of TCP substreams and/or the plurality of TCP substream pairs The radio resource status of the network node in response; the processing unit is specifically configured to perform the measurement according to the Reporting, selecting the at least one TCP substream.
  • a unit configured to, according to an MPTCP connection message, the plurality of TCP substreams and the MPTCP
  • the connection is associated such that the MPTCP connection includes the plurality of TCP substreams
  • the MPTCP connection message is used to indicate the status of the MPTCP connection.
  • the information management list The element is also used to store and update an MPTCP connection information table, the MPTCP connection information table
  • the MPTCP connection message carried in the MPTCP connection message reported by the network node is included
  • the MPTCP connection information is used to connect the multiple TCP substreams to the MPTCP. Connected.
  • One end is a user equipment UE accessing the network node, and the other end is a server;
  • the network node supports the UE to communicate with the server.
  • an embodiment of the present invention provides a network node for controlling multipath Data transmission between the two ends of the communication control protocol MPTCP connection, the MPTCP The connection includes a plurality of Transmission Control Protocol TCP substreams, and the network node is located in the communication two End of the communication link, including,
  • a receiving unit configured to receive, by the data transmission control node, the selected one of the plurality of TCP substreams Information of at least one TCP substream obtained;
  • the receiving unit is further configured to receive an adjustment policy indicated by the data transmission control node
  • the adjustment strategy is an adjustment policy for an input of the MPTCP congestion control algorithm.
  • the output of the MPTCP congestion control algorithm includes determining at least Part or all of the data transmitted on a TCP substream migrates to belong to the MPTCP connection On other TCP substreams.
  • the network node further includes a sending unit, configured to send the adjusted input item to the two ends of the communication, the adjusted Inputting, by the policy execution unit, adjusting data transmission on the at least one TCP substream Obtained after the input process, used to trigger the two ends of the communication to perform the MPTCP congestion control calculation law.
  • the MPTCP The adjustment policy of the input of the plug control algorithm includes: reducing the at least one TCP substream The data transmission rate and the magnitude of the data transmission rate reduction.
  • the MPTCP An input of the plug control algorithm includes a data round trip delay of the at least one TCP substream RTT and/or packet loss rate.
  • the policy execution order Means for adjusting a data transmission process on the at least one TCP substream including delaying transmission Sending data packets on the at least one TCP substream to increase the RTT, delay time and location Determining the magnitude of the data transmission rate reduction; and/or discarding the at least one TCP substream
  • the service data packet is increased to increase the packet loss rate, and the number and the number of the service data packets are discarded.
  • the magnitude of the data transmission rate reduction corresponds.
  • the sending unit further Used to report a measurement report to the data transmission control node, where the measurement report is used to refer to a link status of the communication link of the plurality of TCP substreams and/or the plurality of TCP substreams Corresponding network resource status of the network node, the measurement report is used for the data transmission control
  • the node selects the at least one TCP substream.
  • the sending unit further And signaling, to the data transmission control node, an MPTCP connection message, where the MPTCP is a connection message is used to indicate a status of the MPTCP connection, in the MPTCP connection message Carrying MPTCP connection information, the MPTCP connection information is used for the data transmission control a node associating the plurality of TCP substreams with the MPTCP connection such that The MPTCP connection contains the plurality of TCP substreams.
  • the two ends of the communication One end is a user equipment UE accessing the network node, and the other end is a server;
  • the network node supports the UE to communicate with the server.
  • an embodiment of the present invention provides a communication system, including
  • a data transmission control node configured to select, in the plurality of TCP substreams, a number to be performed According to at least one TCP substream controlled by the transmission; determining an input of the MPTCP congestion control algorithm
  • the adjustment policy is sent to the network corresponding to the at least one TCP substream a network node, the adjustment policy is used to instruct the network node to adjust the at least one TCP
  • the data transmission process on the substream triggers the two ends of the communication to perform the MPTCP congestion control Algorithm
  • the output of the MPTCP congestion control algorithm includes determining at least Part or all of the data transmitted on a TCP substream migrates to belong to the MPTCP connection On other TCP substreams;
  • a network node located in a communication link at both ends of the communication, for receiving data transmission a control node selects a letter of at least one TCP substream obtained from the plurality of TCP substreams Receiving an adjustment policy indicated by the data transmission control node, where the adjustment strategy is a pin An adjustment strategy for the input of the MPTCP congestion control algorithm.
  • an embodiment of the present invention further provides a data transmission management method,
  • the method includes: the data transmission control node selects a number of the plurality of TCP substreams to be selected At least one TCP substream according to transmission control; the data transmission control node determines MPTCP An adjustment policy of an entry of a congestion control algorithm; the data transmission control node indicating the And the network node corresponding to the at least one TCP substream adjusts the at least according to the adjustment policy a data transmission process on a TCP substream, triggering both ends of the communication to execute the MPTCP a congestion control algorithm, the output of the MPTCP congestion control algorithm includes determining Part or all of the data transmitted on the at least one TCP substream migrates to belong to the MPTCP is connected to other TCP substreams.
  • the data transmission control node An adjustment policy for determining an entry of an MPTCP congestion control algorithm includes the data transmission The control node determines to reduce a data transmission rate of the at least one TCP substream and the The magnitude of the data transfer rate reduction.
  • the MPTCP congestion control includes a data round trip delay RTT of the at least one TCP substream and / or packet loss rate.
  • the data transmission control section Pointing the network node to adjust a data transmission process on the at least one TCP substream Including the data transmission control node instructing the network node to delay transmitting the at least one Packets on the TCP substream to increase the RTT, delay time, and data transmission speed
  • the magnitude of the rate reduction corresponds; and/or the data transmission control node indicates the network section Pointing off the service data packet on the at least one TCP substream to increase the packet loss rate, and discarding
  • the number of discarded service packets corresponds to the magnitude of the data transmission rate reduction.
  • the data transmission control section Pointing at least one of the plurality of TCP substreams that needs to perform data transmission control a TCP substream, comprising: the data transmission control node selecting a location according to a carrier management policy Said at least one TCP substream.
  • the data transmission control section Pointing at least one of the plurality of TCP substreams that needs to perform data transmission control a TCP substream, including: the data transmission control node corresponding to the multiple TCP substreams
  • the wireless resource status of the network node determines the network node that needs to perform data transmission control. Selecting at least one TCP substream corresponding to the network node that is undergoing data transmission.
  • the TCP substream includes: the data transmission control node determines that the plurality of TCP substreams are located Whether the link status of the communication link satisfies a preset condition, and does not satisfy the preset condition In the TCP substream, at least one TCP substream that is undergoing data transmission is selected.
  • the data transmission control section Pointing at least one of the plurality of TCP substreams that needs to perform data transmission control a TCP substream, the data transmission control node receiving the measurement reported by the network node a quantity report, the measurement report is used to indicate a link of a communication link of the plurality of TCP substreams Status and/or radio resource status of the network node corresponding to the plurality of TCP substreams; The at least one TCP substream is selected according to the measurement report according to the transmission control node.
  • the method further includes The data transmission control node, according to the MPTCP connection message, the plurality of TCPs a substream is associated with the MPTCP connection such that the MPTCP connection includes the plurality of A TCP substream, the MPTCP connection message is used to indicate the status of the MPTCP connection.
  • the data transmission control section Point storage and update the MPTCP connection information table, where the MPTCP connection information table is included
  • MPTCP connection information carried in the MPTCP connection message reported by the network node
  • MPTCP connection information is used to associate the plurality of TCP substreams with the MPTCP connection Union.
  • the two ends of the communication One end is a user equipment UE accessing the network node, and the other end is a server;
  • the network node supports the UE to communicate with the server.
  • an embodiment of the present invention further provides a data transmission management method,
  • the result includes determining that part or all of the data transmitted on the at least one TCP substream is migrated Moving to another TCP substream belonging to the MPTCP connection, wherein the adjustment policy is An adjustment strategy for the input of the MPTCP congestion control algorithm.
  • the network node is adjusted at the network node After the data transmission process on the at least one TCP substream, the network node further includes The adjusted input item is sent to the two ends of the communication, and the adjusted input item is sent by the network After the network node adjusts the data transmission process on the at least one TCP substream, the network node obtains The MPTCP congestion control algorithm is executed at both ends of the communication.
  • the MPTCP The adjustment policy of the input of the plug control algorithm includes: reducing the at least one TCP substream The data transmission rate and the magnitude of the data transmission rate reduction.
  • the MPTCP An input of the plug control algorithm includes a data round trip delay of the at least one TCP substream RTT and/or packet loss rate.
  • the network node is adjusted
  • the data transmission process on the at least one TCP substream includes: the network node delay Transmitting a data packet on the at least one TCP substream to increase the RTT, delay time and The magnitude of the data transmission rate reduction corresponds; and/or the network node discards the Traffic packets on at least one TCP substream to increase the packet loss rate and discard the service
  • the number of data packets corresponds to the magnitude of the data transmission rate reduction.
  • the method further includes The network node reports a measurement report to the data transmission control node, where the measurement Reporting a link status and/or the plurality of communication links indicating the plurality of TCP substreams Radio resource status of the network node corresponding to the TCP substream, the measurement report is used for the The data transmission control node selects the at least one TCP substream.
  • the method further Including, the network node reports an MPTCP connection message to the data transmission control node,
  • the MPTCP connection message is used to indicate a status of the MPTCP connection, the MPTCP
  • the connection message carries MPTCP connection information, and the MPTCP connection information is used for the number Correlating the plurality of TCP substreams with the MPTCP connection according to a transmission control node, such that The MPTCP connection includes the plurality of TCP substreams.
  • the two ends of the communication One end is a user equipment UE accessing the network node, and the other end is a server;
  • the network node supports the UE to communicate with the server.
  • the technical solution provided by the embodiment of the present invention is selected according to the system operation requirement. Performing a TCP subflow of data offloading by changing the input of the MPTCP congestion control algorithm, Enable the data on the selected TCP substream to migrate to the same MPTCP connection as expected On other TCP substreams. By adding control measures for MPTCP data transmission on the access network side Therefore, the data shunting result on the TCP substream can be adapted to the expected demand of the system operation, and optimized. System performance improves the usability of the MPTCP congestion control algorithm.
  • FIG. 1 is a schematic diagram of an MPTCP protocol stack
  • FIG. 2 is a schematic diagram of an application scenario according to an embodiment of the present invention.
  • FIG. 3 is a schematic block diagram of a data transmission control node according to an embodiment of the present invention.
  • Figure 4 is a schematic diagram of the initial connection establishment process of the MPTCP and the signaling process of the TCP subflow Schematic diagram of the process;
  • FIG. 5 is a schematic block diagram of a data transmission control node according to an embodiment of the present invention.
  • FIG. 6 is a schematic block diagram of another data transmission control node according to an embodiment of the present invention.
  • FIG. 7 is a schematic block diagram of another network node according to an embodiment of the present invention.
  • FIG. 8 is a schematic diagram of a communication system according to an embodiment of the present invention.
  • FIG. 9 is a schematic flowchart of a data transmission management method according to an embodiment of the present invention.
  • FIG. 10 is a diagram of associating a TCP subflow to an MPTCP connection according to an embodiment of the present invention. Schematic diagram of the method flow;
  • FIG. 11 is a schematic flowchart of another data transmission management method according to an embodiment of the present invention.
  • GSM Global System for Mobile Communications
  • CDMA Code Division Multiple Access
  • TDMA Time Division Multiple Access
  • WCDMA Wideband Code Division Multiple Access
  • LTE Long Term Evolution
  • WLAN Wireless Local Area Network
  • WiMax Worldwide Interoperability for WiMax (Worldwide Interoperability for Other wireless broadband access networks such as Microwave Access) systems.
  • the network node described in the embodiment of the present invention may be an access device.
  • it can be a base transceiver station (BTS, Base Transceiver) in GSM or CDMA. Station), can also be Node B in WCDMA, or it can be played in LTE
  • An eNodeB (evolved NodeB), which can also be an access point in the WLAN (AP, Access Point), or other communication or wireless network system with similar functions
  • the network element is not limited in this embodiment of the present invention.
  • Network nodes can also be access control
  • the device may be a base station controller in GSM or CDMA (BSC, Base Station) Controller), which can be a radio network controller in WCDMA (RNC, Radio) Network Controller), which can also be an access controller in the WLAN (AC, Access) Controller) may also be another network element having similar functions, which is used in this embodiment of the present invention. Do not make any restrictions.
  • the data transmission control node proposed in the embodiment of the present invention is a representation
  • the access network side is used to control the unit of MPTCP data transmission, and its deployment location is not limited, It can be deployed separately or in conjunction with network management, access control devices, and access devices.
  • the control node can be a standalone device or can be deployed in other communication devices. The embodiment of the present invention does not limit this.
  • the data transfer control node can be deployed as a control unit in the above In the access control device such as BSC, RNC or AC; the data transmission control node can also be used as A control unit is deployed in the access device such as the BTS, the NodeB, the e-NodeB, and the AP.
  • the data transmission control node can also be an independently deployed device and can The above base station device or base station control device of the network node performs information interaction.
  • the user equipment (UE, User Equipment) appearing in the embodiment of the present invention may be User equipment that wirelessly accesses the access network for data transmission, such as a mobile phone, A personal computer such as a wireless communication function.
  • the access network described in the embodiment of the present invention includes the BTS and the BSC, or a radio access network (RAN, which is composed of a NodeB and an RNC, or an e-NodeB, etc. Radio Access Network), and a broadband access network consisting of devices such as APs and ACs.
  • the AC and the RNC can be integrated, that is, the function of the AC is integrated on the RNC, and the AP is associated with the AP. Communicate.
  • FIG. 1 is a schematic diagram of the MPTCP protocol stack, compared to standard TCP (standard TCP)
  • the MPTCP protocol stack mainly divides the TCP layer into an MPTCP layer and a TCP sub-flow (TCP).
  • the Subflow layer implements concurrent data transmission by: establishing multiple TCPs. Substream, which is equivalent to establishing multiple TCP connections; counting multiple TCP substreams by the MPTCP layer According to the treatment of diversion or convergence.
  • the TCP subflow can be connected to the associated MPTCP through the token.
  • Token is linked. Token is the identifier of the MPTCP connection and has only local significance. And uniqueness, both the initiator and the response end of the MPTCP initial connection establishment process must generate one Token, Token can be regarded as the MPTCP Connection ID.
  • An MPTCP connection can consist of multiple TCP substreams.
  • TCP subflow is at a specific The TCP data stream that is transmitted on the transport path (path).
  • the transmission path means a path between the two ends of the communication of the MPTCP connection, which can pass through the two ends of the communication
  • a pair of addresses such as an IP address, and/or a port number (port) to identify each TCP subflow There are corresponding paths.
  • MPTCP is an end-to-end protocol, so both ends of the communication need interactive control signaling.
  • To complete initial connection establishment such as MPTCP, TCP substream addition or deletion, address notification, etc. Features. These control signaling are transmitted through the TCP option (TCP Option).
  • FIG. 2 is a schematic diagram of an application scenario according to an embodiment of the present invention.
  • UE and The MPTCP connection between Servers has 2 TCP substreams, and one TCP substream is uploaded on the LTE link.
  • the LTE link includes an access network side node e-NodeB and a core network side node independent industry.
  • Service Gateway SGW, Service Gateway
  • PGW Packet Data Network Gateway
  • Another TCP substream is transmitted on the Wi-Fi link.
  • Wi-Fi link includes access network side node AP and AC; core network side node broadband remote access Server (BRAS, Broadband Remote Access Server).
  • BRAS Broadband Remote Access Server
  • this The data transmission control node provided by the embodiment of the invention can be independently set and respectively There is a link connection between the e-NodeB and the AC to facilitate information exchange.
  • e-NodeB and AP The MPTCP connection message can be reported to the data transmission control node separately, and the MPTCP connection is eliminated.
  • the information is obtained by the e-NodeB or the AP to resolve the MPTCP signaling interaction process between the UE and the server. For example, the TCP data packet carrying the MPTCP signaling of the access user is obtained, and the data transmission is performed.
  • the control node will belong to different TCP substreams of the MPTCP connection according to the MPTCP connection message.
  • the condition of the radio resource or the link status of the access user to determine the data is not
  • the allocation strategy between the TCP sub-flows is finally sent to the e-NodeB and the AP. Row.
  • the embodiment of the invention provides a data transmission control node, which can be used for control Data transmission between the two ends of the communication of the MPTCP connection, the MPTCP connection includes multiple pieces Transmission Control Protocol TCP substream.
  • the two ends of the MPTCP connection can be air interface uplink or air interface downlink data. Any two ends of the communication device during the transmission process can initiate or respond to the MPTCP connection.
  • Related processes including MPTCP initial connection establishment process, TCP subflow new creation process, TCP substream deletion process, TCP substream priority update process, and the like. For example, by accessing some The UE of the line access point initiates an initial connection establishment process of the MPTCP, and the server on the network side performs the process. In response, the UE and the server become the communication ends of the MPTCP connection; similarly, The MPTCP initial connection establishment process is initiated by the server, and the UE responds.
  • the originating or responding end of the initial connection establishment process of the MPTCP and the TCP subflow are newly created.
  • the originating or responding end of the process, and the transmitting or receiving end of the subsequent data transmission process There is no corresponding relationship between them, for example, the originating end of the MPTCP initial connection establishment process or The responding end can initiate the TCP subflow new process, and the MPTCP initial connection establishment process
  • the originating end or the responding end, either the originating end or the responding end of the TCP subflow new process Both can send data.
  • the schematic block diagram of the data transmission control node is shown in FIG. 3, and mainly includes:
  • the processing unit 301 can be configured to select, among the multiple TCP substreams, a number to be performed. According to at least one TCP substream controlled by the transmission.
  • the processing unit 301 can determine which air sub-flows on the TCP sub-flow are selected to go down. Or the upstream data is migrated to other TCP substreams that belong to the same MPTCP connection, and will need The TCP substream to be data migrated is classified into the air interface downlink or uplink pipe according to the data transmission direction. Control subflow set.
  • the processing unit 301 can also be configured to determine an input of an MPTCP congestion control algorithm. The adjustment strategy of the item.
  • the sending unit 302 is configured to send the adjustment policy to the at least one TCP sub- Flow corresponding to the network node, the adjustment policy is used to instruct the network node to adjust the Reducing a data transmission process on a TCP substream, triggering both ends of the communication to perform the MPTCP congestion control algorithm, the output of the MPTCP congestion control algorithm includes Part or all of the data transmitted on the at least one TCP substream migrates to the belonging On the other TCP substreams of the MPTCP connection.
  • the network node is located in a communication link at both ends of the communication.
  • the sending process of the foregoing adjustment policy includes: traversing the air interface downlink or uplink control substream The set sends an adjustment policy to the network node where each TCP subflow in the control subflow set is located.
  • the following two methods can be used:
  • Method 1 First, according to the air interface of the wireless access point in the MPTCP connection information table Downstream or uplink radio resource status, get the wireless access point to be controlled, for example, select empty A wireless access point with a downlink or uplink load congestion, optionally, can be combined with carrier management The strategy is judged. Second, iterate through all the TCPs under each wireless access point that needs to be controlled. Subflow, if the current TCP subflow belongs to the MPTCP connection, there is an air interface downstream or uplink. Backup identifies other TCP sub-flows with bit 0, and other TCP sub-flows are located at the wireless access point. Non-current wireless access point, and the air interface of the other wireless TCP access point is down or on If the radio resource status is idle, the current TCP substream is selected to enter the air interface downlink or uplink tube. Control subflow set.
  • Method 2 Traverse all TCP substreams in the MPTCP connection information table. If the current child The air interface link information of the access user to which the flow belongs indicates that the air interface of the link is down or up. Data transmission, for example, the link quality of the current TCP substream is poor or the access user is The edge of the cell, and the current TCP sub-flow belongs to the MPTCP connection. Backup identifies other TCP sub-flows with bit 0, and other TCP sub-flows are located at the wireless access point. A wireless access point that is not a current TCP substream, and the access user to which other TCP substreams belong is empty. The port link information indicates that the air interface downlink or uplink is suitable for data transmission, and the current TCP substream is selected. Enter the air interface downlink or uplink control sub-flow set.
  • the number of TCP substreams in the control subflow set can be defined.
  • the priority of the user can be preferentially selected to preferentially select the current wireless access point.
  • the TCP substream corresponding to the prior user enters the control subflow set.
  • Data migration can be to transfer part of the data of the TCP substream to the same On other TCP substreams connected to MPTCP, all data of the TCP substream can also be transferred to On other TCP substreams that belong to an MPTCP connection, the ratio of data migration can be
  • the embodiment of the present invention is not particularly limited as determined by the purpose of optimization of the operation of the system.
  • the processing unit 301 is configured to determine An adjustment policy for an entry of an MPTCP congestion control algorithm, including determining to reduce the The data transmission rate of one less TCP substream and the extent to which the data transmission rate is reduced.
  • the speed is reduced to a certain range. After the degree, the TCP substream will not meet the performance requirements of data transmission, and both ends of the communication can Detecting or obtaining the rate information of the TCP substream by the network node, and starting MPTCP congestion
  • the control algorithm switches the TCP substream.
  • the processing unit 301 may determine a specific implementation manner of the adjustment policy, It is directly sent by the sending unit 302 to the corresponding network node; the processing unit 301 can also Only the TCP substream migration policy is given and sent to the corresponding network node, and the network node itself Decide which specific form to take to perform the data migration.
  • the offloading strategy can only include The rate at which a TCP substream needs to be reduced, and can be determined by the network node itself. Which specific form should be used, such as delay sending or increasing the packet loss rate to satisfy the score The traffic policy; similarly, the adjustment policy may also directly indicate that the network node sends by delay Or increase the packet loss rate to reduce the rate of the TCP substream.
  • the TCP layer that is an input of the MPTCP congestion control algorithm can be changed.
  • RTT Real Time
  • packet loss rate will be adjusted RTT and / Or the packet loss rate is used as an input to the MPTCP congestion algorithm, affecting the output of the algorithm, so that The data on each TCP substream can complete the migration process that meets the expected requirements.
  • adjusting the policy may include instructing the network node to delay sending the at least Packets on a TCP substream to increase RTT, delay time and the data transmission rate The magnitude of the decrease corresponds.
  • adjusting the policy may include instructing the network node to discard the at least one piece
  • the service data packet on the TCP substream increases the packet loss rate and discards the number of the service data packets.
  • adjusting the policy may include instructing the network node to discard the at least one piece
  • the service data packet on the TCP substream increases the packet loss rate and discards the number of the service data packets.
  • the above process of changing the RTT and the packet loss rate can be performed simultaneously, that is, comprehensive application RTT and packet loss rate change the data transmission rate of the TCP substream; or separately, the present invention
  • the embodiment does not particularly limit this.
  • the data offload scheduling of the MPTCP congestion control algorithm specified by the existing protocol The results mainly depend on RTT and packet loss rate. Therefore, it is possible to use the RTT and the loss of each communication link.
  • the packet rate is used as an input to affect the output of the MPTCP congestion control algorithm, for example by increasing Packet loss rate on a link, so that data migration is performed under the action of the congestion control algorithm Go to another link. Therefore, the above adjustment strategy can be changed by changing the RTT of the TCP layer. / or packet loss rate, indirectly affect the output of the above algorithm, so that the data can be shunted Enough to meet the different needs of the system operation.
  • the MPTCP congestion control algorithm used in the embodiments of the present invention is existing. Coupled Congestion Control algorithm specified by the protocol, the algorithm operation process can refer to Protocol related content, such as slow start, fast retransmission, fast recovery in each TCP substream The items are consistent with the existing algorithms. For details, see RFC6356, which is not described here.
  • the network node may be instructed to actively drop the air interface on the TCP substream or Upstream service data packets to increase the packet loss rate, for example, increasing the packet loss rate to a specified ratio, Or actively discard a packet every few packets.
  • the network node may be instructed to delay the uplink or downlink of the air interface on the substream.
  • ACK packet to increase the RTT of the TCP layer, for example, specifying the air interface uplink or downlink ACK on the substream The packet is delayed for a certain time and then sent.
  • the network node may be instructed to delay downlink or uplink data by delaying the air interface on the substream.
  • the packet enters the scheduler to increase the RTT of the TCP layer, for example, specifying the data delay on the TCP substream. Enter the scheduler after a while.
  • the network node can be instructed to influence the scheduler behavior to increase the RTT of the TCP layer. For example, by not scheduling the air interface downlink or uplink data of the TCP substream, or reducing the TCP sub-
  • the air interface of the stream downlinks or uplinks the priority to a certain degree, or the scheduler senses TCP.
  • the time point of the downlink or uplink data scheduling requirement of the substream is delayed by a certain value.
  • the processing unit 301 may be specifically configured to: According to at least the carrier management policy, the link status of the communication link, and the radio resource status A condition selects a TCP subflow.
  • the processing unit 301 is configured to select, according to an operator management policy, the One less TCP subflow.
  • the carrier management and control strategy includes system load balancing and link resource utilization. Rate and billing strategies, etc., which reflect the operator's tendency to control the entire network.
  • LTE An example of a scenario concurrent with Wi-Fi data: a communication chain of a UE at the edge of an LTE network The quality of the road is not good, operators may prefer to transmit on spectrum-efficient Wi-Fi links. More data, but communication link quality and spectrum efficiency and link RTT or packet loss rate are not Direct correspondence, so only the objective existence of RTT on LTE links and Wi-Fi links And packet loss rate as the basis for data offload scheduling, can not guarantee accurate data distribution to Wi-Fi On the link; for example, considering the reason for classified charging, the operator may want the UE to do as much as possible.
  • the above carrier policy can be configured directly on the data transmission control node, or Is issued by other network elements, for example, by the control device in the core network, and issued.
  • the embodiment does not specifically limit this.
  • the processing unit 301 is further configured to be configured according to the multiple TCP substreams.
  • the wireless resource status of the network node determines the network node that needs to perform data transmission control. Selecting at least one TCP substream corresponding to the network node for data transmission.
  • the processing unit 301 is further configured to determine, where the multiple TCP substreams are located. Whether the link status of the communication link satisfies a preset condition, and does not satisfy the preset condition In the TCP substream, at least one TCP substream that is undergoing data transmission is selected.
  • the data transmission control node may further include a receiving unit 303, configured to receive a measurement report reported by the network node, where the measurement report is used to indicate the multiple TCPs Link status of the communication link of the substream and/or network node corresponding to the plurality of TCP substreams
  • the radio resource status; the processing unit 301 is specifically configured to select according to the measurement report Select at least one TCP substream.
  • the radio resource status may be based on a wireless connection reported by the network node.
  • the measurement report of the in-point is obtained, and the measurement report may include resources of the wireless access point. status.
  • the link status of the communication link may be based on the access oriented reported by the network node.
  • the measurement report of the user is obtained, and the measurement report may include an air interface link of the access user. interest.
  • the access user refers to the UE that successfully accesses the communication network through the access network.
  • the corresponding logical object in the network the UE that is one end of the MPTCP connection accesses the communication network and When performing actual communication, the UE can be regarded as an access user, in particular, accessing different protocols.
  • the same UE that has a network and has a TCP substream in different communication networks is considered to be in phase.
  • Corresponding communication networks correspond to different access users, and have corresponding ones in the corresponding communication networks.
  • User ID For example, if the UE has a radio link in the 3GPP network, the UE is in the 3GPP network.
  • IMSI or GUTI or GCI+ cell user standard can be used.
  • the UE Identify and identify the access user; if the UE has a wireless link in the WLAN network, then The UE corresponds to one access user in the WLAN network, and can adopt a MAC address or an AP. ID+IP address or AC ID+IP address or SSID+IP address to identify this access Household.
  • the data transmission control node may also To include the information management unit 304, which can be used to connect messages according to the MPTCP, and will belong to the same A plurality of TCP substreams connected by an MPTCP are associated with the MPTCP connection, The MPTCP connection message is used to indicate the status of the MPTCP connection.
  • the MPTCP connection message may include: indicating an initial connection establishment process of the MPTCP Completed MPTCP initial connection setup notification message; and indication of initial connection in MPTCP After the establishment process is completed, create a new TCP subflow belonging to the TCP subflow of the MPTCP connection. Notification message.
  • the method further includes: adding an address increase notification message of the available address; Including, deleting the deleted TCP sub-circulation of one or more TCP sub-streams currently in use Knowing the message; may also include a TCP sub-flow optimization indicating a change in the usage priority of the TCP substream Pre-update notification message, etc., wherein changing the TCP sub-flow usage priority means changing a certain The backup status of one or more TCP substreams.
  • the receiving unit 303 is further configured to receive the MPTCP connection reported by the network node.
  • the MPTCP connection message is parsed by the network node to be carried by the access user. TCP packet acquisition for MPTCP signaling.
  • the carrying content of the MPTCP initial connection establishment notification message may include But not limited to the following information:
  • the type of wireless access where the current TCP substream is located such as a 3GPP network or WLAN.
  • Network the identity of the wireless access point where the current TCP subflow is located, for example, for wireless access
  • the type is a 3GPP network, and the Global Cell Identity (GCI) table can be used.
  • the indication that the wireless access type is a WLAN network may be represented by an AP ID or an AC ID;
  • the IMSI or GUTI or GCI+ cell user ID can be used.
  • MAC can be used Address or AP ID + IP address or AC ID + IP address or SSID + IP address, etc.
  • Knowledge; the address information of the current TCP substream, the representation of this information can be the current TCP sub
  • the address of both ends of the stream that is, the IP address obtained from the header of the TCP packet of the access user.
  • the port number (Port) the Key at both ends of the current TCP substream or calculated by Key Token.
  • the carried content of the newly created TCP subflow notification message may include but is not limited to the following information:
  • the representation of this information can be the two ends of the current TCP subflow. Address; Address_ID of both ends of the current TCP subflow; response of the TCP subflow new process Token of the end, that is, the Token carried by the MP_JOIN option in the TCP SYN packet; current TCP Which end of the substream initiates the subflow creation process, for example, the server establishes MPTCP with the UE.
  • the SYN packet indicates that the server initiates a subflow new process, and vice versa indicates that the UE initiates TCP sub-flow new process; the upstream and downstream uplink of the current TCP sub-flow
  • the Backup flag is indicated by the MPTCP signaling carrying the MP_JOIN option. Whether the sender of the MPTCP signaling wants the peer to send to the MPTCP signaling sender
  • the TCP substream is used as a backup when the data is used. Therefore, the air interface of the current TCP substream is uplinked.
  • the Backup ID refers to the air interface that carries the MP_JOIN option during the TCP subflow creation process.
  • the downstream identifier of the current TCP sub-stream Refers to the air interface uplink signaling carried by the MP_JOIN option in the TCP subflow creation process.
  • Server The SYN packet carrying the MP_JOIN option sent to the UE indicates that the Backup flag is 1, and the generation
  • the backup flag of the upstream interface of the TCP sub-flow is 1, indicating that the server indicates that the UE is behind.
  • the new TCP subflow is regarded as the backup substream.
  • the backup flag is also carried in the SYN/ACK message that the UE feeds back to the server. Indicates the backup status of the TCP substream used when the Server sends data to the UE.
  • the carried content of the address increase notification message may include but is not limited to the following information: Address information that needs to be added from the ADD_ADDR option, ADD_ADDR The option is sent by one end of the current TCP subflow.
  • the form of this address information is ADDRESS_ID+ available address, which can be ADDRESS_ID+IP address and Port, or Is the ADDRESS_ID+IP address, which form depends on the contents of the ADD_ADDR option Whether the format carries a Port; the address information of the current TCP substream, and the representation of this information can be Is the address of both ends of the current TCP substream; which of the MPTCP connections to which the current TCP subflow belongs
  • the server establishes an MPTCP connection with the UE.
  • the network node recognizes that the ADD_ADDR option sent by the server indicates that the server needs to increase. Add an available address.
  • Deleting the TCP substream notification message according to the network node determines that the MPTCP substream is deleted.
  • the process can be completed in different ways to carry different information content, including:
  • deleting the content carried in the TCP sub-flow notification message may include, but is not limited to, The following information, the current TCP subflow, that is, the TCP carrying the TCP RST packet or the TCP FIN packet. Address information of the substream, the representation of this information may be the two ends of the current TCP substream site.
  • the content carried in the TCP substream notification message may be deleted. Including but not limited to the following information, the need to resolve from the REMOVE_ADDR option Deleted ADDRESS_ID collection; current TCP subflow, ie bearer REMOVE_ADDR The address information of the TCP subflow of the option. The representation of this information can be the current TCP subflow.
  • the server establishes an MPTCP connection with the UE, if the network node recognizes that The REMOVE_ADDR option sent by the server indicates that the server needs to delete the available land. site.
  • the TCP substream priority update notification message is according to the access node according to the network node.
  • the MP_PRIO option in the TCP packet determines that the MPTCP substream priority update process is complete In the case of a situation, you can carry different information content, including,
  • the TCP substream priority update notification message carrying content may include but is not limited to the following information. Address information of the current TCP substream; information indicating the priority update direction of the substream, this letter The expression of the information may be an MP_PRIO selection indicating that one end of the current TCP substream is sent. Item, you can also directly indicate the need to update the air interface downstream or air interface uplink backup label Knowledge; the value of the Backup flag bit parsed from the MP_PRIO option.
  • the updated uplink backup identifier is used to indicate that the server indicates that the UE is in the subsequent server. Whether the current TCP subflow is regarded as a backup substream when sending data; if it is sent by the UE The MP_PRIO option indicates that the downlink Backup identifier needs to be updated, and the updated The backup flag is used to indicate whether the UE instructs the server to send data to the UE subsequently. Treat the current TCP subflow as a backup subflow.
  • the MP_PRIO option carries the ADDRESS_ID, it means that the MP is excellent.
  • the prior update is applied to all of the MPTCP connections that use the ADDRESS_ID TCP subflow of the address, at this time, the TCP substream priority update notification message carrying content may include But not limited to the following information: the current TCP substream, that is, the TCP subflow that carries the MP_PRIO option.
  • Address information the representation of this information may be the address of the two ends of the current TCP substream; ADDRESS_ID parsed from the MP_PRIO option; indicating that the subflow has a higher priority
  • the new direction of information the form of this information can be indicated as some of the current TCP substream
  • the MP_PRIO option sent by the terminal can also directly indicate that the air interface needs to be updated. Or the upstream port backup identifier; the Backup ID parsed from the MP_PRIO option. The value of the bit. The specific process is similar to the situation 1 and will not be described here.
  • the types of air interface links carrying TCP sub-flows may be the same or not.
  • an MPTCP connection contains two TCP substreams, then the two TCP subflows can Both are established on the LTE link, or one TCP substream is established on the LTE link, and the other The TCP substream is built on the Wi-Fi link.
  • Embodiments of the present invention include the same MPTCP connection
  • the number of TCP substreams and the link type are not limited, and can be operated according to the network. Demand is determined.
  • the receiving unit 303 can receive the report by multiple different types of network nodes.
  • MPTCP messages belonging to multiple different TCP substreams of the same MPTCP connection. For example Receiving at least one TCP substream that is reported by the BTS, the NodeB, and the AP through the network node.
  • MPTCP message may be a base station device, and the different networks are managed.
  • a communication can be established between the base station control device of the network node and the data transmission control node Link for transmitting measurement information and control information. For example, suppose the data transfer control node Integrated in the RNC, the NodeB can directly report the MPTCP connection resolved by the NodeB.
  • the information is sent to the RNC, and the BTS or the AP can report the obtained MPTCP connection message to each.
  • Corresponding BSC and AC according to the link relationship between the BSC and the AC and the above RNC,
  • the received MPTCP connection message is forwarded to the RNC, and the RNC completes the MPTCP. Management of the connection.
  • it can also be directly at each network node and data transmission control node. When a communication link is established, each network node can report directly to the data transmission control node. Related messages, no need to be forwarded through the relevant control device.
  • the transmission control node is integrated in a network device, for example integrated in the e-NodeB, with The scenario is similar, just establish a connection between other network devices at the level corresponding to the e-NodeB.
  • the link is used for information interaction and will not be described here.
  • FIG 4 is the signaling flow of the initial connection establishment process of MPTCP and the new process of TCP subflow Figure.
  • the initial connection establishment process of the MPTCP includes establishing an MPTCP connection.
  • the first TCP substream through the three-way handshake process to establish an MPTCP connection, and establish a common
  • the difference between TCP connections is that TCP SYN packets, SYN/ACK packets, and ACK packets are carried.
  • MP-CAPABLE option and the option will carry a 64bit key (Key).
  • either end of the MPTCP connection can be Create a new TCP subflow with a pair of addresses that are not currently in use.
  • a new A2 ⁇ ->B1 TCP subflow is created between HostA and HostB.
  • A2 is HostA.
  • B1 is the available address of HostB.
  • the new TCP subflow process will carry MP_JOIN The option completes the signaling interaction process in the last 4 steps of Figure 4, where the TCP SYN packet carries
  • the MP_JOIN option contains a 32-bit Token, a 32-bit random number, and an ADDRESS_ID. And the Backup logo.
  • Token-B is the initiator of the TCP subflow new process.
  • the 64-bit key value of the end is passed through a certain encryption algorithm, such as a 32-bit value generated by a hash algorithm.
  • the initiator of the TCP subflow new process identifies this by carrying this parameter in the TCP SYN packet. Which MPTCP connection the TCP substream expects to join.
  • the initiator of the TCP subflow creation process is
  • the response end of the 32-bit random number R-A and TCP sub-flow new process carried in the TCP SYN packet is The 32-bit random number R-B carried in the SYN/ACK packet is used to further ensure the sub-flow establishment process. Security.
  • ADDRESS_ID has only local meaning and is unique, TCP subflow
  • the ADDRESS_ID carried in the TCP SYN packet by the initiator of the new process is used to identify the sender.
  • the IP address used in the current newly created TCP subflow, and the response of the TCP subflow new process is used to identify the responding end.
  • the Backup flag is used to represent the sender of this message. Look at the TCP substream as a backup. By setting 1, the sender requests the peer only if there is no A backup that has a Backup flag of 1 when the subflow with the Backup flag is 0 is available.
  • the substream sends data, for example, the host A of the TCP substream in FIG.
  • the Backup flag is used to represent whether HostA wants HostB to use the TCP subflow as Backup, the Backup of the TCP substream in the figure HostB carried in the SYN/ACK packet
  • the ambition is used to represent whether HostB wants HostA to use the TCP subflow as a backup.
  • the data transmission control node can simultaneously manage multiple MPTCP connections.
  • the device type at each end of the communication of the plurality of MPTCP connections, and each MPTCP connection includes Type of communication link where the TCP substream is located, type of wireless access point accessed by the UE, etc. No restrictions are imposed on the subject.
  • the information management unit 304 further A storage subunit 3041 may be included for storing and updating the MPTCP connection information table.
  • the MPTCP connection information table includes the MPTCP connection message reported by the network node.
  • the information management unit 304 parses out the received various types of MPTCP connection messages.
  • Related MPTCP connection information carried in, and organized into an MPTCP connection information table,
  • the MPTCP connection information includes access user information related to the current MPTCP connection status, Information such as wireless access point information.
  • the MPTCP connection information table may include that is reported by different network nodes.
  • MPTCP connection information about the same MPTCP connection can also include different network sections Point reported MPTCP connection information about different MPTCP connections.
  • Information management unit 304 The TCP sub-flows can be managed based on the MPTCP connection.
  • the information management unit 304 receives the MPTCP initial connection establishment notification Information (hereinafter referred to as "Notification Message 1"), you can maintain MPTCP connection by the following operations Receive the information form, including:
  • the notification message 1 carries the MPTCP initial connection establishment process established
  • the key value of the TCP substream (referred to as "current TCP substream” in this paragraph)
  • the data transmission The input control node can calculate the above key value into a token and save it in the current TCP substream.
  • the closed MPTCP connection information (referred to as "current MPTCP connection information” in this paragraph); If the notification message 1 carries the Tokens at both ends of the current TCP substream, it is directly saved in The current MPTCP connection information; the current TCP subflow carried in the message 1 will be notified.
  • the type of wireless access, the identity of the wireless access point, and the globally unique identifier of the wireless access user The information such as identification is saved in the current MPTCP connection information; the current information carried in the message 1 will be notified.
  • the address information of the TCP substream is saved in the current MPTCP connection information, and will be current.
  • the air interface downstream and uplink Backup identifier positions of the current TCP substream connected by the MPTCP are 0, and set the ADDRESS_ID corresponding to the address of both ends of the current TCP substream to 0.
  • the MPTCP connection information table can be maintained by the following operations. include:
  • the information management unit 301 can receive the address increase notification message.
  • the following operations are performed to maintain the MPTCP connection information table, including: according to the current TCP substream
  • the address information finds the MPTCP connection to which the current TCP subflow belongs, in the associated MPTCP connection. Save the new available address information.
  • the MPTCP connection information table can be maintained by the following operations. include:
  • the notification message 3 carries the address information of the current TCP substream, it is sent.
  • TCP subflow of REMOVE_ADDR option, and does not carry ADDRRESS_ID, then root According to the address information of the current TCP substream, find the MPTCP connection to which it belongs, in the associated Delete the information about the TCP substream in the MPTCP connection; if the message carries the need to delete In addition to the ADDRESS_ID option set, the current TCP substream address information, the current TCP sub Which end of the stream needs to delete the address and other information, then according to the above address information to find the current The MPTCP connection to which the TCP subflow belongs; according to the MPTCP connection to which the current TCP subflow belongs Which end needs to delete the address and the ADDRESS_ID set that needs to be deleted, delete the current The available addresses of the two ends of the MPTCP connection to which the TCP substream belongs, and delete the current TCP All the addresses corresponding to the above ADDRESS_ID are used in the MP
  • the information management unit 304 receives the TCP sub-flow priority update notification Information (hereinafter referred to as "Notification Message 4"), you can maintain the MPTCP connection by the following operations Information sheet, including:
  • the notification message 4 carries the address information of the current TCP substream, indicating the substream
  • the information of the priority update direction and the backup identifier are based on the address of the current TCP substream.
  • the Backup identifier includes: updating the direction according to the substream priority in the notification message 4, ADDRESS_ID and Backup ID to determine which TCP sub-links belong to the MPTCP connection.
  • the stream needs to be updated with priority and how it is updated. For example, if the subflow priority update direction refers to The performance is shown as the MP_PRIO option sent by the air interface downstream data sending end, for example, Server.
  • Information about all TCP substreams contained in the MPTCP connection to which the current TCP subflow belongs if Searching for the address of a certain end of a TCP substream matches the ADDRESS_ID.
  • the Backup identifier in the message 4 refreshes the air interface uplink Backup identifier of the TCP substream.
  • the data transfer node can also identify the backup substream by monitoring the data stream, such as The network node discovers that a certain MPTCP user is passing through a TCP subflow of the network node. If there is no data in the downlink or uplink for a period of time, you can think that this TCP substream is empty.
  • the Backup ID of the downstream or upstream direction is 1. This step can be performed by the information management unit 304. carried out.
  • the information management unit 304 may further include an information management unit 3042, which may be used for After receiving a new TCP subflow notification message, the network node belongs to a certain MPTCP.
  • the connected new TCP subflow is associated with the MPTCP connection.
  • the method includes: parsing the newly created TCP subflow notification message, and determining the TCP
  • the originating end of the substream new process is the end of the communication end of the MPTCP connection, Obtain the address of the current TCP substream and the Tokens at both ends of the subflow new process, and then judge The response end of the TCP subflow new process is the other end of the communication ends of the above MPTCP connection, And get the response address.
  • Search the MPTCP connection information table if the TCP is searched
  • the response Token of the subflow creation process is equal to the Token of the end of an MPTCP connection.
  • the response end address of the TCP subflow new process is connected to one end of the MPTCP connection. If the address in the address set is equal, the new TCP substream belongs to the MPTCP. connection.
  • the available address set includes at least one available address for establishing TCP Subflow.
  • the TCP subflow is newly determined before searching the MPTCP connection information table
  • the response end of the process is one end of the communication ends of the MPTCP connection, and the end is acquired Address, when searching for MPTCP connection information, you can only search for the correlation of the response end.
  • MPTCP connection information no need to search the peer end, saving system computing resources, Increase search efficiency.
  • the multiple access terminal can The peer initiates an address increase notification process to inform the peer of its available address.
  • the single address end of the MPTCP connection may be initiated according to the learned peer available address. TCP subflow new process.
  • the information management unit 3042 can newly create each new one managed by the data transmission control node. If the TCP substream performs the above processing, multiple pieces belonging to the same MPTCP connection can be connected. The TCP substream is associated with the MPTCP connection, including associating the newly created TCP subflow to the Dependent on the MPTCP connection.
  • Address conflict refers to the same IP address of the communication node under different subnets, for example, Host A And Host A’ belong to different WLAN A and WLAN A’ networks respectively, WLAN A and WLAN A' networks assign private IP addresses to Host A and Host A' respectively. such Host A and Host A' may have the same IP address, which is called an address conflict.
  • the data transmission control node may be connected to both the WLAN A and the WLAN A' network at the same time. In this way, Host A and Host A’ in the captured WLAN A or WLAN A’ network The address may be the same.
  • the information management unit 304 can be configured to receive the carried by the network node MPTCP connection message with network identification, network identifier can include access point name (APN, Access Point Name), wireless access point identification, etc.; for MPTCP In the retrieval operation based on the address in the connection, if the same IP address is retrieved For the MPTCP information record, the above network identifier can be used for matching retrieval.
  • network identifier can include access point name (APN, Access Point Name), wireless access point identification, etc.
  • the premise of controlling the data transmitted on the MPTCP connection is that they belong to the same
  • the TCP substream of the MPTCP connection is associated with the MPTCP connection, that is, the current UE needs to be clarified.
  • the TCP substream that can be used for data communication is which MPTCP connection and which network it belongs to. Network node access and other information, and confirm the status of the TCP sub-flow, in order to belong to the same Congestion control of different TCP substreams of MPTCP connection, including between different TCP substreams Data is shunted to obtain the expected shunt effect.
  • Embodiments of the present invention provide a network node for controlling a multipath transmission control association.
  • Data transmission between the two ends of the communication of the MPTCP connection, the MPTCP connection includes multiple a transmission control protocol TCP substream, the network node being located at a communication chain at both ends of the communication
  • the structure is shown in Figure 5, including:
  • the receiving unit 501 is configured to receive, by the data transmission control node, the multiple TCP substreams The information of at least one TCP substream obtained in the selection.
  • the receiving unit 501 is further configured to receive an adjustment policy indicated by the data transmission control node.
  • the adjustment strategy is an adjustment policy for an input of the MPTCP congestion control algorithm.
  • the input item of the MPTCP congestion control algorithm includes: the at least one The data round-trip delay RTT and/or packet loss rate of the TCP substream.
  • the policy execution unit 502 is configured to adjust the at least one piece according to the foregoing adjustment policy.
  • the data transmission process on the TCP substream triggers the two ends of the communication to perform the MPTCP congestion a control algorithm, the output of the MPTCP congestion control algorithm includes determining Part of the data transmitted on one TCP substream is partially or completely migrated to belong to the MPTCP connection. Connected to other TCP substreams.
  • the network node may further include a sending unit 503, configured to adjust the An input is sent to both ends of the communication, and the adjusted entry is executed by the policy
  • the metadata is obtained after adjusting the data transmission process on the at least one TCP substream, and is used to trigger the Both ends of the communication perform the MPTCP congestion control algorithm.
  • the adjustment policy of the input item of the MPTCP congestion control algorithm includes: reducing the a data transmission rate of at least one TCP substream and a reduced rate of the data transmission rate degree.
  • the policy execution unit 502 is configured to adjust the at least one TCP substream.
  • a data transmission process comprising: delaying transmitting a data packet on the at least one TCP substream to Increasing the RTT, the delay time corresponds to the magnitude of the decrease in the data transmission rate.
  • the policy execution unit 502 is further configured to discard the at least one TCP sub- Traffic packets on the flow to increase the packet loss rate, discarding the number of the service data packets and The magnitude of the data transmission rate reduction corresponds.
  • the RTT and the packet loss rate may be detected by the two ends of the communication.
  • the algorithm input item is not specifically limited in this embodiment of the present invention.
  • the sending unit 501 is further configured to send to the data transmission control node. Reporting a measurement report, the measurement report is used to indicate a communication link of the plurality of TCP substreams Link status and/or radio resource status of the network node corresponding to the plurality of TCP substreams, The measurement report is used by the data transmission control node to select the at least one TCP substream.
  • the measurement report may be a measurement report of a network node that is accessed by the UE.
  • the content of the report can be, but is not limited to, the following information: wireless access type, wireless access point mark Knowledge, resource status of the wireless access point, such as system load information.
  • the measurement report may be a measurement report for the access user, and the report is reported.
  • the capacity can be, but is not limited to, the following information: wireless access type, wireless access point identification, access Global unique identifier of the user, air interface link information of the UE, such as link quality and user distribution
  • the situation is, for example, a cell center user or an edge user.
  • the sub-flow reports the measurement report for the network node, or only if the access user has a certain The measurement report for the access user is reported to the TCP sub-flow associated with the MPTCP connection. To reduce the frequency of reporting reports.
  • the measurement report is The reporting mode can be: periodic reporting or event triggering reporting, where the event is triggered.
  • the report includes but is not limited to the following:
  • the measured quantity changes; or, if the measured quantity meets certain conditions, it is reported, for example.
  • the load of the wireless access point exceeds a certain threshold before reporting; or, the data transmission control section is received. After the measurement report request message sent by the point, the measurement result is reported according to the request message.
  • the measurement report is generated and reported.
  • the process can include:
  • the source access point After the source wireless access point finds that the user successfully switches, the source access point reports the access switch notification message to The data transmission control node, the carrying content may include but is not limited to the following information: The user's global unique user ID, wireless access point type, source wireless access point identifier, Target wireless access point identification.
  • the target wireless access point After the target access point is successfully switched, the target wireless access point will face the access user.
  • the measurement report is reported to the data transmission control in the manner described in the above related embodiments. node.
  • the information management unit 304 of the data transmission control node receives the access handover notification message After that, all the TCPs of this access user under this source wireless access point in the MPTCP connection information table will be The wireless access point information of the substream is changed from the source wireless access point information to the target wireless access point information. interest.
  • the sending unit 501 is further configured to: Reporting an MPTCP connection message to the data transmission control node, where the MPTCP connection is The information is used to indicate the status of the MPTCP connection, and the MPTCP connection message carries MPTCP connection information, the MPTCP connection information is used for the data transmission control node Associating the plurality of TCP substreams with the MPTCP connection such that the MPTCP connection The plurality of TCP substreams are included.
  • the sending unit 501 may further include a user identification subunit 5011, configured to Identify the MPTCP connection status of the access user, and generate a connection corresponding to the MPTCP connection status. MPTCP connection message.
  • a user identification subunit 5011 configured to Identify the MPTCP connection status of the access user, and generate a connection corresponding to the MPTCP connection status. MPTCP connection message.
  • the user identification sub-unit 5011 may be specifically configured to analyze the carrying of the access user. TCP packets with MPTCP signaling, which are identified by special fields in the data packet
  • the access user is an MPTCP user and obtains an MPTCP connection message of the access user. specifically, Obtain relevant signaling of the MPTCP connection of the access user, parse the signaling and form a phase
  • the MPTCP connection message should be.
  • the MPTCP connection message is used to indicate the MPTCP connection The state of the connection.
  • the user identification subunit 5011 parses the current access user's carrying MPTCP. Signaling TCP packet, if the access user is found to have completed the initial connection establishment of MPTCP In the process, the sending unit 501 reports the MPTCP initial connection establishment notification message to the data transmission. Input control node.
  • the method by which the network node determines that the MPTCP initial connection establishment process is completed may be: The network node obtains and parses the acknowledgement message carrying the MP_CAPABLE option (ACK, Acknowledgement); or by the network node to get and parse carrying MP_CAPABLE The three-step handshake process of the option, which is carried in turn to carry the MP_CAPABLE option.
  • TCP SYN packet corresponding to the TCP SYN packet carrying the MP_CAPABLE option SYN/ACK
  • ACK corresponding to SYN/ACK carrying the MP_CAPABLE option.
  • the decision process can be performed by the user identification sub-unit 5011.
  • the user identification subunit 5011 parses the current access user's carrying MPTCP. Signaling TCP packets, if the access user is found to have completed the MPTCP connection address increase In the notification process, the sending unit 501 reports the address increase notification message to the data transmission control. System node.
  • the method for determining that the MPTCP address increase notification process is completed by the network node may be: obtaining Take the TCP packet of the access user and parse out the ADD_ADDR option.
  • the judgment The flow can be performed by the user identification sub-unit 5011.
  • the user identification subunit 5011 parses the current access user's carrying MPTCP. Signaling TCP packet, if a user is found to have completed the MPTCP subflow new process, then The sending unit 501 reports the new TCP sub-flow notification message to the data transmission control node.
  • the method for determining that the MPTCP sub-flow new process is completed by the network node may be sequentially obtained. And parsing the TCP SYN packet carrying the MP_JOIN option, and the corresponding ACK message; It can also be to acquire and parse the 4-step handshake process that carries the MP_JOIN option, that is, capture sequentially.
  • the decision process can be The user identification sub-unit 5011 executes.
  • the user identification subunit 5011 parses the TCP packet of the current access user. If an access user is found to have completed the MPTCP substream deletion process, the sending unit 501 The TCP substream notification message is deleted to the data transmission control node.
  • the method for the network node to determine that the MPTCP substream deletion process is completed may be: acquiring and Parse the TCP RST packet; you can also get and parse the TCP FIN packet; or you can get the solution Parse the REMOVE_ADDR option in the TCP packet.
  • the user identification subunit 5011 parses the current access user's carrying MPTCP. Signaling TCP packets, if an access user is found to have completed the MPTCP subflow priority In the new process, the sending unit 501 reports the TCP substream priority update notification message to the data transmission. Input control node.
  • the method for the network node to determine that the MPTCP sub-flow priority update process is completed may be: Obtain and parse the MP_PRIO option contained in the TCP packet of the access user.
  • the data transmission control node and the network node provided by the embodiment of the present invention are based on a system
  • the system needs to select the TCP subflow that needs to be offloaded by the data, by changing the MPTCP congestion. Control the input of the algorithm so that the data on the selected TCP substream can be migrated to the expected genus On other TCP sub-flows connected to the same MPTCP, including the management control policy of the operator. Reflecting the congestion status of the data transmission link carrying the TCP substream and the status of the radio resource Wait.
  • By adding control measures to MPTCP data transmission on the access network side data is divided. Streaming results can be adapted to the expected needs of system operation, optimize system performance, and improve MPTCP The practicability of the congestion control algorithm.
  • FIG. 6 is a schematic block diagram of a data transmission control node according to another embodiment of the present invention.
  • the data transmission control node is configured to control data between two ends of the communication of the MPTCP connection Transmission, the MPTCP connection includes multiple transmission control protocol TCP substreams, and the data transmission
  • the control node includes a memory 601, a processor 602, and a transceiver 603.
  • the memory 601 stores instructions that cause the processor 602 to:
  • the network node corresponding to the at least one TCP substream adjusts the at least according to the adjustment policy a data transmission process on a TCP substream, triggering both ends of the communication to execute the MPTCP a congestion control algorithm
  • the output of the MPTCP congestion control algorithm includes determining Part or all of the data transmitted on the at least one TCP substream migrates to belong to the MPTCP is connected to other TCP substreams.
  • the transceiver 603 can be configured to send the adjustment policy to the at least one TCP The network node corresponding to the subflow.
  • the sending process of the foregoing adjustment policy includes: traversing the air interface downlink or uplink control substream The set sends an adjustment policy to the network node where each TCP subflow in the control subflow set is located.
  • the memory 601 may also store fingers that cause the processor 602 to perform the following operations Determining: reducing a data transmission rate of the at least one TCP substream and the data transmission The magnitude of the reduction in the rate of transmission.
  • the speed is reduced to a certain range. After the degree, the TCP substream will not meet the performance requirements of data transmission, and both ends of the communication can Detecting or obtaining the rate information of the TCP substream by the network node, and starting MPTCP congestion
  • the control algorithm switches the TCP substream. For details, refer to the correlation of the embodiment shown in FIG. 3. Description, do not repeat here.
  • the memory 601 may also store fingers that cause the processor 602 to perform the following operations Deleting: the network node is instructed to delay sending the data packet on the at least one TCP substream to Increasing the RTT, the delay time corresponds to the magnitude of the decrease in the data transmission rate.
  • the memory 601 may also store fingers that cause the processor 602 to perform the following operations Let: adjusting the policy may include instructing the network node to discard the at least one TCP substream The service data packet on the packet to increase the packet loss rate, discarding the number of the service data packet and the number Corresponding to the magnitude of the decrease in transmission rate.
  • the above process of changing the RTT and the packet loss rate can be performed simultaneously, that is, comprehensive application RTT and packet loss rate change the data transmission rate of the TCP substream; or separately, the present invention
  • the embodiment does not particularly limit this.
  • the packet rate is used as an input to the MPTCP congestion algorithm, which affects the output of the algorithm, so that each The data on the TCP substream can complete the migration process that meets the expected requirements.
  • the specific method of changing the packet loss rate or RTT can refer to the correlation in the embodiment shown in FIG. Description, do not repeat here.
  • the memory 601 may also store fingers that cause the processor 602 to perform the following operations Order: According to the carrier management and control strategy, the link status of the communication link, and the MPTCP connection At least one condition of the radio resource status of the network node corresponding to each TCP substream is selected Select the TCP subflow.
  • the transceiver 603 is configured to receive the measurement report reported by the network node. Obtaining the link status of the communication link of each TCP substream, and/or corresponding to each TCP substream The wireless resource status of the network node, the specific content of the measurement report can be seen in Figure 3 The relevant description in the example is not described here.
  • the memory 601 can also store instructions that cause the processor 602 to perform the following operations: MPTCP connection message, which will belong to multiple TCP sub-flows of the same MPTCP connection and Associated with an MPTCP connection, the MPTCP connection message is used to indicate the MPTCP connection status.
  • the MPTCP connection message may include: indicating that the MPTCP initial connection is established.
  • the MPTCP initial connection establishment notification message completed by the process; and the indication at the beginning of MPTCP After the initial connection establishment process is completed, a new TCP subflow belonging to the MPTCP connection is newly created. TCP subflow notification message.
  • the method further includes: adding an address increase notification message of the available address; Including, deleting the deleted TCP sub-circulation of one or more TCP sub-streams currently in use Knowing the message; may also include a TCP sub-flow optimization indicating a change in the usage priority of the TCP substream Pre-update notification message, etc., wherein changing the TCP sub-flow usage priority means changing a certain The backup status of one or more TCP substreams.
  • the method may further include: deleting the deletion of a currently used TCP substream TCP sub-flow notification message, TCP sub-flow priority indicating change of TCP sub-flow usage priority Update notification messages, etc., where changing the TCP substream usage priority means changing the current usage The backup status of the TCP substream.
  • the data transmission control node can receive multiple different types of network nodes.
  • Reported MPTCP information of multiple different TCP substreams belonging to the same MPTCP connection For example, receiving at least one TCP that is reported by the BTS, the NodeB, and the AP through the network node.
  • Substream's MPTCP information For the specific reporting manner, refer to the phase in the embodiment shown in FIG. 3. The content is not described here.
  • the memory 601 can also store instructions that cause the processor 602 to:
  • the MPTCP connection information table may include that is reported by different network nodes.
  • MPTCP connection messages for the same MPTCP connection may also include different network sections Point to report MPTCP connection messages about different MPTCP connections, you can use MPTCP
  • the connection is a classification basis, and each TCP substream is managed.
  • the data transmission control node receives the MPTCP connection containing different content
  • the above MPTCP connection table can be maintained accordingly, and the corresponding
  • the relevant description in the description will not be repeated here.
  • each TCP can be obtained.
  • the sub-flow is associated with the MPTCP connection.
  • the relevant description in the description will not be repeated here.
  • the data transmission control node may further include an antenna 604 and a bus system. 605 and so on.
  • the processor 602 controls the operation of the device, and the processor 602 may also be referred to as a CPU. (Central Processing Unit).
  • Memory 601 can include read only The memory and random access memory provide instructions and data to processor 602. Memory A portion of 601 may also include non-volatile random access memory (NVRAM).
  • NVRAM non-volatile random access memory
  • the transceiver 603 can be coupled to the antenna 605.
  • the various components of the device pass the total Line system 605 is coupled together, wherein bus system 605 can include, in addition to a data bus To include the power bus, control bus and status signal bus. But for clarity See, various buses are labeled as bus system 605 in the figure.
  • FIG. 7 is a schematic block diagram of a network node according to an embodiment of the present invention, including storage The device 701, the processor 702, and the transceiver 703.
  • a transceiver 703 configured to receive a data transmission control node in the multiple TCP substreams Select the information of at least one TCP substream obtained. And for receiving the data transmission An adjustment policy indicated by the control node, where the adjustment policy is calculated for the MPTCP congestion control The adjustment strategy of the input of the law.
  • the input item of the MPTCP congestion control algorithm includes: the at least one The data round-trip delay RTT and/or packet loss rate of the TCP substream.
  • the memory 701 can also store instructions that cause the processor 702 to:
  • Adjusting the at least one according to the foregoing adjustment policy sent by the data transmission control node The data transmission process on the TCP substream triggers the two ends of the communication to perform the MPTCP congestion a control algorithm, the output of the MPTCP congestion control algorithm includes determining Part of the data transmitted on one TCP substream is partially or completely migrated to belong to the MPTCP connection. Connected to other TCP substreams.
  • the memory 701 may also store fingers that cause the processor 702 to perform the following operations Deciling: reducing a data transmission rate of the at least one TCP substream and the data transmission speed The rate of decrease in rate.
  • the memory 701 may also store fingers that cause the processor 702 to perform the following operations Let: discard the service data packet on the at least one TCP substream to increase the packet loss rate, The number of discarded service data packets corresponds to the magnitude of the data transmission rate reduction.
  • the data transmission speed of a specific TCP substream can be reduced. Rate, trigger the communication ends to change the data transmission path, and transmit data through other TCP substreams. Guarantee the quality of data transmission.
  • the transceiver 703 is further configured to send the adjusted input to the At both ends of the communication, the adjusted input is adjusted by the processor 702 to the at least one TCP Obtained after the data transmission process on the substream, used to trigger the two ends of the communication to perform the MPTCP congestion control algorithm.
  • the RTT and the packet loss rate may be detected by the two ends of the communication.
  • the algorithm input item is not specifically limited in this embodiment of the present invention.
  • the memory 701 can also store instructions that cause the processor 702 to:
  • the MPTCP connection message is used to indicate a status of the MPTCP connection.
  • the memory 701 can also store instructions that cause the processor 702 to:
  • the measurement report is used to indicate Link status of the communication link of the plurality of TCP substreams and/or the plurality of TCP substream pairs
  • the radio resource status of the intended network node the measurement report is used for the data transmission control
  • the node selects the at least one TCP substream.
  • the transceiver 703 can also be used to The data transmission control node reports an MPTCP connection message, and the MPTCP connection message Used to indicate the status of the MPTCP connection, which is carried in the MPTCP connection message MPTCP connection information, the MPTCP connection information is used for the data transmission control node Associating the plurality of TCP substreams with the MPTCP connection such that the MPTCP connection The plurality of TCP substreams are included.
  • the processor 702 is further configured to identify an MPTCP connection of the access user. State, generating an MPTCP connection message corresponding to the MPTCP connection state.
  • the specific identification mode, the type, content and reporting mode of the MPTCP connection message can be The related description of other embodiments of the present invention is not described herein.
  • the network node may also include an antenna 704, a bus system 705, etc., a specific connection For the relationship, reference may be made to the related description in the embodiment shown in FIG. 6, and details are not described herein again.
  • the data transmission control node and the network node provided by the embodiment of the present invention are based on a system
  • the system needs to select the TCP subflow that needs to be offloaded by the data, by changing the MPTCP congestion. Control the input of the algorithm so that the data on the selected TCP substream can be migrated to the expected genus On other TCP sub-flows connected to the same MPTCP, including the management control policy of the operator. Reflecting the congestion status of the data transmission link carrying the TCP substream and the status of the radio resource Wait.
  • By adding control measures to MPTCP data transmission on the access network side data is divided. Streaming results can be adapted to the expected needs of system operation, optimize system performance, and improve MPTCP The practicability of the congestion control algorithm.
  • FIG. 8 is a schematic diagram of a communication system according to an embodiment of the present invention. Including, network node 801, data transmission control node 802, both of which may be wired or not The line link performs direct or indirect communication.
  • a data transmission control node 801 configured to select, in the multiple TCP substreams At least one TCP substream controlled by the data transmission; determining the transmission of the MPTCP congestion control algorithm An adjustment policy of the entry; sending the adjustment policy to the at least one TCP substream Network node, the adjustment policy is used to instruct the network node to adjust the at least one piece
  • the data transmission process on the TCP substream triggers the two ends of the communication to perform the MPTCP congestion a control algorithm, the output of the MPTCP congestion control algorithm includes determining Part of the data transmitted on one TCP substream is partially or completely migrated to belong to the MPTCP connection. Connected to other TCP substreams;
  • a network node 802 located in a communication link at both ends of the communication, for receiving data At least one TCP substream selected by the transmission control node in the plurality of TCP substreams Information; receiving an adjustment policy indicated by the data transmission control node, the adjustment strategy An adjustment strategy for an entry for the MPTCP congestion control algorithm.
  • the data transmission control node 801 in the communication system may be implemented as shown in FIG. 3 or FIG.
  • the data transmission control node in the example, the network node 802 can be the embodiment shown in FIG. 5 or FIG. Network node in which the implementation of the embodiment shown in Figures 9-11 can be performed
  • Data transmission control method and TCP sub-flow association method the details of which can be referred to the present invention The relevant description of his embodiment is not described here.
  • the communication system provided by the embodiment of the invention is selected according to the system operation requirements. Performing a TCP subflow of data offloading by changing the input of the MPTCP congestion control algorithm, Enable the data on the selected TCP substream to migrate to the same MPTCP connection as expected On other TCP substreams. By adding control measures for MPTCP data transmission on the access network side Therefore, the data shunting result on the TCP substream can be adapted to the expected demand of the system operation, and optimized. System performance improves the usability of the MPTCP congestion control algorithm.
  • the embodiment of the invention provides a data transmission management method, which can be represented by FIG. 3 or FIG. 6
  • the data transmission control node in the illustrated embodiment is executed to control the multipath transmission control association Data transmission between the two ends of the communication of the MPTCP connection, the MPTCP connection includes multiple Strip Transmission Control Protocol TCP Substream.
  • Figure 9 is a schematic flow chart of the method, including steps S901-S903:
  • the data transmission control node selects, in the multiple TCP substreams, a number to be performed. According to at least one TCP substream controlled by the transmission.
  • the data transmission control node may decide which air ports on the TCP substream are selected.
  • the downlink or uplink data is migrated to other TCP substreams that belong to the same MPTCP connection, and
  • the TCP substream that needs to be migrated is classified into the air interface down or on the air transmission direction.
  • Manage the subflow set may be used to decide which air ports on the TCP substream are selected.
  • the data transmission control node determines an input of an MPTCP congestion control algorithm. Adjust the strategy.
  • the data transmission control node in S902 determines an MPTCP congestion control algorithm.
  • the adjustment policy of the input of the method includes: the data transmission control node determines to decrease the The data transmission rate of one less TCP substream and the extent to which the data transmission rate is reduced.
  • the TCP layer that is an input of the MPTCP congestion control algorithm can be changed.
  • RTT Real Time
  • packet loss rate adjusted RTT and/or The packet loss rate is used as an input to the algorithm, which affects the output of the algorithm, so that each TCP substream is The data can be completed to meet the expected needs of the migration process.
  • adjusting the policy may include instructing the network node to delay sending the at least Packets on a TCP substream to increase RTT, delay time and the data transmission rate The magnitude of the decrease corresponds.
  • adjusting the policy may include instructing the network node to discard the at least one piece
  • the service data packet on the TCP substream increases the packet loss rate and discards the number of the service data packets.
  • adjusting the policy may include instructing the network node to discard the at least one piece
  • the service data packet on the TCP substream increases the packet loss rate and discards the number of the service data packets.
  • the above process of changing the RTT and the packet loss rate can be performed simultaneously, that is, comprehensive application RTT and packet loss rate change the data transmission rate of the TCP substream; or separately, the present invention
  • the embodiment does not particularly limit this.
  • the specific method for changing the packet loss rate or RTT can be referred to the relevant embodiment of the embodiment shown in FIG. Rong, do not repeat here.
  • the data transmission control node indicates the network corresponding to the at least one TCP substream.
  • the node adjusts data transmission on the at least one TCP substream according to the adjustment policy End, the two ends of the communication are triggered to execute the MPTCP congestion control algorithm, the MPTCP
  • the output of the congestion control algorithm includes determining to transmit on the at least one TCP substream The data is partially or completely migrated to other TCP substreams belonging to the MPTCP connection.
  • the sending process of the foregoing adjusting policy includes: traversing the air interface downlink or uplink pipe
  • the control sub-flow set delivers adjustments to the network node where each TCP sub-flow in the control sub-flow set is located.
  • Strategy There are multiple ways to decide which of the TCP substreams to choose for the air interface to go down or up.
  • the data is migrated to other TCP sub-flows that belong to the same MPTCP connection, and the embodiment of the present invention is This is not particularly limited. For a specific data migration method, reference may be made to the phase of the embodiment shown in FIG. 3. The description is not repeated here.
  • the data transmission control node is in the Selecting at least one TCP subflow packet that needs to perform data transmission control among multiple TCP substreams Including, according to the operator's management and control strategy, the link status of the communication link, and the MPTCP connection.
  • At least one condition of the radio resource status of the network node corresponding to each TCP substream is selected Select the TCP substream for shunting.
  • the link status and/or radio resource status of the communication link It can be obtained by related information in the measurement report reported by the network node. Specific content can be referenced The related description of the embodiment shown in FIG. 3 is not described herein.
  • the data transmission control node root Associating the plurality of TCP substreams with the MPTCP connection according to an MPTCP connection message, Causing the MPTCP connection to include the plurality of TCP substreams, the MPTCP connection message Used to indicate the status of the MPTCP connection.
  • the MPTCP connection message may include: indicating that the MPTCP initial connection is established.
  • the MPTCP initial connection establishment notification message completed by the process; and the indication at the beginning of MPTCP After the initial connection establishment process is completed, a new TCP subflow belonging to the MPTCP connection is newly created. TCP subflow notification message.
  • the method may further include: after the initial connection establishment process of the MPTCP is completed, Add an address increase notification message of the available address to the multiple access responder of the new TCP substream.
  • the method may further include: deleting the deletion of a currently used TCP substream TCP sub-flow notification message, TCP sub-flow priority indicating change of TCP sub-flow usage priority Update notification messages, etc., where changing the TCP substream usage priority means changing the current usage The backup state of the TCP substream.
  • each TCP subflow associated with the data transmission control node is associated with a corresponding
  • the process of MPTCP connection is as shown in FIG. 10, including steps S1001-1004:
  • S1002 Analyze a new TCP subflow notification message reported by the network node, and determine the The originating end of the TCP subflow new process is which of the two ends of the MPTCP connection communication End, obtain the address of the current TCP substream and the Tokens of both ends of the subflow new process.
  • S1004 Search for information corresponding to the MPTCP connection in the MPTCP connection information table, Determine the MPTCP connection to which the new TCP subflow belongs.
  • the response end Token of the TCP subflow new process is searched for The Tokens of one end of the MPTCP connection are equal, and the response end of the TCP subflow new process is The address is equal to an address in the available address set at one end of the MPTCP connection.
  • the newly created TCP substream belongs to the MPTCP connection.
  • Steps can be taken for each TCP subflow managed by the data transfer control node S1001-S1004 judging process, thereby dividing multiple TCP sub-links belonging to the same MPTCP connection
  • the stream is associated with the MPTCP connection.
  • the data transmission control method provided by the embodiment of the present invention is based on system operation requirements. Select the TCP substream that needs to be offloaded by changing the MPTCP congestion control algorithm. Inputs so that the data on the selected TCP substream can be migrated to the expected one MPTCP is connected to other TCP substreams. By adding MPTCP data on the access network side The control measures of the transmission enable the data shunting result on the TCP substream to adapt to the pre-operation of the system. Demand, optimize system performance, and improve the practicability of the MPTCP congestion control algorithm.
  • the embodiment of the invention further provides a data transmission management method for controlling multipath Data transmission between the two ends of the communication control protocol MPTCP connection, the MPTCP The connection contains multiple Transmission Control Protocol TCP substreams.
  • Figure 11 is a schematic flow chart of the method, As shown in the figure, both ends of the communication can be user equipment and servers, including steps S1101-S1107:
  • the network node identifies whether the access user belongs to the initiator end of the MPTCP connection or Receiving end.
  • the MPTCP connection message is used to indicate the MPTCP connection Status, through the special field in the packet, can identify whether the access user is MPTCP user.
  • An access user refers to a logical unit corresponding to a UE accessing a communication network through an access network.
  • the specific content of the MPTCP connection message and the status of the MPTCP connection For the specific determination method, refer to the related description in the corresponding device embodiment, and no longer Said.
  • the network node reports the MPTCP connection to the data transmission control node.
  • the MPTCP connection message is used to indicate a status of the MPTCP connection,
  • the MPTCP connection message carries MPTCP connection information, and the MPTCP connection information is used.
  • the data transmission control node associates the plurality of TCP substreams with the MPTCP connection And causing the MPTCP connection to include the plurality of TCP substreams.
  • MPTCP information of multiple different TCP substreams connected by MPTCP For example, receiving BTS at the same time, An MPTCP message reported by the NodeB and the AP through at least one TCP substream of the network node. interest.
  • specific reporting method refer to related content in the device embodiment, and no longer ⁇ Said.
  • step S1102' may also be included:
  • the network node reports a measurement report to the data transmission management node, where the measurement report is used Indicates the link status and/or radio resource status of the communication link.
  • the sub-flow is reported to the measurement report for the network node, or only if the access user has a certain When the TCP sub-network related to the TCP TCP connection is reported, the measurement report for the access user is reported.
  • Steps S1102 and S1102' do not have a limitation of the execution order, and may also execute S1102' first. Then execute S1102.
  • the data transmission control node stores the MPTCP connection information table, and obtains according to the The MPTCP connection message is updated, and the network section is included in the MPTCP connection information table. Point the reported MPTCP connection message.
  • the MPTCP connection information table may include that is reported by different network nodes.
  • MPTCP connection messages for the same MPTCP connection may also include different network sections Point reported MPTCP connection messages for different MPTCP connections. Connect to MPTCP as Based on the classification, each TCP substream is managed.
  • the above MPTCP connection table performs corresponding maintenance, and corresponds to various types of MPTCP connection messages.
  • body processing process reference may be made to the related description in the device embodiment, and details are not described herein.
  • the stored MPTCP connection can be updated when the measurement status of the wireless resource status of the network node is reported. Relevant network node information and/or access user information in the information table.
  • S1104 The data transmission control node belongs to the received MPTCP connection message according to The plurality of TCP substreams of the MPTCP connection are associated with the MPTCP connection.
  • the MPTCP connection message further includes deleting a TCP sub-flow notification message and/or Or TCP substream priority update notification message.
  • deleting a TCP sub-flow notification message and/or Or TCP substream priority update notification message Corresponding deletion of TCP subflow process and TCP sub
  • the flow priority update process may refer to the related description in the device embodiment, and is no longer referred to herein. Said.
  • the communication ends of the MPTCP connection are one-way multiple access, one side single address, and A scenario in which a TCP subflow is created by a single-address party, or a scenario where both ends are multiple-access
  • the multiple access end can First initiate the address increase process, telling the peer that it needs to increase when creating a new TCP subflow. Available address. If the network node resolves the access user, the above address increase notification is completed. Process, then report the address increase notification message to the data transmission control node, including the new multiple access The TCP substream responder needs to increase the available address.
  • step S1004 may also include, in step S1104', adding a new TCP subflow in the MPTCP connection information table.
  • the available address of the responder which is the multiple access initiator of the MPTCP initial connection or Multiple access responder.
  • the premise of congestion control for TCP substreams is that they will belong to the same MPTCP connection.
  • the TCP substream is associated with the MPTCP connection, that is, the current UE needs to be clarified for data communication. Which TCP TCP stream is used and which network node is connected to, etc. Information, and confirm the status of the TCP subflow, in order to be different for the same MPTCP connection.
  • the TCP substream performs congestion control, and data is split between different TCP substreams.
  • the data transmission control node selects the plurality of TCP substreams to be performed. At least one TCP substream controlled by data transmission.
  • S1106 The data transmission control node sends the determined adjustment policy to the network node.
  • the network node performs the foregoing adjustment policy, and adjusts the at least one TCP sub- The data transfer process on the stream.
  • the network node may delay sending data on the at least one TCP substream Packet to increase the RTT, and the delay time corresponds to a magnitude of the data transmission rate reduction; And/or discarding the service data packet on the at least one TCP substream to increase the packet loss Rate, the number of discarded service packets and the magnitude of the data transmission rate reduction should.
  • the network node will adjust the input obtained after the data transmission process of the TCP substream
  • the item is sent to both ends of the communication, and the MPTCP congestion control algorithm is executed at both ends of the trigger communication.
  • the adjusted input is adjusted by the network node to the at least one TCP sub Obtained after the data transfer process on the stream.
  • the output result of the MPTCP congestion control algorithm executed at both ends of the above communication includes determining Part or all of the data transmitted on the at least one TCP substream migrates to belong to the On the other TCP substream of the MPTCP connection, where the adjustment policy is for MPTCP The adjustment strategy of the input of the congestion control algorithm.
  • the specific implementation of the foregoing adjustment policy may be determined by the data transmission control node.
  • the form is sent to the corresponding network node for direct execution; it can also be controlled by the data transmission control node. Only the TCP sub-flow migration policy is sent to the corresponding network node, and the network node decides itself. What specific form is to be used to perform data migration.
  • the data transmission control method provided by the embodiment of the present invention is based on system operation requirements. Select the TCP substream that needs to be offloaded by changing the MPTCP congestion control algorithm. Inputs so that the data on the selected TCP substream can be migrated to the expected one Other TCP sub-flows connected to the MPTCP, including the management control strategy of the operator, Congestion status of the data transmission link carrying the TCP substream, and radio resource status. by Add control measures for MPTCP data transmission on the access network side to enable data splitting results Adapt to the expected requirements of system operation, optimize system performance, and improve MPTCP congestion control The practicality of the algorithm. At the same time, there is no need to change the congestion control algorithm of the existing protocol, by changing The input of the algorithm indirectly affects the output of the algorithm, and the expected data shunting result is obtained, and the process is simple. Easy to implement.
  • the form of the software product is embodied, the computer software product is stored in a storage medium Including a number of instructions to make a computer device (can be a personal computer, service) Server, or network device, etc.) performing all or part of the methods described in various embodiments of the present invention Step by step.
  • the foregoing storage medium includes: a U disk, a mobile hard disk, a read only memory (ROM), Various types of random access memory (RAM), disk or optical disk that can store program code medium.
  • the disclosed apparatus may be implemented in other ways without departing from the scope of the present application.
  • the device embodiments described above are merely illustrative, for example, the modules or The division of units is only a logical function division, and there may be another scheme in actual implementation.
  • a sub-mode such as multiple units or components, may be combined or integrated into another system. Or some features can be ignored or not executed.
  • the unit described as a separate component May be or may not be physically separate, the component displayed as a unit may be or Can also be not a physical unit, that is, it can be located in one place, or it can be distributed On a network unit.
  • Some or all of the modules can be selected according to actual needs. The purpose of implementing the solution of this embodiment is achieved. Those skilled in the art do not pay creative labor In the case of that, it can be understood and implemented.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Environmental & Geological Engineering (AREA)
  • Computer Security & Cryptography (AREA)
  • Mobile Radio Communication Systems (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

本发明实施例提供了一种数据传输控制节点、***及数据传输管理方法，依据***运行需求选择需要进行数据分流的TCP子流，通过改变MPTCP拥塞控制算法的输入项，使选择出的TCP子流上数据能够迁移到预期的属于同一MPTCP连接的其他TCP子流上。通过在接入网侧增加对MPTCP数据传输的管控措施，使TCP子流上数据分流结果能够适应***运行的预期需求，优化***性能，提升MPTCP拥塞控制算法的实用性。

Description

数据传输控制节点、通信***及数据传输管理方法 技术领域
本发明实施例涉及通信技术领域，尤其涉及一种数据传输控制节 点、通信***及数据传输管理方法。
背景技术
多种无线宽带接入技术的日益发展，使得一个用户设备(UE， User Equipment)可以同时具有到目标节点的多条数据传输链路。多 路径数据传输协议(MPTCP，Multi-Path Transmission Control  Protocol)在传统TCP协议的基础上发展而来，可以为用户终端提供 端到端的多链路通信，以资源共享的方式，把数据流分发到多条链路 上来提高网络带宽。
为了提升数据传输质量，MPTCP协议中定义了一种称为耦合的 拥塞控制(Coupled Congestion Control)的拥塞控制算法，其预期目 标之一就是平衡拥塞，即尽可能将数据从最拥塞的链路上分流到其他 链路上。采用现有协议规定的端到端的拥塞控制算法，虽然可以在不 同通信链路间进行数据分流，但数据分流结果不能充分适应***运行 的各种预期需求，进而无法使得***性能达到最优化，拥塞控制算法 的实用性较差。
发明内容
本发明实施例提供了一种数据传输控制节点、***及数据传输管 理方法，能够使得MPTCP数据分流结果适应***运行的预期需求。
第一方面，本发明实施例提供了一种数据传输控制节点，用于控 制多路径传输控制协议MPTCP连接的通信两端之间的数据传输，所 述MPTCP连接包含多条传输控制协议TCP子流，所述数据传输控制节 点包括，处理单元，用于在所述多条TCP子流中选择需要进行数据传 输控制的至少一条TCP子流；所述处理单元还用于，确定MPTCP拥塞 控制算法的输入项的调整策略；发送单元，用于将所述调整策略发送 给所述至少一条TCP子流对应的网络节点，所述调整策略用于指示所 述网络节点调整所述至少一条TCP子流上的数据传输过程，触发所述 通信两端执行所述MPTCP拥塞控制算法，所述MPTCP拥塞控制算法 的输出结果包括确定在所述至少一条TCP子流上传输的数据部分或 全部迁移到属于所述MPTCP连接的其他TCP子流上，其中，所述网络 节点位于所述通信两端的通信链路中。
在第一方面的第一种可能的实现方式中，所述处理单元用于确定 MPTCP拥塞控制算法的输入项的调整策略，包括，确定降低所述至 少一条TCP子流的数据传输速率以及所述数据传输速率降低的幅度。
结合第一方面的以上任意一种可能的实现方式，所述MPTCP拥 塞控制算法的输入项包括，所述至少一条TCP子流的数据往返时延 RTT和/或丢包率。
结合第一方面的以上任意一种可能的实现方式，所述发送单元用 于指示所述网络节点调整所述至少一条TCP子流上的数据传输过程 以获得调整后的输入项，包括，指示所述网络节点延迟发送所述至少 一条TCP子流上的数据包以增大所述RTT，延迟时间与所述数据传输 速率降低的幅度对应；和/或，指示所述网络节点丢弃所述至少一条 TCP子流上的业务数据包以增大所述丢包率，丢弃所述业务数据包的 数量与所述数据传输速率降低的幅度对应。
结合第一方面的以上任意一种可能的实现方式，所述处理单元用 于在所述多条TCP子流中选择需要进行数据传输控制的至少一条 TCP子流，包括，根据运营商管控策略选择所述至少一条TCP子流。
结合第一方面的以上任意一种可能的实现方式，所述处理单元用 于在所述多条TCP子流中选择需要进行数据传输控制的至少一条 TCP子流，包括，根据所述多条TCP子流对应的网络节点的无线资源 状况确定需要进行数据传输控制的网络节点，选择所述网络节点对应 的正在进行数据传输的至少一条TCP子流。
结合第一方面的以上任意一种可能的实现方式，所述处理单元用 于在所述多条TCP子流中选择需要进行数据传输控制的至少一条 TCP子流，包括，判断所述多条TCP子流所在的通信链路的链路状况 是否满足预设条件，在不满足所述预设条件的TCP子流中，选择正在 进行数据传输的至少一条TCP子流。
结合第一方面的以上任意一种可能的实现方式，还包括接收单 元，用于接收所述网络节点上报的测量报告，所述测量报告用于指示 所述多条TCP子流的通信链路的链路状况和/或所述多条TCP子流对 应的网络节点的无线资源状况；所述处理单元具体用于根据所述测量 报告，选择所述至少一条TCP子流。
结合第一方面的以上任意一种可能的实现方式，还包括信息管理 单元，用于根据MPTCP连接消息，将所述多条TCP子流与所述MPTCP 连接相关联，使得所述MPTCP连接包含所述多条TCP子流，所述 MPTCP连接消息用于指示所述MPTCP连接的状态。
结合第一方面的以上任意一种可能的实现方式，所述信息管理单 元还用于，存储并更新MPTCP连接信息表，所述MPTCP连接信息表 中包括所述网络节点上报的MPTCP连接消息中携带的MPTCP连接信 息，所述MPTCP连接信息用于将所述多条TCP子流与所述MPTCP连 接相关联。
结合第一方面的以上任意一种可能的实现方式，所述通信两端中 的一端是接入所述网络节点的用户设备UE，另一端是服务器；所述 网络节点支持所述UE与所述服务器相通信。
第二方面，本发明实施例提供了一种网络节点，用于控制多路径 传输控制协议MPTCP连接的通信两端之间的数据传输，所述MPTCP 连接包含多条传输控制协议TCP子流，所述网络节点位于所述通信两 端的通信链路中，包括，
接收单元，用于接收数据传输控制节点在所述多条TCP子流中选 择得到的至少一条TCP子流的信息；
所述接收单元还用于，接收所述数据传输控制节点指示的调整策 略，所述调整策略为针对MPTCP拥塞控制算法的输入项的调整策略。
策略执行单元，用于根据所述调整策略，调整所述至少一条TCP 子流上的数据传输过程，触发所述通信两端执行所述MPTCP拥塞控 制算法，所述MPTCP拥塞控制算法的输出结果包括确定在所述至少 一条TCP子流上传输的数据部分或全部迁移到属于所述MPTCP连接 的其他TCP子流上。
在第二方面的第一种可能的实现方式中，所述网络节点还包括发 送单元，用于将调整后的输入项发送给所述通信两端，所述调整后的 输入项由所述策略执行单元调整所述至少一条TCP子流上的数据传 输过程后获得，用于触发所述通信两端执行所述MPTCP拥塞控制算 法。
结合第二方面的以上任意一种可能的实现方式，所述MPTCP拥 塞控制算法的输入项的调整策略包括，降低所述至少一条TCP子流的 数据传输速率以及所述数据传输速率降低的幅度。
结合第二方面的以上任意一种可能的实现方式，所述MPTCP拥 塞控制算法的输入项包括，所述至少一条TCP子流的数据往返时延 RTT和/或丢包率。
结合第二方面的以上任意一种可能的实现方式，所述策略执行单 元用于调整所述至少一条TCP子流上的数据传输过程，包括，延迟发 送所述至少一条TCP子流上的数据包以增大所述RTT，延迟时间与所 述数据传输速率降低的幅度对应；和/或，丢弃所述至少一条TCP子流 上的业务数据包以增大所述丢包率，丢弃所述业务数据包的数量与所 述数据传输速率降低的幅度对应。
结合第二方面的以上任意一种可能的实现方式，所述发送单元还 用于，向所述数据传输控制节点上报测量报告，所述测量报告用于指 示所述多条TCP子流的通信链路的链路状况和/或所述多条TCP子流 对应的网络节点的无线资源状况，所述测量报告用于所述数据传输控 制节点选择所述至少一条TCP子流。
结合第二方面的以上任意一种可能的实现方式，所述发送单元还 用于，向所述数据传输控制节点上报MPTCP连接消息，所述MPTCP 连接消息用于指示所述MPTCP连接的状态，所述MPTCP连接消息中 携带MPTCP连接信息，所述MPTCP连接信息用于所述数据传输控制 节点将所述多条TCP子流与所述MPTCP连接相关联，使得所述 MPTCP连接包含所述多条TCP子流。
结合第二方面的以上任意一种可能的实现方式，所述通信两端中 的一端是接入所述网络节点的用户设备UE，另一端是服务器；所述 网络节点支持所述UE与所述服务器相通信。
第三方面，本发明实施例提供了一种通信***，包括，
数据传输控制节点，用于在所述多条TCP子流中选择需要进行数 据传输控制的至少一条TCP子流；确定MPTCP拥塞控制算法的输入项 的调整策略；将所述调整策略发送给所述至少一条TCP子流对应的网 络节点，所述调整策略用于指示所述网络节点调整所述至少一条TCP 子流上的数据传输过程，触发所述通信两端执行所述MPTCP拥塞控 制算法，所述MPTCP拥塞控制算法的输出结果包括确定在所述至少 一条TCP子流上传输的数据部分或全部迁移到属于所述MPTCP连接 的其他TCP子流上；
网络节点，位于所述通信两端的通信链路中，用于接收数据传输 控制节点在所述多条TCP子流中选择得到的至少一条TCP子流的信 息；接收所述数据传输控制节点指示的调整策略，所述调整策略为针 对MPTCP拥塞控制算法的输入项的调整策略。
第四方面，本发明实施例还提供了一种数据传输管理方法，其特 征在于，用于控制多路径传输控制协议MPTCP连接的通信两端之间 的数据传输，所述MPTCP连接包含多条传输控制协议TCP子流，所述 方法包括，数据传输控制节点在所述多条TCP子流中选择需要进行数 据传输控制的至少一条TCP子流；所述数据传输控制节点确定MPTCP 拥塞控制算法的输入项的调整策略；所述数据传输控制节点指示所述 至少一条TCP子流对应的网络节点根据所述调整策略，调整所述至少 一条TCP子流上的数据传输过程，触发所述通信两端执行所述MPTCP 拥塞控制算法，所述MPTCP拥塞控制算法的输出结果包括确定在所 述至少一条TCP子流上传输的数据部分或全部迁移到属于所述 MPTCP连接的其他TCP子流上。
在第四方面的第一种可能的实现方式中，所述数据传输控制节点 确定MPTCP拥塞控制算法的输入项的调整策略包括，所述数据传输 控制节点确定降低所述至少一条TCP子流的数据传输速率以及所述 数据传输速率降低的幅度。
结合第四方面的任意一种可能的实现方式，所述MPTCP拥塞控 制算法的输入项包括，所述至少一条TCP子流的数据往返时延RTT和 /或丢包率。
结合第四方面的任意一种可能的实现方式，所述数据传输控制节 点指示所述网络节点调整所述至少一条TCP子流上的数据传输过程 包括，所述数据传输控制节点指示所述网络节点延迟发送所述至少一 条TCP子流上的数据包以增大所述RTT，延迟时间与所述数据传输速 率降低的幅度对应；和/或，所述数据传输控制节点指示所述网络节 点丢弃所述至少一条TCP子流上的业务数据包以增大所述丢包率，丢 弃所述业务数据包的数量与所述数据传输速率降低的幅度对应。
结合第四方面的任意一种可能的实现方式，所述数据传输控制节 点在所述多条TCP子流中选择需要进行数据传输控制的至少一条 TCP子流，包括，所述数据传输控制节点根据运营商管控策略选择所 述至少一条TCP子流。
结合第四方面的任意一种可能的实现方式，所述数据传输控制节 点在所述多条TCP子流中选择需要进行数据传输控制的至少一条 TCP子流，包括，所述数据传输控制节点根据所述多条TCP子流对应 的网络节点的无线资源状况确定需要进行数据传输控制的网络节点， 选择所述网络节点对应的正在进行数据传输的至少一条TCP子流。
结合第四方面的任意一种可能的实现方式，所述数据传输控制节 点在所述多条TCP子流中选择需要进行数据传输控制的至少一条 TCP子流包括，所述数据传输控制节点判断所述多条TCP子流所在的 通信链路的链路状况是否满足预设条件，在不满足所述预设条件的 TCP子流中，选择正在进行数据传输的至少一条TCP子流。
结合第四方面的任意一种可能的实现方式，所述数据传输控制节 点在所述多条TCP子流中选择需要进行数据传输控制的至少一条 TCP子流，包括，所述数据传输控制节点接收所述网络节点上报的测 量报告，所述测量报告用于指示所述多条TCP子流的通信链路的链路 状况和/或所述多条TCP子流对应的网络节点的无线资源状况；所述数 据传输控制节点根据所述测量报告，选择所述至少一条TCP子流。
结合第四方面的任意一种可能的实现方式，所述方法进一步包 括，所述数据传输控制节点根据MPTCP连接消息，将所述多条TCP 子流与所述MPTCP连接相关联，使得所述MPTCP连接包含所述多条 TCP子流，所述MPTCP连接消息用于指示所述MPTCP连接的状态。
结合第四方面的任意一种可能的实现方式，所述数据传输控制节 点存储并更新MPTCP连接信息表，所述MPTCP连接信息表中包括所 述网络节点上报的MPTCP连接消息中携带的MPTCP连接信息，所述 MPTCP连接信息用于将所述多条TCP子流与所述MPTCP连接相关 联。
结合第四方面的以上任意一种可能的实现方式，所述通信两端中 的一端是接入所述网络节点的用户设备UE，另一端是服务器；所述 网络节点支持所述UE与所述服务器相通信。
第五方面，本发明实施例还提供了一种数据传输管理方法，其特 征在于，用于控制多路径传输控制协议MPTCP连接的通信两端之间 的数据传输，所述MPTCP连接包含多条传输控制协议TCP子流，网络 节点位于所述通信两端的通信链路中，所述方法包括，所述网络节点 接收数据传输控制节点在所述多条TCP子流中选择得到的至少一条 TCP子流的信息，所述网络节点根据数据传输控制节点指示的调整策 略，调整所述至少一条TCP子流上的数据传输过程，触发所述通信两 端执行所述MPTCP拥塞控制算法，所述MPTCP拥塞控制算法的输出 结果包括确定在所述至少一条TCP子流上传输的数据部分或全部迁 移到属于所述MPTCP连接的其他TCP子流上，其中，所述调整策略为 针对MPTCP拥塞控制算法的输入项的调整策略。
在第五方面的第一种可能的实现方式中，在所述网络节点调整所 述至少一条TCP子流上的数据传输过程之后还包括，所述网络节点将 调整后的输入项发送给所述通信两端，所述调整后的输入项由所述网 络节点调整所述至少一条TCP子流上的数据传输过程后获得，用于触 发所述通信两端执行所述MPTCP拥塞控制算法。
结合第五方面的以上任意一种可能的实现方式，所述MPTCP拥 塞控制算法的输入项的调整策略包括，降低所述至少一条TCP子流的 数据传输速率以及所述数据传输速率降低的幅度。
结合第五方面的以上任意一种可能的实现方式，所述MPTCP拥 塞控制算法的输入项包括，所述至少一条TCP子流的数据往返时延 RTT和/或丢包率。
结合第五方面的以上任意一种可能的实现方式，所述网络节点调 整所述至少一条TCP子流上的数据传输过程包括，所述网络节点延迟 发送所述至少一条TCP子流上的数据包以增大所述RTT，延迟时间与 所述数据传输速率降低的幅度对应；和/或，所述网络节点丢弃所述 至少一条TCP子流上的业务数据包以增大所述丢包率，丢弃所述业务 数据包的数量与所述数据传输速率降低的幅度对应。
结合第五方面的以上任意一种可能的实现方式，所述方法还包 括，所述网络节点向所述数据传输控制节点上报测量报告，所述测量 报告用于指示所述多条TCP子流的通信链路的链路状况和/或所述多 条TCP子流对应的网络节点的无线资源状况，所述测量报告用于所述 数据传输控制节点选择所述至少一条TCP子流。
结合第五方面的以上任意一种可能的实现方式，所述方法进一步 包括，所述网络节点向所述数据传输控制节点上报MPTCP连接消息， 所述MPTCP连接消息用于指示所述MPTCP连接的状态，所述MPTCP 连接消息中携带MPTCP连接信息，所述MPTCP连接信息用于所述数 据传输控制节点将所述多条TCP子流与所述MPTCP连接相关联，使得 所述MPTCP连接包含所述多条TCP子流。
结合第五方面的以上任意一种可能的实现方式，所述通信两端中 的一端是接入所述网络节点的用户设备UE，另一端是服务器；所述 网络节点支持所述UE与所述服务器相通信。
采用本发明实施例提供的技术方案，依据***运行需求选择需要 进行数据分流的TCP子流，通过改变MPTCP拥塞控制算法的输入项， 使选择出的TCP子流上数据能够迁移到预期的属于同一MPTCP连接 的其他TCP子流上。通过在接入网侧增加对MPTCP数据传输的管控措 施，使TCP子流上数据分流结果能够适应***运行的预期需求，优化 ***性能，提升MPTCP拥塞控制算法的实用性。
附图说明
图1是一种MPTCP协议栈的示意图；
图2是本发明实施例的一种应用场景示意图；
图3是本发明实施例提供的一种数据传输控制节点的示意框图；
图4是一种MPTCP初始连接建立过程与TCP子流新建过程的信令 流程的示意图；
图5是本发明实施例提供的一种数据传输控制节点的示意框图；
图6是本发明实施例提供的另一种数据传输控制节点的示意框 图；
图7是本发明实施例提供的另一种网络节点的示意框图；
图8是本发明实施例提供的一种通信***示意图；
图9是本发明实施例提供的一种数据传输管理方法流程示意图；
图10是本发明实施例提供的一种将TCP子流关联到MPTCP连接 的方法流程示意图；
图11是本发明实施例提供的另一种数据传输管理方法流程示意 图。
具体实施方式
为使本发明的目的、技术方案和优点更加清楚，下面将结合附图 对本发明实施方式作进一步详细描述。
本文中描述的各种技术可用于多种通信***，例如当前2G、3G 通信***和下一代通信***，例如全球移动通信***(GSM，Global  System for Mobile Communication)，码分多址(CDMA，Code Division  Multiple Access)***，时分多址(TDMA，Time Division Multiple  Access)***，宽带码分多址(WCDMA，Wideband Code Division  Multiple Access)***，长期演进(LTE，Long Term Evolution)*** 及后续演进***等；以及无线局域网(WLAN，Wireless Local Area  Network)、全球微波互联接入(WiMax，Worldwide Interoperability for  Microwave Access)***等其他无线宽带接入网络。
需要说明的是，本发明实施例中所述的网络节点可以是接入设 备，例如可以是GSM或CDMA中的基站收发台(BTS，Base Transceiver  Station)，也可以是WCDMA中的节点B(NodeB)，也可以是LTE中的演 进型节点B(e-NodeB，evolved NodeB)，还可以是WLAN中的接入点 (AP，Access Point)，或者其他通信或无线网络***中具有类似功能的 网元，本发明实施例对此不做任何限定。网络节点也可以是接入控制 设备，例如可以是GSM或CDMA中的基站控制器(BSC，Base Station  Controller)，可以是WCDMA中的无线网络控制器(RNC，Radio  Network Controller)，也可以是WLAN中的接入控制器(AC，Access  Controller)，也可以是其他具有类似功能的网元，本发明实施例对此 不做任何限定。
需要说明的是，本发明实施例中提出的数据传输控制节点是表示 接入网侧用于管控MPTCP数据传输的单元，其部署位置不限，可以 单独部署，也可以与网管、接入控制设备、接入设备共同部署，换言 之，该控制节点可以是独立的设备，也可以部署在其他通信设备中， 本发明实施例对此不做任何限定。
例如，数据传输控制节点可以作为一个控制单元部署在上述 BSC、RNC或AC等接入控制设备中；数据传输控制节点也可以作为 一个控制单元部署在上述BTS、NodeB、e-NodeB、AP等接入设备中。 又例如，数据传输控制节点也可以是独立部署的设备，并能够与作为 网络节点的上述基站设备或基站控制设备进行信息交互。
本发明实施例出现的用户设备(UE，User Equipment)可以是以 无线方式接入接入网进行数据传输的用户设备，例如移动电话，具备 无线通信功能的个人计算机等。
需要说明的是，本发明实施例中所述的接入网包括由BTS与 BSC、或NodeB与RNC、或e-NodeB等设备组成的无线接入网(RAN， Radio Access Network)，以及由AP与AC等设备组成的宽带接入网。 其中，AC与RNC可以进行融合，即AC的功能集成在RNC上，并与AP 进行通信。
图1是MPTCP协议栈的示意图，相对标准TCP(standard TCP)而 言，MPTCP协议栈主要是将TCP层分为MPTCP层和TCP子流(TCP Subflow)层，通过如下方式实现数据并发传输，包括：建立多个TCP 子流，相当于建立多条TCP连接；由MPTCP层对多个TCP子流进行数 据分流或汇聚等处理。TCP子流与所属的MPTCP连接可以通过令牌 (Token)联系起来。Token是MPTCP连接的标识，只具有本地意义 和唯一性，MPTCP初始连接建立过程的发起端和响应端均要生成一 个Token，Token可以看作是MPTCP连接标识(MPTCP Connection ID)。
MPTCP连接可以由多条TCP子流构成。TCP子流是在某一特定的 传输路径(path)上进行传输的TCP数据流。其中，传输路径是指存 在于MPTCP连接的通信两端之间的通路，可以通过所述通信两端的 一对地址，例如IP地址，和/或端口号(port)来标识，每条TCP子流 都有对应的路径。
MPTCP是一个端到端的协议，因此通信两端需要交互控制信令 来完成诸如MPTCP初始连接建立、TCP子流增加或删除、地址通知等 功能。这些控制信令都是通过TCP选项(TCP Option)来传输的。
需要说明的是，本发明实施例中涉及到的具体MPTCP信令的含 义及功能可以参见RFC6824协议的规定，本说明书中不做过多赘述。
图2是本发明实施例的一种应用场景示意图。如图2所示，UE与 Server间的MPTCP连接有2条TCP子流，一条TCP子流在LTE链路上传 输，该LTE链路上包括接入网侧节点e-NodeB、核心网侧节点独立业 务网关(SGW，Service Gateway)及分组数据网网关(PGW，Packet  Data Network Gateway)。另外一条TCP子流在Wi-Fi链路上传输，该 Wi-Fi链路上包括接入网侧节点AP及AC；核心网侧节点宽带远程接入 服务器(BRAS，Broadband Remote Access Server)。如图2所示，本 发明实施例提供的数据传输控制节点可以独立设置，并分别与 e-NodeB及AC之间存在链路连接，便于进行信息交互。e-NodeB和AP 可以分别向数据传输控制节点上报MPTCP连接消息，MPTCP连接消 息由e-NodeB或AP解析UE与Server之间的MPTCP信令交互过程获得， 例如解析接入用户的携带有MPTCP信令的TCP数据包获得，数据传输 控制节点根据MPTCP连接消息将属于该MPTCP连接的不同TCP子流 关联起来，并可以根据运营商策略、或接入用户接入的接入网侧节点 的无线资源状况、或接入用户所在的链路状况等条件来确定数据在不 同TCP子流间的分配策略，最终将分配策略下发给e-NodeB及AP执 行。
本发明实施例提供了一种数据传输控制节点，可以用于控制 MPTCP连接的通信两端之间的数据传输，所述MPTCP连接包含多条 传输控制协议TCP子流。
其中，MPTCP连接的通信两端可以是空口上行或空口下行数据 传输过程中的两端通信设备，任意一端均可以发起或响应MPTCP连 接的相关过程，包括MPTCP初始连接建立过程、TCP子流新建过程、 TCP子流删除过程、TCP子流优先级更新过程等。例如，由接入某无 线接入点的UE发起MPTCP初始连接建立过程，网络侧的Server进行 响应，则UE与Server成为MPTCP连接的通信两端；类似地，也可以 由Server发起MPTCP初始连接建立过程，UE进行响应。需要说明的 是，所述MPTCP初始连接建立过程的发起端或响应端与TCP子流新建 过程的发起端或响应端、及后续数据传输过程的发送端或接收端相互 之间不存在对应关系，例如，MPTCP初始连接建立过程的发起端或 响应端均可以发起TCP子流新建过程，MPTCP初始连接建立过程的 发起端或响应端、TCP子流新建过程的发起端或响应端中的任意一端 均可以发送数据。
该数据传输控制节点的示意框图如图3所示，主要包括：
处理单元301，可以用于在所述多条TCP子流中选择需要进行数 据传输控制的至少一条TCP子流。
具体地，处理单元301可以决定选择哪些TCP子流上的空口下行 或上行的数据迁移到同属一个MPTCP连接的其他TCP子流上，并将需 要进行数据迁移的TCP子流按数据传输方向归入空口下行或上行管 控子流集。
所述处理单元301还可以用于，确定MPTCP拥塞控制算法的输入 项的调整策略。
发送单元302，用于将所述调整策略发送给所述至少一条TCP子 流对应的网络节点，所述调整策略用于指示所述网络节点调整所述至 少一条TCP子流上的数据传输过程，触发所述通信两端执行所述 MPTCP拥塞控制算法，所述MPTCP拥塞控制算法的输出结果包括确 定在所述至少一条TCP子流上传输的数据部分或全部迁移到属于所 述MPTCP连接的其他TCP子流上。
其中，所述网络节点位于所述通信两端的通信链路中。
上述调整策略的发送过程包括，遍历空口下行或上行管控子流 集，向管控子流集中的每个TCP子流所在的网络节点下发调整策略。
可选地，可以有多种途径决定选择哪些TCP子流上的空口下行或 上行的数据迁移到同属一个MPTCP连接的其他TCP子流上，本发明实 施例对此不做特别限定。例如可以采用以下两种方法：
方法一：首先，根据MPTCP连接信息表中的无线接入点的空口 下行或上行无线资源状态，得到需要控制的无线接入点，例如选择空 口下行或上行负载拥塞的无线接入点，可选地，可以结合运营商管控 策略进行判断。其次，遍历每个需要控制的无线接入点下的所有TCP 子流，如果当前TCP子流所属的MPTCP连接存在空口下行或上行 Backup标识位为0的其他TCP子流，且其他TCP子流所在无线接入点 非当前无线接入点，且其他TCP子流所在无线接入点的空口下行或上 行无线资源状态为空闲，则选定当前TCP子流进入空口下行或上行管 控子流集。
方法二：遍历MPTCP连接信息表中所有的TCP子流。如果当前子 流所属接入用户的空口链路信息表明此链路的空口下行或上行不适 合数据传输，例如当前TCP子流所在的链路质量差或者接入用户处于 小区边缘，且当前TCP子流所属的MPTCP连接存在空口下行或上行 Backup标识位为0的其他TCP子流，且其他TCP子流所在无线接入点 非当前TCP子流的无线接入点，且其他TCP子流所属的接入用户的空 口链路信息表明空口下行或上行适合数据传输，则选定当前TCP子流 进入空口下行或上行管控子流集。
可选地，可以限定管控子流集中的TCP子流的数量。
可选地，可以参考用户优先级来优先选择当前无线接入点上高优 先级用户对应的TCP子流进入管控子流集。
数据迁移可以是将TCP子流的一部分数据转移到同属于一个 MPTCP连接的其他TCP子流上，也可是将TCP子流的全部数据转移到 同属于一个MPTCP连接的其他TCP子流上，对数据迁移的比例可以按 照***运行的优化目的确定，本发明实施例不做特别限定。
可选地，作为本发明的另一个实施例，处理单元301用于确定确 定MPTCP拥塞控制算法的输入项的调整策略，包括确定降低所述至 少一条TCP子流的数据传输速率以及所述数据传输速率降低的幅度。
具体地，当某条通信两端正在使用TCP子流的速度降低到一定程 度后，该TCP子流将不满足数据传输的性能要求，通信两端可以自行 检测或由网络节点处获取该TCP子流的速率信息，并启动MPTCP拥塞 控制算法，切换TCP子流。
可选地，可以由处理单元301确定所述调整策略的具体实施形式， 由发送单元302下发给对应的网络节点直接执行；处理单元301也可以 仅给出TCP子流迁移策略并下发给对应的网络节点，由网络节点自行 决定采取何种具体形式来执行数据迁移。例如，分流策略可以仅包括 某一个TCP子流的速率需要降低到哪种程度，可以由网络节点自行决 定采用何种具体形式，例如延时发送或增大丢包率等方法来满足该分 流策略；类似地，调整策略中也可以直接指示网络节点通过延时发送 或增大丢包率来降低TCP子流的速率。
可选地，可以改变作为MPTCP拥塞控制算法输入项的TCP层的往 返时延(RTT，Round Trip Time)和/或丢包率，将调整后的RTT和/ 或丢包率作为MPTCP拥塞算法的输入项，影响算法的输出结果，使 得各条TCP子流上的数据能够完成满足预期需求的迁移过程。
可选地，调整策略可以包括指示所述网络节点延迟发送所述至少 一条TCP子流上的数据包以增大RTT，延迟时间与所述数据传输速率 降低的幅度对应。
可选地，调整策略可以包括指示所述网络节点丢弃所述至少一条 TCP子流上的业务数据包以增大丢包率，丢弃所述业务数据包的数量 与所述数据传输速率降低的幅度对应。
上述改变RTT和丢包率的处理过程可以同时进行，即综合运用 RTT和丢包率改变TCP子流的数据传输速率；或者单独进行，本发明 实施例对此不做特别限定。
具体地，现有协议规定的MPTCP拥塞控制算法的数据分流调度 结果主要取决于RTT和丢包率。因此，可以以各通信链路的RTT和丢 包率作为输入来影响MPTCP拥塞控制算法输出结果，例如通过增大 某条链路上的丢包率，使得在所述拥塞控制算法的作用下将数据迁移 到另一条链路上。因此，上述调整策略可以通过改变TCP层的RTT和 /或丢包率，间接影响上述算法的输出结果，使得数据的分流结果能 够满足***运行的不同需求。
可以理解，本发明实施例所使用的MPTCP拥塞控制算法是现有 协议规定的Coupled Congestion Control算法，算法运算过程可以参照 协议相关内容，例如每个TCP子流中的慢启动、快速重传、快速恢复 等项目与现有算法一致，具体参见RFC6356，在此不做赘述。
以下描述几种改变丢包率或RTT的具体方法，仅作示意之用，不 构成对本发明实施例的任何限定，本领域的普通技术人员可以了解到 采取其他方式也可以改变丢包率或RTT。
例如，可以指示网络节点通过主动丢弃TCP子流上的空口下行或 上行的业务数据包来增大丢包率，例如增大丢包率到某一指定比例， 或每隔多少个数据包主动丢弃一个数据包。
例如，可以指示网络节点通过延迟发送子流上的空口上行或下行 ACK包来增大TCP层的RTT，例如指定子流上的空口上行或下行ACK 包延迟一定时间再发送。
例如，可以指示网络节点通过推迟子流上的空口下行或上行数据 包进入调度器来增大TCP层的RTT，例如指定TCP子流上的数据延迟 一段时间后进入调度器。
例如，可以指示网络节点影响调度器行为来增大TCP层的RTT， 例如通过不调度TCP子流的空口下行或上行数据，或者是降低TCP子 流的空口下行或上行调度优先级到某一程度，或者将调度器感知TCP 子流的下行或上行数据调度需求的时间点延迟一定数值。
作为本发明的另一个实施例，所述处理单元301可以具体用于， 根据运营商管控策略、通信链路的链路状况、无线资源状况中的至少 一个条件选择TCP子流。
可选地，所述处理单元301用于根据运营商管控策略选择所述至 少一条TCP子流。
其中，所述运营商管控策略包括***负载平衡，链路资源利用效 率以及计费策略等体现运营商对整个网络的管控倾向的项目。以LTE 和Wi-Fi数据并发的场景举例说明：处于LTE网络边缘的UE的通信链 路质量不好，运营商可能倾向于在频谱效率较高的Wi-Fi链路上传输 更多的数据，但通信链路质量及频谱效率与链路的RTT或丢包率没有 直接的对应关系，因此仅凭LTE链路和Wi-Fi链路上客观存在的RTT 和丢包率作为数据分流调度的依据，不能保证数据准确分流到Wi-Fi 链路上；再例如，考虑到分类计费的原因，运营商可能希望UE尽可 能在LTE链路上传输更多的数据，如上例所述，以客观的RTT和丢包 率作为分流依据，不能保证实现运营商的管控目的。因此，通过改变 LTE链路和Wi-Fi链路的RTT和/或丢包率并达到一定的值，可以影响 MPTCP拥塞控制算法输出，将数据迁移到相应链路上，满足运营商 的管控需求。
上述运营商策略的可以直接在数据传输控制节点上进行配置，或 者通过其他网元下发，例如由核心网中的控制设备制定并下发，本发 明实施例对此不做特别限定。
可选地，处理单元301还可以用于根据所述多条TCP子流对应的 网络节点的无线资源状况确定需要进行数据传输控制的网络节点，选 择所述网络节点对应的正在进行数据传输的至少一条TCP子流。
可选地，处理单元301还可以用于判断所述多条TCP子流所在的 通信链路的链路状况是否满足预设条件，在不满足所述预设条件的 TCP子流中，选择正在进行数据传输的至少一条TCP子流。
可选地，数据传输控制节点还可以包括接收单元303，用于接收 所述网络节点上报的测量报告，所述测量报告用于指示所述多条TCP 子流的通信链路的链路状况和/或所述多条TCP子流对应的网络节点 的无线资源状况；所述处理单元301具体用于根据所述测量报告，选 择所述至少一条TCP子流。
具体地，所述无线资源状况可以根据网络节点上报的面向无线接 入点的测量报告分析得到，测量报告中可以包括该无线接入点的资源 状态。所述通信链路的链路状况可以根据网络节点上报的面向接入用 户的测量报告分析得到，测量报告中可以包括接入用户的空口链路信 息。
其中，接入用户是指通过接入网成功接入通信网络的UE在接入 网中对应的逻辑对象，作为MPTCP连接的一端的UE接入通信网络并 进行实际通信时，可以将该UE视为接入用户，特别地，接入不同通 信网络并在不同通信网络均存在TCP子流的同一个UE被认为是在相 应的通信网络下对应不同的接入用户，在相应的通信网络中有各自的 用户标识。例如，UE在3GPP网络中存在无线链路，则UE在3GPP网 络中对应一个接入用户，可以采用IMSI或GUTI或GCI+小区内用户标 识等方法来标识此接入用户；UE在WLAN网络中存在无线链路，则 UE在WLAN网络中对应一个接入用户，可以采用MAC地址或AP ID+IP地址或AC ID+IP地址或SSID+IP地址等方法来标识此接入用 户。
可以了解，上述三种条件可以任意组合，或者单独作为选择TCP 子流的依据。同时，选择何条TCP子流进行分流处理并不限定在上述 三种条件，可以根据***运行的需求选择合适的参考量作为制定依 据，本发明实施例对此不做特别限定。
可选地，作为本发明的另一个实施例，该数据传输控制节点还可 以包括信息管理单元304，可以用于根据MPTCP连接消息，将属于同 一个MPTCP连接的多条TCP子流与该MPTCP连接相关联，所述 MPTCP连接消息用于指示所述MPTCP连接的状态。
所述MPTCP连接消息可以包括，指示MPTCP初始连接建立过程 完成的MPTCP初始连接建立通知消息；以及指示在MPTCP初始连接 建立过程完成后新建属于该MPTCP连接的TCP子流的新建TCP子流 通知消息。
可选地，还可以包括增加可用地址的地址增加通知消息；还可以 包括，指示删除当前使用的某一条或多条TCP子流的删除TCP子流通 知消息；还可以包括，指示改变TCP子流的使用优先级的TCP子流优 先级更新通知消息等，其中，改变TCP子流使用优先级是指改变某一 条或多条TCP子流的备份状态。
接收单元303还可以用于，接收网络节点上报的MPTCP连接消 息，所述MPTCP连接消息由所述网络节点解析接入用户的携带 MPTCP信令的TCP数据包获得。
不同类型的MPTCP连接消息携带了与对应的MPTCP连接状态相 关的信息。
例如，所述MPTCP初始连接建立通知消息的携带内容可以包括 但不限于以下信息：
当前TCP子流所在的无线接入类型，例如是3GPP网络或WLAN 网络；当前TCP子流所在的无线接入点的标识，例如对于无线接入类 型是3GPP网络，可以用全球小区标识(GCI，Global Cell Identity)表 示，对于无线接入类型是WLAN网络，可以用AP ID或AC ID表示； 当前TCP子流所属的接入用户的全球唯一标识，例如，对于3GPP网 络中的接入用户来说，可以采用IMSI或GUTI或GCI+小区内用户标识 等方法来标识；对于WLAN网络中的接入用户来说，可以采用MAC 地址或AP ID+IP地址或AC ID+IP地址或SSID+IP地址等方法来标 识；当前TCP子流的地址信息，此信息的表现形式可以是当前TCP子 流的两端地址，即从接入用户的TCP数据包的包头中获取到的IP地址 与端口号(Port)；当前TCP子流的两端的Key或由Key计算得到的 Token。
所述新建TCP子流通知消息的携带内容可以包括但不限于以下 信息：
当前TCP子流所在的无线接入的类型；当前TCP子流所在的无线 接入点的标识；当前TCP子流所属的接入用户的全球唯一标识；当前 TCP子流的地址信息，此信息的表现形式可以是当前TCP子流的两端 地址；当前TCP子流的两端的Address_ID；TCP子流新建过程的响应 端的Token，即TCP SYN包中MP_JOIN选项携带的Token；当前TCP 子流的哪一端发起子流新建过程，例如，Server与UE建立了MPTCP 连接，网络节点如果识别出是Server发送的携带MP_JOIN选项的TCP SYN包，则指示是Server发起子流新建过程，反之则指示是UE发起 TCP子流新建过程；当前TCP子流的空口下行和空口上行的Backup标 识，Backup标志位由携带MP_JOIN选项的MPTCP信令所指示，表示 该MPTCP信令的发送端是否希望对端在向该MPTCP信令发送端发送 数据时将该TCP子流用作备份。因此，当前TCP子流的空口上行的 Backup标识指的是TCP子流新建过程中携带MP_JOIN选项的空口下 行信令所指示的Backup标识；当前TCP子流的空口下行的Backup标识 指的是TCP子流新建过程中携带MP_JOIN选项的空口上行信令所指 示的Backup标识。以上述Server发起TCP子流新建过程为例，Server 向UE发送的携带MP_JOIN选项的SYN包指示了Backup标识位为1，代 表TCP子流的空口上行的Backup标识位为1，表明Server指示UE在后 续向Server发送数据时，将该条新建TCP子流视为备份子流。类似地， 在UE向Server反馈的SYN/ACK消息中也会携带Backup标识位，用以 指示Server向UE发送数据时使用的TCP子流的备份状态。
所述地址增加通知消息的携带内容可以包括但不限于以下信息： 从ADD_ADDR选项中解析出来的需要增加的地址信息，ADD_ADDR 选项由当前TCP子流的一端发送，此地址信息的形式是 ADDRESS_ID+可用地址，可以是ADDRESS_ID+IP地址及Port，或者 是ADDRESS_ID+IP地址，哪种形式取决于ADD_ADDR选项中的内容 格式是否携带Port；当前TCP子流的地址信息，此信息的表现形式可 以是当前TCP子流的两端地址；当前TCP子流所属的MPTCP连接的哪 一端需要增加可用地址，例如，Server与UE建立了MPTCP连接，网 络节点识别出是Server发送的ADD_ADDR选项，则指示Server需要增 加可用地址。
所述删除TCP子流通知消息按照网络节点判定MPTCP子流删除 过程完成的方法不同可以携带不同的信息内容，包括：
通过获取并解析TCP RST包或TCP FIN包判定MPTCP子流删除 过程完成，则删除TCP子流通知消息携带的内容可以包括但不限于以 下信息，当前TCP子流，即承载TCP RST包或者是TCP FIN包的TCP 子流的地址信息，此信息的表现形式可以是当前TCP子流的两端地 址。
通过获取并解析TCP数据包的REMOVE_ADDR选项判定 MPTCP子流删除过程完成，则删除TCP子流通知消息携带的内容可以 包括但不限于以下信息，从REMOVE_ADDR选项中解析出来的需要 删除的ADDRESS_ID集合；当前TCP子流，即承载REMOVE_ADDR 选项的TCP子流的地址信息，此信息的表现形式可以是当前TCP子流 的两端地址；当前TCP子流所属的MPTCP连接的哪一端需要删除地 址，例如，Server与UE建立了MPTCP连接，如果网络节点识别出是 Server发送的REMOVE_ADDR选项，则指示是Server需要删除可用地 址。
所述TCP子流优先级更新通知消息按照网络节点根据接入用户 的TCP数据包中的MP_PRIO选项判定MPTCP子流优先级更新过程完 成情况，可以携带不同的信息内容，包括，
情况一：如果MP_PRIO选项如果不带ADDRESS_ID，则说明需 要改变当前TCP子流，即承载MP_PRIO选项的子流的优先级，此时 TCP子流优先级更新通知消息携带内容可以包括但不限于以下信息， 当前TCP子流的地址信息；表明子流的优先级更新方向的信息，此信 息的表现形式可以为指示当前TCP子流的某一端发送的MP_PRIO选 项，也可以直接为指示需要更新的是空口下行或空口上行Backup标 识；从MP_PRIO选项中解析出来的Backup标识位的取值。例如，如 果是Server发送的MP_PRIO选项，则指示上行Backup标识需要更新， 所述更新后的上行Backup标识用于表示Server指示UE在后续向Server 发送数据时，是否将当前TCP子流视为备份子流；如果是UE发送的 MP_PRIO选项，则指示下行Backup标识需要更新，所述更新后的下 行Backup标识用于表示UE指示Server在后续向UE发送数据时，是否 将当前TCP子流视为备份子流。
情况二：MP_PRIO选项如果携带了ADDRESS_ID，则是说明优 先级更新应用于此MPTCP连接上所有使用了ADDRESS_ID所对应地 址的TCP子流，此时TCP子流优先级更新通知消息携带内容可以包括 但不限于以下信息：当前TCP子流，即承载MP_PRIO选项的TCP子流 的地址信息，此信息的表现形式可以是当前TCP子流的两端的地址； 从MP_PRIO选项中解析出来的ADDRESS_ID；表明子流的优先级更 新方向的信息，此信息的表现形式可以为指示是当前TCP子流的某一 端发送的MP_PRIO选项，也可以直接为指示需要更新的是空口下行 或空口上行Backup标识；从MP_PRIO选项中解析出来的Backup标识 位的取值。具体过程与情况一类似，在此不做赘述。
应理解，承载TCP子流的空口链路的类型可以相同，也可以不相 同，假设某个MPTCP连接包含两条TCP子流，则这两条TCP子流可以 都建立在LTE链路上，或者一条TCP子流建立在LTE链路上，另一条 TCP子流建立在Wi-Fi链路上。本发明实施例对同一MPTCP连接包含 的TCP子流的数量及链路类型不做特别限定，可以根据网络实际运行 需求确定。
可选地，接收单元303可以接收多个不同类型的网络节点上报的 属于同一个MPTCP连接的多个不同TCP子流的MPTCP消息。例如同 时接收BTS、NodeB与AP上报的经过该网络节点的至少一条TCP子流 的MPTCP消息。可选地，网络节点可以是基站设备，则管理不同网 络节点的基站控制设备与所述数据传输控制节点之间可以建立通信 链路，用于传输测量信息和控制信息。例如，假设数据传输控制节点 集成在RNC中，则NodeB可以直接上报经NodeB解析的MPTCP连接消 息给RNC，而BTS或AP则可以将各自获取的MPTCP连接消息上报给 对应的BSC及AC，由BSC及AC根据与上述RNC之间的链接关系，将 接收到的MPTCP连接消息转发给上述RNC，由RNC完成对该MPTCP 连接的管理。可选地，也可以直接在各网络节点与数据传输控制节点 之间建立通信链路，则各网络节点可以直接向数据传输控制节点上报 相关消息，无需通过相关控制设备转发。此外，如果网络节点与数据 传输控制节点集成在一个网络设备中，例如集成在e-NodeB中，与上 述场景类似，只需在与e-NodeB对应层级的其他网络设备之间建立通 信链路以进行信息交互，在此不再赘述。
图4是MPTCP初始连接建立过程与TCP子流新建过程的信令流程 图。其中，MPTCP初始连接建立过程包括建立属于该MPTCP连接的 第一条TCP子流，经过三次握手过程建立MPTCP连接，与建立普通 TCP连接的区别在于，TCP SYN包、SYN/ACK包、ACK包会携带 MP-CAPABLE选项，且选项中会携带64bit键值(Key)。
在MPTCP初始连接建立过程完成后，MPTCP连接的任意一端可 以采用一对当前没有使用的地址来新建一个TCP子流。如图4所示， HostA和HostB之间新建了一条A2<->B1的TCP子流，A2是HostA的可 用地址，B1是HostB的可用地址。新建TCP子流过程会携带MP_JOIN 选项完成图4中后4步的信令交互过程，其中TCP SYN包中携带的 MP_JOIN选项的包含32位(bit)的Token、32bit随机数、ADDRESS_ID 和Backup标识。其中，Token-B是TCP子流新建过程的发起端用响应 端的64bit键值通过一定的加密算法，例如哈希算法生成的32bit值， TCP子流新建过程的发起端通过在TCP SYN包中携带此参数来标识 该TCP子流期望加入哪个MPTCP连接。TCP子流新建过程的发起端在 TCP SYN包中携带的32bit随机数R-A和TCP子流新建过程的响应端在 SYN/ACK包中携带的32bit随机数R-B用于进一步保证子流建立过程 的安全性。ADDRESS_ID只具有本地意义，且具有唯一性，TCP子流 新建过程的发起端在TCP SYN包中携带的ADDRESS_ID用于标识发 起端在当前新建TCP子流中用到的IP地址，TCP子流新建过程的响应 端在SYN/ACK包中携带的ADDRESS_ID用于标识响应端在当前新建 子流中用到的IP地址。Backup标志位用于代表此信息的发送端是否希 望对端将该TCP子流用作备份，通过置1，发送端请求对端仅在没有 可用的即Backup标志位为0的子流时才在此Backup标志位为1的备份 子流上发数据，例如图4中TCP子流的发起端HostA在SYN包中携带的 Backup标志位就是用于代表HostA是否希望HostB将该TCP子流用作 备份，图中TCP子流的响应端HostB在SYN/ACK包中携带的Backup标 志位就是用于代表HostB是否希望HostA将该TCP子流用作备份。
可选地，所述数据传输控制节点可以同时管理多个MPTCP连接， 对所述多个MPTCP连接的通信两端的设备类型、各MPTCP连接包含 的TCP子流所在的通信链路类型、UE接入的无线接入点的类型等项 目都不做任何限定。
可选地，作为本发明的另一个实施例，所述信息管理单元304还 可以包括存储子单元3041，用于存储并更新MPTCP连接信息表，所 述MPTCP连接信息表中包括所述网络节点上报的MPTCP连接消息中 携带的MPTCP连接信息。
具体地，信息管理单元304解析出接收到的各类MPTCP连接消息 中携带的相关MPTCP连接信息，并整理成MPTCP连接信息表，所述 MPTCP连接信息包括与当前MPTCP连接状态相关的接入用户信息、 无线接入点信息等信息。
可选地，所述MPTCP连接信息表可以包括不同网络节点上报的 关于同一个MPTCP连接的MPTCP连接信息；也可以包括不同网络节 点上报的关于不同MPTCP连接的MPTCP连接信息。信息管理单元304 可以以MPTCP连接为分类依据，对各条TCP子流进行管理。
可选地，信息管理单元304如果收到MPTCP初始连接建立通知消 息(以下简称“通知消息1”)，可以通过以下操作来维护MPTCP连 接信息表，包括：
如果通知消息1中携带的是MPTCP初始连接建立过程所建立的 TCP子流(本段中称为“当前TCP子流”)的两端的键值，则数据传 输控制节点可以将上述键值计算成Token后保存在与当前TCP子流相 关的MPTCP连接信息(本段中称为“当前MPTCP连接信息”)中；如 果通知消息1中携带的是当前TCP子流的两端的Token，则直接保存在 当前MPTCP连接信息中；将通知消息1中携带的当前TCP子流所在的 无线接入类型、无线接入点的标识、所属无线接入用户的全球唯一标 识等信息保存在当前MPTCP连接信息中；将通知消息1中携带的当前 TCP子流的地址信息保存在当前MPTCP连接信息中，并且将当前 MPTCP连接的当前TCP子流的空口下行和上行的Backup标识位置为 0，并且将当前TCP子流的两端地址对应的ADDRESS_ID置为0。
可选地，信息管理单元301如果收到新建TCP子流通知消息(以 下简称“通知消息2”)，可以通过以下操作来维护MPTCP连接信息表， 包括：
将TCP子流新建过程建立的新建TCP子流关联到对应的MPTCP 连接上；将通知消息2中携带的新建TCP子流所在的无线接入类型、 无线接入点的标识、所属无线接入用户的全球唯一标识保存在新建 TCP子流所属的MPTCP连接对应的MPTCP连接信息中；将通知消息2 中携带的新建TCP子流的地址信息、新建TCP子流的空口下行及空口 上行的Backup标识保存在上述对应的MPTCP连接信息中，并且将通 知消息2中携带的新建TCP子流的两端地址和对应的ADDRESS_ID保 存在所属的MPTCP连接的通信两端的可用地址信息中。
可选地，信息管理单元301如果收到地址增加通知消息，可以通 过以下操作来维护MPTCP连接信息表，包括：根据当前TCP子流的地 址信息找到当前TCP子流所属的MPTCP连接，在所属的MPTCP连接 中保存新增的可用地址信息。
可选地，信息管理单元304如果收到删除TCP子流通知消息(以 下简称“通知消息3”)，可以通过以下操作来维护MPTCP连接信息表， 包括：
如果通知消息3中携带的是当前TCP子流的地址信息，即发送 REMOVE_ADDR选项的TCP子流，且未携带ADDRRESS_ID，则根 据当前TCP子流的地址信息找到所属的MPTCP连接，在所属的 MPTCP连接中删除该TCP子流的相关信息；如果消息中携带了需要删 除的ADDRESS_ID选项集合、当前TCP子流的地址信息、当前TCP子 流的哪一端需要删除地址等信息，则根据上述地址信息去找到当前 TCP子流所属的MPTCP连接；根据当前TCP子流所属的MPTCP连接 的哪一端需要删除地址以及需要删除的ADDRESS_ID集合，删除当前 TCP子流所属的MPTCP连接的两端的可用地址，并且删除当前TCP 子流所属的MPTCP连接中使用了上述ADDRESS_ID对应地址的所有 TCP子流的相关信息。
可选地，信息管理单元304如果收到TCP子流优先级更新通知消 息(以下简称“通知消息4”)，可以通过以下操作来维护MPTCP连接 信息表，包括：
如果通知消息4中携带的是当前TCP子流的地址信息、表明子流 优先级更新方向的信息、Backup标识，则根据当前TCP子流的地址信 息去找到当前TCP子流所属的MPTCP连接；刷新该MPTCP连接中待 更新子流的空口下行或者上行的Backup标识。
上述刷新所属的MPTCP连接中待更新子流的下行或者上行的 Backup标识包括：根据通知消息4中的子流优先级更新方向、 ADDRESS_ID、Backup标识去确定所属的MPTCP连接中哪些TCP子 流需要更新优先级以及如何更新。例如，如果子流优先级更新方向指 示表现为空口下行数据发送端例如Server发送的MP_PRIO选项，则搜 索当前TCP子流所属的MPTCP连接包含的所有TCP子流的信息，如果 搜索到某条TCP子流的某一端的地址与ADDRESS_ID相匹配，则用通 知消息4中的Backup标识刷新此TCP子流的空口上行Backup标识。
可选地，除了基于MP_PRIO选项可以识别TCP子流是否是备份子 流之外，数据传输节点还可以通过监控数据流进行备份子流识别，如 果网络节点发现某个MPTCP用户通过该网络节点的某个TCP子流在 空口下行或上行一段时间内没有数据，则可以认为这条TCP子流在空 口下行或上行方向的Backup标识是1。该步骤可以由信息管理单元304 执行。
所述信息管理单元304还可以包括信息管理单元3042，可以用于 在接收到网络节点上报新建TCP子流通知消息后，将属于某一MPTCP 连接的新建TCP子流关联到该MPTCP连接。
具体地，包括，解析所述新建TCP子流通知消息，判断所述TCP 子流新建过程的发起端为所述MPTCP连接的通信两端中的哪一端， 获得当前TCP子流的地址以及子流新建过程的两端Token，进而判断 TCP子流新建过程的响应端是上述MPTCP连接的通信两端中另一端， 并获得响应端地址。搜索MPTCP连接信息表，如果搜索到所述TCP 子流新建过程的响应端Token与某MPTCP连接的一端的Token相等， 且所述TCP子流新建过程的响应端地址与该MPTCP连接的一端的可 用地址集中的一个地址相等，则所述新建TCP子流属于所述MPTCP 连接。其中，所述可用地址集中包括至少一个可用地址，用于建立TCP 子流。
可选地，如果在搜索MPTCP连接信息表前已经确定TCP子流新建 过程的响应端是MPTCP连接的通信两端的其中一端，并获取该端地 址，则在进行MPTCP连接信息搜索时，可以只搜索该响应端的相关 MPTCP连接信息，无需再对对端进行搜索，节约***计算资源，提 升搜索效率。
可选地，在MPTCP连接的通信两端为一方多址、一方单址或者 双方多址的场景中，在MPTCP初始建立过程完成后，多址端可以向 对端发起地址增加通知过程，告知对端其可用地址。网络节点上报给 数据传输控制节点的地址增加通知消息中包括MPTCP连接的哪一端 需要增加可用地址、需要增加的地址信息和其对应的ADDRESS_ID。 可选地，MPTCP连接的单址端可以根据获悉的对端可用地址来发起 TCP子流新建过程。
信息管理单元3042可以对由数据传输控制节点管理的每条新建 TCP子流进行上述处理过程，则可以将属于同一MPTCP连接的多条 TCP子流关联到该MPTCP连接上，包括将上述新建TCP子流关联到所 属的MPTCP连接上。
可选地，在MPTCP连接的一端的IP地址冲突的场景下，此处的 地址冲突指的是不同子网下的通信节点的IP地址相同，例如，Host A 和Host A’各自分别属于不同的WLAN A及WLAN A’网络，WLAN  A及WLAN A’网络分别给Host A及Host A’分配私有IP地址。这样 Host A和Host A’的IP地址可能相同，这种情况即称为地址冲突。由 于数据传输控制节点可能会同时连接到WLAN A和WLAN A’网络， 这样，捕获到的WLAN A或WLAN A’网络中的Host A和Host A’的 地址可能相同。这种情况下，为了能在数据传输控制节点中区分不同 的MPTCP连接，信息管理单元304可以用于，接收网络节点上报的携 带网络标识的MPTCP连接消息，网络标识可以包括接入点名称 (APN，Access Point Name)、无线接入点标识等；在针对MPTCP 连接中根据地址所做的检索操作中，如果对同样一个IP地址检索到多 条MPTCP信息记录，则可以用上述网络标识进行匹配检索。
对MPTCP连接上传输的数据进行管控的前提是将属于同一个 MPTCP连接的TCP子流与MPTCP连接相关联，即需要明确当前UE进 行数据通信可使用的TCP子流是属于哪个MPTCP连接以及从哪个网 络节点接入等信息，并确认TCP子流的状态，才能对属于同一个 MPTCP连接的不同TCP子流进行拥塞控制，包括在不同TCP子流之间 进行数据分流，得到预期的分流效果。
本发明实施例提供了一种网络节点，用于控制多路径传输控制协 议MPTCP连接的通信两端之间的数据传输，所述MPTCP连接包含多 条传输控制协议TCP子流，所述网络节点位于所述通信两端的通信链 路中，结构如图5所示，包括：
接收单元501，用于接收数据传输控制节点在所述多条TCP子流 中选择得到的至少一条TCP子流的信息。
接收单元501，还用于接收所述数据传输控制节点指示的调整策 略，所述调整策略为针对MPTCP拥塞控制算法的输入项的调整策略。
可选地，所述MPTCP拥塞控制算法的输入项包括，所述至少一 条TCP子流的数据往返时延RTT和/或丢包率。
策略执行单元502，用于根据上述调整策略，调整所述至少一条 TCP子流上的数据传输过程，触发所述通信两端执行所述MPTCP拥塞 控制算法，所述MPTCP拥塞控制算法的输出结果包括确定在所述至 少一条TCP子流上传输的数据部分或全部迁移到属于所述MPTCP连 接的其他TCP子流上。
可选地，该网络节点还可以包括发送单元503，用于将调整后的 输入项发送给所述通信两端，所述调整后的输入项由所述策略执行单 元调整所述至少一条TCP子流上的数据传输过程后获得，用于触发所 述通信两端执行所述MPTCP拥塞控制算法。
所述MPTCP拥塞控制算法的输入项的调整策略包括，降低所述 至少一条TCP子流的数据传输速率以及所述数据传输速率降低的幅 度。
可选地，策略执行单元502用于调整所述至少一条TCP子流上的 数据传输过程，包括，延迟发送所述至少一条TCP子流上的数据包以 增大所述RTT，延迟时间与所述数据传输速率降低的幅度对应。
可选地，策略执行单元502还可以用于丢弃所述至少一条TCP子 流上的业务数据包以增大所述丢包率，丢弃所述业务数据包的数量与 所述数据传输速率降低的幅度对应。
在另一个实施例中，也可以由通信两端自行检测RTT及丢包率等 算法输入项，本发明实施例对此不做特别限定。
可选地，发送单元501还可以用于，向所述数据传输控制节点上 报测量报告，所述测量报告用于指示所述多条TCP子流的通信链路的 链路状况和/或所述多条TCP子流对应的网络节点的无线资源状况，所 述测量报告用于所述数据传输控制节点选择所述至少一条TCP子流。
可选地，所述测量报告可以是面向UE接入的网络节点的测量报 告，上报内容可以但不限于以下信息：无线接入类型、无线接入点标 识、无线接入点的资源状态，例如***负载信息等。
可选地，所述测量报告可以是面向接入用户的测量报告，上报内 容可以但不限于以下信息：无线接入类型、无线接入点标识、接入用 户的全球唯一标识、UE的空口链路信息，例如链路质量、用户分布 情况例如是小区中心用户还是边缘用户等。
可选地，可以仅在该网络节点处有与某MPTCP连接相关的TCP 子流时才上报面向该网络节点的测量报告，或者仅在接入用户有与某 MPTCP连接相关的TCP子流时才上报面向该接入用户的测量报告，可 以减少测量报告上报的频度。
可选地，在UE稳定存在某一个网络节点的场景下，测量报告的 上报方式可以是：周期上报或者是事件触发上报，其中，事件触发上 报包括但不限于以下几种：
测量量发生变化时上报；或者，测量量满足一定条件才上报例如 无线接入点的负载超过一定门限才上报；或者，收到数据传输控制节 点发送的测量上报请求消息后按照该请求消息上报测量结果。
可选地，如果UE发生了网络节点切换，测量报告的生成及上报 过程可以包括：
源无线接入点在发现用户成功切换后，上报接入切换通知消息给 数据传输控制节点，携带内容可以包括但不限于以下信息：切换的接 入用户的全球唯一用户标识、无线接入点类型、源无线接入点标识， 目标无线接入点标识。
目标无线接入点在发现接入用户成功切换后，将面向此接入用户 的测量报告按照上述相关实施例中所述的方式上报给数据传输控制 节点。
数据传输控制节点的信息管理单元304收到接入切换通知消息 后，将MPTCP连接信息表中此源无线接入点下此接入用户的所有TCP 子流的无线接入点信息由源无线接入点信息改为目标无线接入点信 息。
可选地，作为本发明的另一个实施例，发送单元501还可以用于， 向所述数据传输控制节点上报MPTCP连接消息，所述MPTCP连接消 息用于指示所述MPTCP连接的状态，所述MPTCP连接消息中携带 MPTCP连接信息，所述MPTCP连接信息用于所述数据传输控制节点 将所述多条TCP子流与所述MPTCP连接相关联，使得所述MPTCP连 接包含所述多条TCP子流。
可选地，发送单元501中还可以包括用户识别子单元5011，用于 识别接入用户的MPTCP连接状态，生成与MPTCP连接状态对应的 MPTCP连接消息。
具体地，用户识别子单元5011可以具体用于，解析接入用户的携 带MPTCP信令的TCP数据包，通过数据包中的特殊字段，识别出哪些 接入用户是MPTCP用户，获得接入用户的MPTCP连接消息。具体地， 可以获得接入用户的MPTCP连接的相关信令，解析该信令并组成相 应的MPTCP连接消息。所述MPTCP连接消息用于指示所述MPTCP连 接的状态。
可选地，用户识别子单元5011解析当前接入用户的携带MPTCP 信令的TCP数据包，如果发现该接入用户完成了MPTCP初始连接建立 过程，则由发送单元501上报MPTCP初始连接建立通知消息给数据传 输控制节点。
网络节点判定MPTCP初始连接建立过程完成的方法可以是：由 网络节点获取并解析携带MP_CAPABLE选项的确认消息(ACK， Acknowledgement)；或者由网络节点获取并解析携带MP_CAPABLE 选项的三步握手过程，即先后依次承载到携带MP_CAPABLE选项的 TCP SYN包、与TCP SYN包对应的携带MP_CAPABLE选项的 SYN/ACK、与SYN/ACK对应的携带MP_CAPABLE选项的ACK。该 判定流程可以由用户识别子单元5011执行。
可选地，用户识别子单元5011解析当前接入用户的携带MPTCP 信令的TCP数据包，如果发现接入用户完成了MPTCP连接的地址增加 通告过程，则发送单元501上报所述地址增加通知消息给数据传输控 制节点。
网络节点判定MPTCP地址增加通告过程完成的方法可以是：获 取接入用户的TCP数据包并解析出其中的ADD_ADDR选项。该判定 流程可以由用户识别子单元5011执行。
可选地，用户识别子单元5011解析当前接入用户的携带MPTCP 信令的TCP数据包，如果发现某用户完成了MPTCP子流新建过程，则 发送单元501上报所述新建TCP子流通知消息给数据传输控制节点。
网络节点判定MPTCP子流新建过程完成的方法可以是依次获取 并解析携带MP_JOIN选项的TCP SYN包、以及与之对应的ACK消息； 也可以是获取并解析携带MP_JOIN选项的4步握手过程，即依次捕获 到携带MP_JOIN选项的TCP SYN包、与TCP SYN包对应的携带 MP_JOIN选项的SYN/ACK、与SYN/ACK对应的携带MP_JOIN选项的 ACK、与携带MP_JOIN选项的ACK对应的ACK。该判定流程可以由 用户识别子单元5011执行。
可选地，用户识别子单元5011解析当前接入用户的TCP数据包， 如果发现某接入用户完成了MPTCP子流删除过程，则发送单元501上 报删除TCP子流通知消息给数据传输控制节点。
网络节点判定MPTCP子流删除过程完成的方法可以是：获取并 解析TCP RST包；也可以获取并解析TCP FIN包；也可以是获取并解 析TCP数据包中REMOVE_ADDR选项。
可选地，用户识别子单元5011解析当前接入用户的携带MPTCP 信令的TCP数据包，如果发现某接入用户完成了MPTCP子流优先级更 新过程，则发送单元501上报TCP子流优先级更新通知消息给数据传 输控制节点。
网络节点判定MPTCP子流优先级更新过程完成的方法可以是： 获取并解析接入用户的TCP数据包中包含的MP_PRIO选项。
需要说明的是，以上各类与MPTCP连接相关的通知消息携带的 具体内容可以参照图3所示实施例中相关描述，在此不再赘述。
采用本发明实施例提供的数据传输控制节点及网络节点，依据系 统运行需求选择需要进行数据分流的TCP子流，通过改变MPTCP拥塞 控制算法的输入项，使选择出的TCP子流上数据能够迁移到预期的属 于同一MPTCP连接的其他TCP子流上，包括满足运营商的管控策略， 反映承载TCP子流的数据传输链路的拥塞状况、以及无线资源状况 等。通过在接入网侧增加对MPTCP数据传输的管控措施，使数据分 流结果能够适应***运行的预期需求，优化***性能，提升MPTCP 拥塞控制算法的实用性。同时不需要改变现有协议的拥塞控制算法， 通过改变算法的输入间接影响算法输出，得到预期的数据分流结果， 流程简单，易于实现。
图6是本发明另一实施例提供的数据传输控制节点的示意框图， 所述数据传输控制节点用于控制MPTCP连接的通信两端之间的数据 传输，所述MPTCP连接包含多条传输控制协议TCP子流，该数据传输 控制节点包括存储器601、处理器602、收发器603。
存储器601存储使得处理器602执行以下操作的指令：
在所述多条TCP子流中选择需要进行数据传输控制的至少一条 TCP子流；确定MPTCP拥塞控制算法的输入项的调整策略；指示所述 至少一条TCP子流对应的网络节点根据所述调整策略，调整所述至少 一条TCP子流上的数据传输过程，触发所述通信两端执行所述MPTCP 拥塞控制算法，所述MPTCP拥塞控制算法的输出结果包括确定在所 述至少一条TCP子流上传输的数据部分或全部迁移到属于所述 MPTCP连接的其他TCP子流上。
收发器603可以用于，将所述调整策略发送给所述至少一条TCP 子流对应的网络节点。
上述调整策略的发送过程包括，遍历空口下行或上行管控子流 集，向管控子流集中的每个TCP子流所在的网络节点下发调整策略。
可选地，可以有多种途径决定选择哪些TCP子流上的空口下行或 上行的数据迁移到同属一个MPTCP连接的其他TCP子流上，本发明实 施例对此不做特别限定。具体内容可以参照图3所示实施例的相关描 述，在此不做赘述。
可选地，存储器601还可以存储使得处理器602执行以下操作的指 令：确定降低所述至少一条TCP子流的数据传输速率以及所述数据传 输速率降低的幅度。
具体地，当某条通信两端正在使用TCP子流的速度降低到一定程 度后，该TCP子流将不满足数据传输的性能要求，通信两端可以自行 检测或由网络节点处获取该TCP子流的速率信息，并启动MPTCP拥塞 控制算法，切换TCP子流。具体内容可以参照图3所示实施例的相关 描述，在此不做赘述。
可选地，存储器601还可以存储使得处理器602执行以下操作的指 令：指示所述网络节点延迟发送所述至少一条TCP子流上的数据包以 增大RTT，延迟时间与所述数据传输速率降低的幅度对应。
可选地，存储器601还可以存储使得处理器602执行以下操作的指 令：调整策略可以包括指示所述网络节点丢弃所述至少一条TCP子流 上的业务数据包以增大丢包率，丢弃所述业务数据包的数量与所述数 据传输速率降低的幅度对应。
上述改变RTT和丢包率的处理过程可以同时进行，即综合运用 RTT和丢包率改变TCP子流的数据传输速率；或者单独进行，本发明 实施例对此不做特别限定。
通过改变TCP层的往返时延和/或丢包率，调整后的RTT和/或丢 包率作为MPTCP拥塞算法的输入项，影响算法的输出结果，使得各 条TCP子流上的数据能够完成满足预期需求的迁移过程。
改变丢包率或RTT的具体方法可以参照图3所示实施例中的相关 描述，在此不做赘述。
可选地，存储器601还可以存储使得处理器602执行以下操作的指 令：根据运营商管控策略、通信链路的链路状况、MPTCP连接下的 各条TCP子流对应的网络节点的无线资源状况中的至少一个条件选 择TCP子流。具体判断方式及过程可以参照图3所示实施例的相关描 述，在此不做赘述。
可选地，收发器603开可以用于，接收网络节点上报的测量报告， 获取各条TCP子流的通信链路的链路状况，和/或各条TCP子流对应的 网络节点的无线资源状况，测量报告的具体内容可以参照图3所示实 施例中的相关描述，在此不做赘述。
可选地，作为本发明的另一个实施例，
存储器601还可以存储使得处理器602执行以下操作的指令：根据 MPTCP连接消息，将属于同一个MPTCP连接的多条TCP子流与该 MPTCP连接相关联，所述MPTCP连接消息用于指示所述MPTCP连接 的状态。
其中，所述MPTCP连接消息可以包括，指示MPTCP初始连接建 立过程完成的MPTCP初始连接建立通知消息；以及指示在MPTCP初 始连接建立过程完成后新建属于该MPTCP连接的TCP子流的新建 TCP子流通知消息。
可选地，还可以包括增加可用地址的地址增加通知消息；还可以 包括，指示删除当前使用的某一条或多条TCP子流的删除TCP子流通 知消息；还可以包括，指示改变TCP子流的使用优先级的TCP子流优 先级更新通知消息等，其中，改变TCP子流使用优先级是指改变某一 条或多条TCP子流的备份状态。
可选地，还可以包括，指示删除当前使用的某条TCP子流的删除 TCP子流通知消息、指示改变TCP子流使用优先级的TCP子流优先级 更新通知消息等，其中，改变TCP子流使用优先级是指更改当前使用 的TCP子流的备份状态。
可选地，数据传输控制节点可以接收多个不同类型的网络节点上 报的属于同一个MPTCP连接的多个不同TCP子流的MPTCP信息。例 如同时接收BTS、NodeB与AP上报的经过该网络节点的至少一条TCP 子流的MPTCP信息。具体的上报方式可以参照图3所示实施例中的相 关内容，在此不再赘述。
存储器601还可以存储使得处理器602执行以下操作的指令：
存储并更新MPTCP连接信息表，所述MPTCP连接信息表中包括 所述网络节点上报的MPTCP连接消息中携带的MPTCP连接信息。
可选地，所述MPTCP连接信息表可以包括不同网络节点上报的 关于同一个MPTCP连接的MPTCP连接消息；也可以包括不同网络节 点上报的关于不同MPTCP连接的MPTCP连接消息，则可以以MPTCP 连接为分类依据，对各条TCP子流进行管理。
可选地，当数据传输控制节点接收到包含不同内容的MPTCP连 接消息时，可以对上述MPTCP连接表进行相应的维护，对应包含不 同内容的MPTCP连接消息的具体维护过程可以参照图3所示实施例 中的相关描述，在此不做赘述。
通过查找上述MPTCP连接表，匹配表中的信息，可以将各条TCP 子流与MPTCP连接相关联，具体的关联过程可以参照图3所示实施例 中的相关描述，在此不做赘述。
此外，该数据传输控制节点还可以包括天线604、以及总线*** 605等。处理器602控制该装置的操作，处理器602还可以称为CPU (Central Processing Unit，中央处理单元)。存储器601可以包括只读 存储器和随机存取存储器，并向处理器602提供指令和数据。存储器 601的一部分还可以包括非易失性随机存取存储器(NVRAM)。具体 的应用中，收发器603可以耦合到天线605。该装置的各个组件通过总 线***605耦合在一起，其中总线***605除包括数据总线之外，还可 以包括电源总线、控制总线和状态信号总线等。但是为了清楚说明起 见，在图中将各种总线都标为总线***605。
图7是本发明实施例提供的一种网络节点的示意框图，包括存储 器701、处理器702、收发器703。
收发器703，用于接收数据传输控制节点在所述多条TCP子流中 选择得到的至少一条TCP子流的信息。以及，用于接收所述数据传输 控制节点指示的调整策略，所述调整策略为针对MPTCP拥塞控制算 法的输入项的调整策略。
可选地，所述MPTCP拥塞控制算法的输入项包括，所述至少一 条TCP子流的数据往返时延RTT和/或丢包率。
存储器701还可以存储使得处理器702执行以下操作的指令：
根据数据传输控制节点发送的上述调整策略，调整所述至少一条 TCP子流上的数据传输过程，触发所述通信两端执行所述MPTCP拥塞 控制算法，所述MPTCP拥塞控制算法的输出结果包括确定在所述至 少一条TCP子流上传输的数据部分或全部迁移到属于所述MPTCP连 接的其他TCP子流上。
可选地，存储器701还可以存储使得处理器702执行以下操作的指 令：降低所述至少一条TCP子流的数据传输速率以及所述数据传输速 率降低的幅度。
可选地，存储器701还可以存储使得处理器702执行以下操作的指 令：丢弃所述至少一条TCP子流上的业务数据包以增大所述丢包率， 丢弃所述业务数据包的数量与所述数据传输速率降低的幅度对应。
通过改变RTT和/或丢包率，可以降低特定TCP子流的数据传输速 率，触发通信两端改变数据传输路径，通过其他TCP子流传输数据， 保证数据传输质量。
可选地，收发器703还可以用于，将调整后的输入项发送给所述 通信两端，所述调整后的输入项由处理器702调整所述至少一条TCP 子流上的数据传输过程后获得，用于触发所述通信两端执行所述 MPTCP拥塞控制算法。
在另一个实施例中，也可以由通信两端自行检测RTT及丢包率等 算法输入项，本发明实施例对此不做特别限定。
存储器701还可以存储使得处理器702执行以下操作的指令：
解析所述接入用户的TCP数据包，获得所述接入用户的MPTCP 连接状态，所述MPTCP连接消息用于指示所述MPTCP连接的状态。
其中，MPTCP连接消息包含的具体内容以及MPTCP连接状态的 具体判断方法可以参照图6所示实施例中的相关描述，在此不再赘述。
存储器701还可以存储使得处理器702执行以下操作的指令：
向所述数据传输控制节点上报测量报告，所述测量报告用于指示 所述多条TCP子流的通信链路的链路状况和/或所述多条TCP子流对 应的网络节点的无线资源状况，所述测量报告用于所述数据传输控制 节点选择所述至少一条TCP子流。
具体的测量报告内容以及上报方式可以参照图5所示实施例中的 相关描述，在此不做赘述。
可选地，作为本发明的另一实施例，收发器703还可以用于，向 所述数据传输控制节点上报MPTCP连接消息，所述MPTCP连接消息 用于指示所述MPTCP连接的状态，所述MPTCP连接消息中携带 MPTCP连接信息，所述MPTCP连接信息用于所述数据传输控制节点 将所述多条TCP子流与所述MPTCP连接相关联，使得所述MPTCP连 接包含所述多条TCP子流。
可选地，处理器702还可以用于识别接入用户的MPTCP连接状 态，生成与MPTCP连接状态对应的MPTCP连接消息。
具体的识别方式、MPTCP连接消息的类型、内容及上报方式可 以参照本发明其他实施例的相关描述，在此不做赘述。
网络节点还可以包括天线704、以及总线***705等，具体的连接 关系可以参照图6所示实施例中相关描述，在此不再赘述。
采用本发明实施例提供的数据传输控制节点及网络节点，依据系 统运行需求选择需要进行数据分流的TCP子流，通过改变MPTCP拥塞 控制算法的输入项，使选择出的TCP子流上数据能够迁移到预期的属 于同一MPTCP连接的其他TCP子流上，包括满足运营商的管控策略， 反映承载TCP子流的数据传输链路的拥塞状况、以及无线资源状况 等。通过在接入网侧增加对MPTCP数据传输的管控措施，使数据分 流结果能够适应***运行的预期需求，优化***性能，提升MPTCP 拥塞控制算法的实用性。同时不需要改变现有协议的拥塞控制算法， 通过改变算法的输入间接影响算法输出，得到预期的数据分流结果， 流程简单，易于实现。
图8是本发明实施例提供的一种通信***示意图，所述通信*** 包括，网络节点801，数据传输控制节点802，两者可以通过有线或无 线链路进行直接或间接通信。
数据传输控制节点801，用于在所述多条TCP子流中选择需要进 行数据传输控制的至少一条TCP子流；确定MPTCP拥塞控制算法的输 入项的调整策略；将所述调整策略发送给所述至少一条TCP子流对应 的网络节点，所述调整策略用于指示所述网络节点调整所述至少一条 TCP子流上的数据传输过程，触发所述通信两端执行所述MPTCP拥塞 控制算法，所述MPTCP拥塞控制算法的输出结果包括确定在所述至 少一条TCP子流上传输的数据部分或全部迁移到属于所述MPTCP连 接的其他TCP子流上；
网络节点802，位于所述通信两端的通信链路中，用于接收数据 传输控制节点在所述多条TCP子流中选择得到的至少一条TCP子流 的信息；接收所述数据传输控制节点指示的调整策略，所述调整策略 为针对MPTCP拥塞控制算法的输入项的调整策略。
该通信***中的数据传输控制节点801可以是图3或图6所示实施 例中的数据传输控制节点，网络节点802可以是图5或图7所示实施例 中的网络节点，在该通信***中，可以执行图9-图11所示实施例中的 数据传输控制方法及TCP子流关联方法，详细内容可以参照本发明其 他实施例的相关描述，在此不做赘述。
采用本发明实施例提供的通信***，依据***运行需求选择需要 进行数据分流的TCP子流，通过改变MPTCP拥塞控制算法的输入项， 使选择出的TCP子流上数据能够迁移到预期的属于同一MPTCP连接 的其他TCP子流上。通过在接入网侧增加对MPTCP数据传输的管控措 施，使TCP子流上数据分流结果能够适应***运行的预期需求，优化 ***性能，提升MPTCP拥塞控制算法的实用性。
本发明实施例提供了一种数据传输管理方法，可以由图3或图6 所示实施例中的数据传输控制节点执行，用于控制多路径传输控制协 议MPTCP连接的通信两端之间的数据传输，所述MPTCP连接包含多 条传输控制协议TCP子流。图9是该方法的流程示意图，包括步骤 S901-S903：
S901，数据传输控制节点在所述多条TCP子流中选择需要进行数 据传输控制的至少一条TCP子流。
具体地，数据传输控制节点可以决定选择哪些TCP子流上的空口 下行或上行的数据迁移到同属一个MPTCP连接的其他TCP子流上，并 将需要进行数据迁移的TCP子流按数据传输方向归入空口下行或上 行管控子流集。
S902，数据传输控制节点确定MPTCP拥塞控制算法的输入项的 调整策略。
可选地，S902中所述数据传输控制节点确定MPTCP拥塞控制算 法的输入项的调整策略包括，所述数据传输控制节点确定降低所述至 少一条TCP子流的数据传输速率以及所述数据传输速率降低的幅度。
可选地，可以改变作为MPTCP拥塞控制算法输入项的TCP层的往 返时延(RTT，Round Trip Time)和/或丢包率，调整后的RTT和/或 丢包率作为算法的输入，影响算法的输出结果，使得各条TCP子流上 的数据能够完成满足预期需求的迁移过程。
可选地，调整策略可以包括指示所述网络节点延迟发送所述至少 一条TCP子流上的数据包以增大RTT，延迟时间与所述数据传输速率 降低的幅度对应。
可选地，调整策略可以包括指示所述网络节点丢弃所述至少一条 TCP子流上的业务数据包以增大丢包率，丢弃所述业务数据包的数量 与所述数据传输速率降低的幅度对应。
上述改变RTT和丢包率的处理过程可以同时进行，即综合运用 RTT和丢包率改变TCP子流的数据传输速率；或者单独进行，本发明 实施例对此不做特别限定。
改变丢包率或RTT的具体方法可以参照图3所示实施例的相关内 容，在此不做赘述。
S903，数据传输控制节点指示所述至少一条TCP子流对应的网络 节点根据所述调整策略，调整所述至少一条TCP子流上的数据传输过 程，触发所述通信两端执行所述MPTCP拥塞控制算法，所述MPTCP 拥塞控制算法的输出结果包括确定在所述至少一条TCP子流上传输 的数据部分或全部迁移到属于所述MPTCP连接的其他TCP子流上。
可选地，上述调整策略的下发过程包括，遍历空口下行或上行管 控子流集，向管控子流集中的每个TCP子流所在的网络节点下发调整 策略。可以有多种途径决定选择哪些TCP子流上的空口下行或上行的 数据迁移到同属一个MPTCP连接的其他TCP子流上，本发明实施例对 此不做特别限定。具体的数据迁移方法可以参照图3所示实施例的相 关描述，在此不做赘述。
作为本发明的另一个实施例，S901中，数据传输控制节点在所述 多条TCP子流中选择需要进行数据传输控制的至少一条TCP子流包 括，根据运营商管控策略、通信链路的链路状况、MPTCP连接下的 各条TCP子流对应的网络节点的无线资源状况中的至少一个条件选 择TCP子流进行分流。其中，通信链路的链路状况和/或无线资源状况 可以由网络节点上报的测量报告中的相关信息获取。具体内容可以参 照图3所示实施例的相关描述，在此不做赘述。
可选地，作为本发明的另一个实施例，所述数据传输控制节点根 据MPTCP连接消息，将所述多条TCP子流与所述MPTCP连接相关联， 使得所述MPTCP连接包含所述多条TCP子流，所述MPTCP连接消息 用于指示所述MPTCP连接的状态。
其中，所述MPTCP连接消息可以包括，指示MPTCP初始连接建 立过程完成的MPTCP初始连接建立通知消息；以及指示在MPTCP初 始连接建立过程完成后新建属于该MPTCP连接的TCP子流的新建 TCP子流通知消息。
可选地，还可以包括指示在MPTCP初始连接建立过程完成后， 为新建TCP子流的多址响应端增加可用地址的地址增加通知消息。
可选地，还可以包括，指示删除当前使用的某条TCP子流的删除 TCP子流通知消息、指示改变TCP子流使用优先级的TCP子流优先级 更新通知消息等，其中，改变TCP子流使用优先级是指更改当前使用 的TCP子流的备份(Backup)状态。
具体地，为数据传输控制节点管理的每条TCP子流关联对应的 MPTCP连接的过程如图10所示，包括步骤S1001-1004：
S1001，解析网络节点上报的MPTCP初始连接建立通知消息，获 得当前MPTCP连接的通信两端的地址及Token。
S1002，解析网络节点上报的新建TCP子流通知消息，判断所述 TCP子流新建过程的发起端为所述MPTCP连接的通信两端中的哪一 端，获得当前TCP子流的地址以及子流新建过程的两端Token。
S1003，根据TCP子流新建过程的发起端的Token及当前TCP子流 的地址，确定TCP子流新建过程的响应端是上述MPTCP连接的通信两 端中的另一端，并获得响应端地址。
S1004，搜索MPTCP连接信息表中与该MPTCP连接对应的信息， 确定新建TCP子流所属的MPTCP连接。
具体地，如果搜索到所述TCP子流新建过程的响应端Token与某 MPTCP连接的一端的Token相等，且所述TCP子流新建过程的响应端 地址与该MPTCP连接的一端的可用地址集中的一个地址相等，则所 述新建TCP子流属于所述MPTCP连接。
对数据传输控制节点管理的每条TCP子流都可以进行步骤 S1001-S1004的判断过程，从而将属于同一MPTCP连接的多条TCP子 流关联到该MPTCP连接上。
采用本发明实施例提供的数据传输控制方法，依据***运行需求 选择需要进行数据分流的TCP子流，通过改变MPTCP拥塞控制算法的 输入项，使选择出的TCP子流上数据能够迁移到预期的属于同一 MPTCP连接的其他TCP子流上。通过在接入网侧增加对MPTCP数据 传输的管控措施，使TCP子流上数据分流结果能够适应***运行的预 期需求，优化***性能，提升MPTCP拥塞控制算法的实用性。
本发明实施例还提供了一种数据传输管理方法，用于控制多路径 传输控制协议MPTCP连接的通信两端之间的数据传输，所述MPTCP 连接包含多条传输控制协议TCP子流。图11是该方法的流程示意图， 如图所示，通信两端可以是用户设备及服务器，包括步骤 S1101-S1107：
S1101，网络节点识别接入用户是否属于MPTCP连接的发起端或 接收端。
其中，解析所述接入用户的TCP数据包，获得所述接入用户的 MPTCP连接状态，所述MPTCP连接消息用于指示所述MPTCP连接的 状态，通过数据包中的特殊字段，可以识别接入用户是否是MPTCP 用户。接入用户是指UE通过接入网接入通信网络后对应的逻辑单元。
其中，MPTCP连接消息包含的具体内容以及MPTCP连接状态的 具体判断方法可以参照对应装置实施例中的相关描述，在此不再赘 述。
S1102，网路节点向所述数据传输控制节点上报MPTCP连接消 息，所述MPTCP连接消息用于指示所述MPTCP连接的状态，所述 MPTCP连接消息中携带MPTCP连接信息，所述MPTCP连接信息用于 所述数据传输控制节点将所述多条TCP子流与所述MPTCP连接相关 联，使得所述MPTCP连接包含所述多条TCP子流。
可选地，可以接收多个不同类型的网络节点上报的属于同一个 MPTCP连接的多条不同TCP子流的MPTCP信息。例如同时接收BTS、 NodeB与AP上报的经过该网络节点的至少一条TCP子流的MPTCP信 息。具体的上报方式可以参照装置实施例中的相关内容，在此不再赘 述。
可选地，在本发明的另一个实施例中，还可以包括步骤S1102’： 网络节点向所述数据传输管理节点上报测量报告，所述测量报告用于 指示通信链路的链路状况和/或无线资源状况。
可选地，可以仅在该网络节点处有与某MPTCP连接相关的TCP 子流时才上报面向网络节点的测量报告，或者仅在接入用户有与某 MPTCP连接相关的TCP子流时上报面向接入用户的测量报告。
具体的测量报告内容以及上报方式可以参照对应装置实施例中 的相关描述，在此不做赘述。
步骤S1102与S1102’没有执行顺序的限制，也可以先执行S1102’ 再执行S1102。
S1103，数据传输控制节点存储MPTCP连接信息表，并根据获取 的MPTCP连接消息更新，所述MPTCP连接信息表中包括所述网络节 点上报的MPTCP连接消息。
可选地，所述MPTCP连接信息表可以包括不同网络节点上报的 关于同一个MPTCP连接的MPTCP连接消息；也可以包括不同网络节 点上报的关于不同MPTCP连接的MPTCP连接消息。以MPTCP连接为 分类依据，对各条TCP子流进行管理。
可选地，当接收到包含不同内容的MPTCP连接消息时，可以对 上述MPTCP连接表进行相应的维护，对应各类MPTCP连接消息的具 体处理过程可以参照装置实施例中的相关描述，在此不做赘述。
可选地，当接收到网络节点上报的关于通信链路的链路状况和/ 网络节点的无线资源状况的测量报告时，可以更新存储的MPTCP连 接信息表中相关的网络节点信息和/或接入用户信息。
S1104，数据传输控制节点根据接收到的MPTCP连接消息将属于 所述MPTCP连接的多条TCP子流与所述MPTCP连接相关联。
将TCP子流与对应的MPTCP连接关联的具体方法与流程可以参 照本发明其他实施例中的相关描述，在此不再赘述。
可选地，所述MPTCP连接消息还包括删除TCP子流通知消息和/ 或TCP子流优先级更新通知消息。对应的删除TCP子流过程及TCP子 流优先级更新过程可以参照装置实施例中的相关描述，在此不再赘 述。
可选地，在MPTCP连接的通信两端为一方多址、一方单址，且 由单址方发起TCP子流新建过程的场景，或者两端均为多址的场景 中，在MPTCP连接的一端发起新建TCP子流过程之前，多址一端可以 先发起地址增加通过过程，告知对端在新建TCP子流时自身需要增加 的可用地址。如果网络节点解析出接入用户完成了上述地址增加通告 过程，则上报地址增加通知消息给数据传输控制节点，包括多址的新 建TCP子流响应端需要增加的可用地址。对应地，在步骤S1004之前 还可以包括，步骤S1104’：在MPTCP连接信息表中增加新建TCP子流 的响应端的可用地址，该响应端为MPTCP初始连接的多址发起端或 多址响应端。
对TCP子流的拥塞控制的前提是将属于同一个MPTCP连接的 TCP子流与MPTCP连接相关联，即需要明确当前UE进行数据通信可 使用的TCP子流是属于哪个MPTCP连接以及从哪个网络节点接入等 信息，并确认TCP子流的状态，才能对属于同一个MPTCP连接的不同 TCP子流进行拥塞控制，在不同TCP子流之间进行数据分流。
S1105，数据传输控制节点在所述多条TCP子流中选择需要进行 数据传输控制的至少一条TCP子流。
S1106，数据传输控制节点将确定的调整策略下发至网络节点。
S1107，网络节点执行上述调整策略，调整所述至少一条TCP子 流上的数据传输过程。
可选地，网络节点可以延迟发送所述至少一条TCP子流上的数据 包以增大所述RTT，延迟时间与所述数据传输速率降低的幅度对应； 和/或，丢弃所述至少一条TCP子流上的业务数据包以增大所述丢包 率，丢弃所述业务数据包的数量与所述数据传输速率降低的幅度对 应。
S1108，网络节点将调整TCP子流的数据传输过程后得到的输入 项发送给通信两端，触发通信两端执行MPTCP拥塞控制算法。
其中，调整后的输入项由所述网络节点调整所述至少一条TCP子 流上的数据传输过程后获得。
上述通信两端执行的MPTCP拥塞控制算法的输出结果包括确定 在所述至少一条TCP子流上传输的数据部分或全部迁移到属于所述 MPTCP连接的其他TCP子流上，其中，所述调整策略为针对MPTCP 拥塞控制算法的输入项的调整策略。
可选地，可以由数据传输控制节点确定上述调整策略的具体实施 形式，下发给对应的网络节点直接执行；也可以由数据传输控制节点 仅给出TCP子流迁移策略下发给对应的网络节点，由网络节点自行决 定采取何种具体形式来执行数据迁移。
该实施例中各步骤的具体描述可以参照本发明其他实施例的相 关内容，在此不做赘述。
采用本发明实施例提供的数据传输控制方法，依据***运行需求 选择需要进行数据分流的TCP子流，通过改变MPTCP拥塞控制算法的 输入项，使选择出的TCP子流上数据能够迁移到预期的属于同一 MPTCP连接的其他TCP子流上，包括满足运营商的管控策略，反映承 载TCP子流的数据传输链路的拥塞状况、以及无线资源状况等。通过 在接入网侧增加对MPTCP数据传输的管控措施，使数据分流结果能 够适应***运行的预期需求，优化***性能，提升MPTCP拥塞控制 算法的实用性。同时不需要改变现有协议的拥塞控制算法，通过改变 算法的输入间接影响算法输出，得到预期的数据分流结果，流程简单， 易于实现。
本所属领域的技术人员可以清楚地了解到，为描述的方便和简 洁，上述描述的***、装置及方法的具体工作过程，可以互相参考， 在此不再赘述。
通过以上的实施例的描述，所属领域的技术人员可以清楚地了解 到本发明可借助软件加必需的通用硬件平台的方式来实现，当然也可 以通过硬件，但很多情况下前者是更佳的实施方式。基于这样的理解， 本发明的技术方案本质上或者说对现有技术做出贡献的部分可以以 软件产品的形式体现出来，该计算机软件产品存储在一个存储介质 中，包括若干指令用以使得一台计算机设备(可以是个人计算机，服 务器，或者网络设备等)执行本发明各个实施例所述方法的全部或部 分步骤。而前述的存储介质包括：U盘、移动硬盘、只读存储器(ROM)、 随机存取存储器(RAM)、磁碟或者光盘等各种可以存储程序代码的 介质。
在本申请所提供的几个实施例中，应该理解到，所揭露的装置、 ***和方法，在没有超过本申请的范围内，可以通过其他的方式实现。 例如，以上所描述的装置实施例仅仅是示意性的，例如，所述模块或 单元的划分，仅仅为一种逻辑功能划分，实际实现时可以有另外的划 分方式，例如多个单元或组件可以结合或者可以集成到另一个***， 或一些特征可以忽略，或不执行。其中所述作为分离部件说明的单元 可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或 者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多 个网络单元上。可以根据实际的需要选择其中的部分或者全部模块来 实现本实施例方案的目的。本领域普通技术人员在不付出创造性劳动 的情况下，即可以理解并实施。
另外，所描述装置、***和方法以及不同实施例的示意图，在不 超出本申请的范围内，可以与其它***，模块，技术或方法结合或集 成。另一点，所显示或讨论的相互之间的耦合或直接耦合或通信连接 可以是通过一些接口，基站或单元的间接耦合或通信连接，可以是电 子、机械或其它的形式。
以上所述，仅为本发明的具体实施方式，但本发明的保护范围并 不局限于此，任何熟悉本技术领域的技术人员在本发明揭露的技术范 围内，可轻易想到变化或替换，都应涵盖在本发明的保护范围之内。 因此，本发明的保护范围应所述以权利要求的保护范围为准。

Claims (39)

  1. 一种数据传输控制节点，其特征在于，用于控制多路径传输 控制协议MPTCP连接的通信两端之间的数据传输，所述MPTCP连接 包含多条传输控制协议TCP子流，所述数据传输控制节点包括，
    处理单元，用于在所述多条TCP子流中选择需要进行数据传输控 制的至少一条TCP子流；
    所述处理单元还用于，确定MPTCP拥塞控制算法的输入项的调 整策略；
    发送单元，用于将所述调整策略发送给所述至少一条TCP子流对 应的网络节点，所述调整策略用于指示所述网络节点调整所述至少一 条TCP子流上的数据传输过程，触发所述通信两端执行所述MPTCP 拥塞控制算法，所述MPTCP拥塞控制算法的输出结果包括确定在所 述至少一条TCP子流上传输的数据部分或全部迁移到属于所述 MPTCP连接的其他TCP子流上，
    其中，所述网络节点位于所述通信两端的通信链路中。
  2. 根据权利要求1所述的数据传输控制节点，其特征在于，所述 处理单元用于确定MPTCP拥塞控制算法的输入项的调整策略，包括， 确定降低所述至少一条TCP子流的数据传输速率以及所述数据传输 速率降低的幅度。
  3. 根据权利要求1或2所述的数据传输控制节点，其特征在于， 所述MPTCP拥塞控制算法的输入项包括，所述至少一条TCP子流的数 据往返时延RTT和/或丢包率。
  4. 根据权利要求1-3所述的数据传输控制节点，其特征在于，所 述发送单元用于指示所述网络节点调整所述至少一条TCP子流上的 数据传输过程以获得调整后的输入项，包括，
    指示所述网络节点延迟发送所述至少一条TCP子流上的数据包 以增大所述RTT，延迟时间与所述数据传输速率降低的幅度对应；和 /或，
    指示所述网络节点丢弃所述至少一条TCP子流上的业务数据包 以增大所述丢包率，丢弃所述业务数据包的数量与所述数据传输速率 降低的幅度对应。
  5. 根据权利要求1-4所述的数据传输控制节点，其特征在于，所 述处理单元用于在所述多条TCP子流中选择需要进行数据传输控制 的至少一条TCP子流，包括，根据运营商管控策略选择所述至少一条 TCP子流。
  6. 根据权利要求1-5任一所述的数据传输控制节点，其特征在于， 所述处理单元用于在所述多条TCP子流中选择需要进行数据传输控 制的至少一条TCP子流，包括，
    根据所述多条TCP子流对应的网络节点的无线资源状况确定需 要进行数据传输控制的网络节点，选择所述网络节点对应的正在进行 数据传输的至少一条TCP子流。
  7. 根据权利要求1-6任一所述的数据传输控制节点，其特征在于， 所述处理单元用于在所述多条TCP子流中选择需要进行数据传输控 制的至少一条TCP子流，包括，判断所述多条TCP子流所在的通信链 路的链路状况是否满足预设条件，在不满足所述预设条件的TCP子流 中，选择正在进行数据传输的至少一条TCP子流。
  8. 根据权利要求6或7所述的数据传输控制节点，其特征在于， 还包括接收单元，用于接收所述网络节点上报的测量报告，所述测量 报告用于指示所述多条TCP子流的通信链路的链路状况和/或所述多 条TCP子流对应的网络节点的无线资源状况；
    所述处理单元具体用于根据所述测量报告，选择所述至少一条 TCP子流。
  9. 根据权利要求1-8任一所述的数据传输控制节点，其特征在于， 还包括信息管理单元，用于根据MPTCP连接消息，将所述多条TCP 子流与所述MPTCP连接相关联，使得所述MPTCP连接包含所述多条 TCP子流，所述MPTCP连接消息用于指示所述MPTCP连接的状态。
  10. 根据权利要求9所述的数据传输控制节点，其特征在于，所 述信息管理单元还用于，存储并更新MPTCP连接信息表，所述MPTCP 连接信息表中包括所述网络节点上报的MPTCP连接消息中携带的 MPTCP连接信息，所述MPTCP连接信息用于将所述多条TCP子流与 所述MPTCP连接相关联。
  11. 根据权利要求1-10任一所述的数据传输控制节点，其特征在 于，所述通信两端中的一端是接入所述网络节点的用户设备UE，另 一端是服务器；
    所述网络节点支持所述UE与所述服务器相通信。
  12. 一种网络节点，其特征在于，用于控制多路径传输控制协议 MPTCP连接的通信两端之间的数据传输，所述MPTCP连接包含多条 传输控制协议TCP子流，所述网络节点位于所述通信两端的通信链路 中，包括，
    接收单元，用于接收数据传输控制节点在所述多条TCP子流中选 择得到的至少一条TCP子流的信息；
    所述接收单元还用于，接收所述数据传输控制节点指示的调整策 略，所述调整策略为针对MPTCP拥塞控制算法的输入项的调整策略。
    策略执行单元，用于根据所述调整策略，调整所述至少一条TCP 子流上的数据传输过程，触发所述通信两端执行所述MPTCP拥塞控 制算法，所述MPTCP拥塞控制算法的输出结果包括确定在所述至少 一条TCP子流上传输的数据部分或全部迁移到属于所述MPTCP连接 的其他TCP子流上。
  13. 根据权利要求12所述的网络节点，其特征在于，所述网络节 点还包括发送单元，用于将调整后的输入项发送给所述通信两端，所 述调整后的输入项由所述策略执行单元调整所述至少一条TCP子流 上的数据传输过程后获得，用于触发所述通信两端执行所述MPTCP 拥塞控制算法。
  14. 根据权利要求12或13所述的网络节点，其特征在于，所述 MPTCP拥塞控制算法的输入项的调整策略包括，降低所述至少一条 TCP子流的数据传输速率以及所述数据传输速率降低的幅度。
  15. 根据权利要求14所述的网络节点，其特征在于，所述MPTCP 拥塞控制算法的输入项包括，所述至少一条TCP子流的数据往返时延 RTT和/或丢包率。
  16. 根据权利要求15所述的网络节点，其特征在于，所述策略执 行单元用于调整所述至少一条TCP子流上的数据传输过程，包括，延 迟发送所述至少一条TCP子流上的数据包以增大所述RTT，延迟时间 与所述数据传输速率降低的幅度对应；和/或，
    丢弃所述至少一条TCP子流上的业务数据包以增大所述丢包率， 丢弃所述业务数据包的数量与所述数据传输速率降低的幅度对应。
  17. 根据权利要求12-16任一所述的网络节点，其特征在于，所 述发送单元还用于，向所述数据传输控制节点上报测量报告，
    所述测量报告用于指示所述多条TCP子流的通信链路的链路状 况和/或所述多条TCP子流对应的网络节点的无线资源状况，所述测量 报告用于所述数据传输控制节点选择所述至少一条TCP子流。
  18. 根据权利要求12-17任一所述的网络节点，其特征在于，所 述发送单元还用于，向所述数据传输控制节点上报MPTCP连接消息， 所述MPTCP连接消息用于指示所述MPTCP连接的状态，
    所述MPTCP连接消息中携带MPTCP连接信息，所述MPTCP连接 信息用于所述数据传输控制节点将所述多条TCP子流与所述MPTCP 连接相关联，使得所述MPTCP连接包含所述多条TCP子流。
  19. 根据权利要求12-18任一所述的网络节点，其特征在于，所 述通信两端中的一端是接入所述网络节点的用户设备UE，另一端是 服务器；
    所述网络节点支持所述UE与所述服务器相通信。
  20. 一种网络***，其特征在于，包括，
    数据传输控制节点，用于在所述多条TCP子流中选择需要进行数 据传输控制的至少一条TCP子流；确定MPTCP拥塞控制算法的输入项 的调整策略；将所述调整策略发送给所述至少一条TCP子流对应的网 络节点，所述调整策略用于指示所述网络节点调整所述至少一条TCP 子流上的数据传输过程，触发所述通信两端执行所述MPTCP拥塞控 制算法，所述MPTCP拥塞控制算法的输出结果包括确定在所述至少 一条TCP子流上传输的数据部分或全部迁移到属于所述MPTCP连接 的其他TCP子流上；
    网络节点，位于所述通信两端的通信链路中，用于接收数据传输 控制节点在所述多条TCP子流中选择得到的至少一条TCP子流的信 息；接收所述数据传输控制节点指示的调整策略，所述调整策略为针 对MPTCP拥塞控制算法的输入项的调整策略。
  21. 一种数据传输管理方法，其特征在于，用于控制多路径传输 控制协议MPTCP连接的通信两端之间的数据传输，所述MPTCP连接 包含多条传输控制协议TCP子流，所述方法包括，
    数据传输控制节点在所述多条TCP子流中选择需要进行数据传 输控制的至少一条TCP子流；
    所述数据传输控制节点确定MPTCP拥塞控制算法的输入项的调 整策略；
    所述数据传输控制节点指示所述至少一条TCP子流对应的网络 节点根据所述调整策略，调整所述至少一条TCP子流上的数据传输过 程，触发所述通信两端执行所述MPTCP拥塞控制算法，所述MPTCP 拥塞控制算法的输出结果包括确定在所述至少一条TCP子流上传输 的数据部分或全部迁移到属于所述MPTCP连接的其他TCP子流上。
  22. 根据权利要求21所述的方法，其特征在于，所述数据传输控 制节点确定MPTCP拥塞控制算法的输入项的调整策略包括，
    所述数据传输控制节点确定降低所述至少一条TCP子流的数据 传输速率以及所述数据传输速率降低的幅度。
  23. 根据权利要求22所述的方法，其特征在于，所述MPTCP拥 塞控制算法的输入项包括，所述至少一条TCP子流的数据往返时延 RTT和/或丢包率。
  24. 根据权利要求23所述的方法，其特征在于，所述数据传输控 制节点指示所述网络节点调整所述至少一条TCP子流上的数据传输 过程包括，
    所述数据传输控制节点指示所述网络节点延迟发送所述至少一 条TCP子流上的数据包以增大所述RTT，延迟时间与所述数据传输速 率降低的幅度对应；和/或，
    所述数据传输控制节点指示所述网络节点丢弃所述至少一条 TCP子流上的业务数据包以增大所述丢包率，丢弃所述业务数据包的 数量与所述数据传输速率降低的幅度对应。
  25. 根据权利要求21-24任一所述的方法，其特征在于，所述数 据传输控制节点在所述多条TCP子流中选择需要进行数据传输控制 的至少一条TCP子流，包括，
    所述数据传输控制节点根据运营商管控策略选择所述至少一条 TCP子流。
  26. 根据权利要求21-25任一所述的方法，其特征在于，所述数 据传输控制节点在所述多条TCP子流中选择需要进行数据传输控制 的至少一条TCP子流，包括，
    所述数据传输控制节点根据所述多条TCP子流对应的网络节点 的无线资源状况确定需要进行数据传输控制的网络节点，选择所述网 络节点对应的正在进行数据传输的至少一条TCP子流。
  27. 根据权利要求21-26任一所述的方法，其特征在于，所述数 据传输控制节点在所述多条TCP子流中选择需要进行数据传输控制 的至少一条TCP子流包括，
    所述数据传输控制节点判断所述多条TCP子流所在的通信链路 的链路状况是否满足预设条件，在不满足所述预设条件的TCP子流 中，选择正在进行数据传输的至少一条TCP子流。
  28. 根据权利要求26或27所述的方法，其特征在于，所述数据传 输控制节点在所述多条TCP子流中选择需要进行数据传输控制的至 少一条TCP子流，包括，
    所述数据传输控制节点接收所述网络节点上报的测量报告，所述 测量报告用于指示所述多条TCP子流的通信链路的链路状况和/或所 述多条TCP子流对应的网络节点的无线资源状况；
    所述数据传输控制节点根据所述测量报告，选择所述至少一条 TCP子流。
  29. 根据权利要求21-28任一所述的方法，其特征在于，所述方 法进一步包括，
    所述数据传输控制节点根据MPTCP连接消息，将所述多条TCP 子流与所述MPTCP连接相关联，使得所述MPTCP连接包含所述多条 TCP子流，所述MPTCP连接消息用于指示所述MPTCP连接的状态。
  30. 根据权利要求29所述的方法，其特征在于，还包括，所述数 据传输控制节点存储并更新MPTCP连接信息表，所述MPTCP连接信 息表中包括所述网络节点上报的MPTCP连接消息中携带的MPTCP连 接信息，所述MPTCP连接信息用于将所述多条TCP子流与所述 MPTCP连接相关联。
  31. 根据权利要求21-30任一所述的方法，其特征在于，
    所述通信两端中的一端是接入所述网络节点的用户设备UE，另 一端是服务器；
    所述网络节点支持所述UE与所述服务器相通信。
  32. 一种数据传输管理方法，其特征在于，用于控制多路径传输 控制协议MPTCP连接的通信两端之间的数据传输，所述MPTCP连接 包含多条传输控制协议TCP子流，网络节点位于所述通信两端的通信 链路中，所述方法包括，
    所述网络节点接收数据传输控制节点在所述多条TCP子流中选 择得到的至少一条TCP子流的信息，
    所述网络节点根据数据传输控制节点指示的调整策略，调整所述 至少一条TCP子流上的数据传输过程，触发所述通信两端执行所述 MPTCP拥塞控制算法，所述MPTCP拥塞控制算法的输出结果包括确 定在所述至少一条TCP子流上传输的数据部分或全部迁移到属于所 述MPTCP连接的其他TCP子流上，其中，所述调整策略为针对MPTCP 拥塞控制算法的输入项的调整策略。
  33. 根据权利要求32所述的方法，其特征在于，在所述网络节点 调整所述至少一条TCP子流上的数据传输过程之后还包括，
    所述网络节点将调整后的输入项发送给所述通信两端，所述调整 后的输入项由所述网络节点调整所述至少一条TCP子流上的数据传 输过程后获得，用于触发所述通信两端执行所述MPTCP拥塞控制算 法。
  34. 根据权利要求32或33所述的方法，其特征在于，所述MPTCP 拥塞控制算法的输入项的调整策略包括，降低所述至少一条TCP子流 的数据传输速率以及所述数据传输速率降低的幅度。
  35. 根据权利要求34所述的方法，其特征在于，所述MPTCP拥 塞控制算法的输入项包括，所述至少一条TCP子流的数据往返时延 RTT和/或丢包率。
  36. 根据权利要求35所述的方法，其特征在于，所述网络节点调 整所述至少一条TCP子流上的数据传输过程包括，
    所述网络节点延迟发送所述至少一条TCP子流上的数据包以增 大所述RTT，延迟时间与所述数据传输速率降低的幅度对应；和/或，
    所述网络节点丢弃所述至少一条TCP子流上的业务数据包以增 大所述丢包率，丢弃所述业务数据包的数量与所述数据传输速率降低 的幅度对应。
  37. 根据权利要求32-36任一所述的方法，其特征在于，所述方 法还包括，所述网络节点向所述数据传输控制节点上报测量报告，
    所述测量报告用于指示所述多条TCP子流的通信链路的链路状 况和/或所述多条TCP子流对应的网络节点的无线资源状况，所述测量 报告用于所述数据传输控制节点选择所述至少一条TCP子流。
  38. 根据权利要求32-37任一所述的方法，其特征在于，所述方 法进一步包括，
    所述网络节点向所述数据传输控制节点上报MPTCP连接消息， 所述MPTCP连接消息用于指示所述MPTCP连接的状态，
    所述MPTCP连接消息中携带MPTCP连接信息，所述MPTCP连接 信息用于所述数据传输控制节点将所述多条TCP子流与所述MPTCP 连接相关联，使得所述MPTCP连接包含所述多条TCP子流。
  39. 根据权利要求32-38任一所述的方法，其特征在于，
    所述通信两端中的一端是接入所述网络节点的用户设备UE，另 一端是服务器；
    所述网络节点支持所述UE与所述服务器相通信。
PCT/CN2014/077061 2014-05-08 2014-05-08 数据传输控制节点、通信***及数据传输管理方法 WO2015168909A1 (zh)

Priority Applications (4)

Application Number Priority Date Filing Date Title
CN201480029919.1A CN105264845B (zh) 2014-05-08 2014-05-08 数据传输控制节点、通信***及数据传输管理方法
EP14891549.9A EP3133784B1 (en) 2014-05-08 2014-05-08 Data transmission control node, communication system and data transmission management method
PCT/CN2014/077061 WO2015168909A1 (zh) 2014-05-08 2014-05-08 数据传输控制节点、通信***及数据传输管理方法
US15/345,022 US10511535B2 (en) 2014-05-08 2016-11-07 Data transmission control node, communications system, and data transmission management method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2014/077061 WO2015168909A1 (zh) 2014-05-08 2014-05-08 数据传输控制节点、通信***及数据传输管理方法

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US15/345,022 Continuation US10511535B2 (en) 2014-05-08 2016-11-07 Data transmission control node, communications system, and data transmission management method

Publications (2)

Publication Number Publication Date
WO2015168909A1 true WO2015168909A1 (zh) 2015-11-12
WO2015168909A9 WO2015168909A9 (zh) 2016-03-17

Family

ID=54391995

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2014/077061 WO2015168909A1 (zh) 2014-05-08 2014-05-08 数据传输控制节点、通信***及数据传输管理方法

Country Status (4)

Country Link
US (1) US10511535B2 (zh)
EP (1) EP3133784B1 (zh)
CN (1) CN105264845B (zh)
WO (1) WO2015168909A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111213401A (zh) * 2017-08-18 2020-05-29 诺基亚技术有限公司 用于无线混合接入网络的业务分布控制
WO2022002215A1 (zh) * 2020-07-03 2022-01-06 华为技术有限公司 传输数据的方法和装置
US11258877B2 (en) 2018-07-26 2022-02-22 Netapp, Inc. Methods for managing workloads in a storage system and devices thereof

Families Citing this family (27)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3133784B1 (en) * 2014-05-08 2019-07-03 Huawei Technologies Co., Ltd. Data transmission control node, communication system and data transmission management method
KR101975684B1 (ko) 2014-07-21 2019-05-07 후아웨이 테크놀러지 컴퍼니 리미티드 링크 제어 노드, 링크 제어 방법 및 통신 시스템
EP3151504B1 (en) * 2015-09-30 2018-06-06 Tessares SA Method and device for establishing multipath network connections
US10505848B2 (en) * 2015-12-24 2019-12-10 Intel Corporation Congestion management techniques for communication networks
CN109451821B (zh) * 2017-06-30 2022-03-18 北京小米移动软件有限公司 数据分流指示方法及装置、数据分流方法及装置和接入点
US10389541B1 (en) 2017-10-19 2019-08-20 Amdocs Development Limited System, method, and computer program for implementing rating, charging and policy control of multipath communication services
KR101929995B1 (ko) * 2017-11-30 2018-12-17 울산과학기술원 무선통신 시스템의 단말 및 그의 데이터 전송 방법
US10931587B2 (en) * 2017-12-08 2021-02-23 Reniac, Inc. Systems and methods for congestion control in a network
EP3744063A1 (en) * 2018-01-26 2020-12-02 IDAC Holdings, Inc. Application mobility based on enhanced mptcp
CN108366104B (zh) * 2018-01-29 2020-11-13 北京奇艺世纪科技有限公司 一种数据下载控制方法和装置
EP3534574B1 (en) * 2018-02-28 2020-02-05 Deutsche Telekom AG Techniques for policy management of multi-connectivity network protocols
ES2903241T3 (es) * 2018-06-07 2022-03-31 Deutsche Telekom Ag Un sistema de comunicación para transmitir un segmento de protocolo de control de transmisión a través de una red de comunicación mediante el uso de un protocolo de control de transmisión de múltiples rutas, procedimiento y programa informático correspondientes
CN109150743B (zh) * 2018-09-14 2021-11-26 清华大学 一种网络拥塞控制策略切换方法及***
US20230144074A1 (en) * 2018-10-11 2023-05-11 Nokia Technologies Oy Active mode traffic steering and switching between cellular and non-cellular systems
US10659569B1 (en) * 2019-01-18 2020-05-19 Hewlett Packard Enterprise Development Lp End-to-end multipath TCP through network gateways
CN113316973A (zh) * 2019-01-18 2021-08-27 中兴通讯股份有限公司 移除多连接***中的用户平面连接的方法和装置
US11558733B2 (en) * 2019-07-10 2023-01-17 Samsung Electronics Co., Ltd. Managing sub-flow communications in user equipment
US11445431B2 (en) * 2019-07-12 2022-09-13 Qualcomm Incorporated Multi-link communication
CN110417650B (zh) * 2019-08-02 2021-12-10 广东中兴新支点技术有限公司 多链路数据分配方法、装置及电子设备
CN110475315B (zh) * 2019-08-19 2021-05-14 Oppo广东移动通信有限公司 网络连接控制方法及相关产品
CN112243253B (zh) * 2019-10-24 2022-07-08 北京大学 一种通信设备
CN111064704B (zh) * 2019-11-19 2021-02-09 中国科学院计算技术研究所 一种基于mptcp启动窗口自适应的数据传输方法、装置和介质
TWI793399B (zh) * 2020-02-14 2023-02-21 緯創資通股份有限公司 使用者裝置及資料流量傳遞之排程方法
CN111970334A (zh) * 2020-07-30 2020-11-20 杭州超级科技有限公司 基于tcp的数据传输方法、装置、电子设备及介质
CN112054966B (zh) * 2020-09-01 2023-11-07 首都师范大学 多路传输数据调度方法、装置及设备
CN114039926B (zh) * 2021-11-05 2023-10-03 北京字节跳动网络技术有限公司 传输控制协议确定方法、装置、可读介质及电子设备
US11962488B2 (en) 2022-07-01 2024-04-16 Cisco Technology, Inc. Supporting multipath transmission control protocol subflows using multipath links

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102404077A (zh) * 2011-11-30 2012-04-04 清华大学 基于喷泉码的多径tcp协议
CN103346963A (zh) * 2013-06-26 2013-10-09 南京昭视畅网络科技有限公司 一种基于预测到达时间的mptcp数据调度方法
WO2014044333A1 (en) * 2012-09-24 2014-03-27 Telefonaktiebolaget L M Ericsson (Publ) Traffic shaping and steering for a multipath transmission control protocol connection

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9503223B2 (en) * 2011-03-04 2016-11-22 Blackberry Limited Controlling network device behavior
US20130064198A1 (en) * 2011-09-14 2013-03-14 Qualcomm Incorporated Multipath transport tunnel over multiple air interfaces connecting wireless stations
US9537759B2 (en) * 2012-01-31 2017-01-03 Massachusetts Institute Of Technology Multi-path data transfer using network coding
US10075987B2 (en) * 2013-12-18 2018-09-11 Telefonaktiebolaget Lm Ericsson (Publ) Multipath TCP subflow establishing on single IP connection
EP3133784B1 (en) * 2014-05-08 2019-07-03 Huawei Technologies Co., Ltd. Data transmission control node, communication system and data transmission management method

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102404077A (zh) * 2011-11-30 2012-04-04 清华大学 基于喷泉码的多径tcp协议
WO2014044333A1 (en) * 2012-09-24 2014-03-27 Telefonaktiebolaget L M Ericsson (Publ) Traffic shaping and steering for a multipath transmission control protocol connection
CN103346963A (zh) * 2013-06-26 2013-10-09 南京昭视畅网络科技有限公司 一种基于预测到达时间的mptcp数据调度方法

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
See also references of EP3133784A4 *
TAO, YANG ET AL.: "Congestion Control for Multipath Transport Control Protocol Based on Compensating for RTT Mismatch", VIDEO ENGINEERING, vol. 37, no. 7, 2 April 2013 (2013-04-02), pages 97 - 100, XP008182930 *

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111213401A (zh) * 2017-08-18 2020-05-29 诺基亚技术有限公司 用于无线混合接入网络的业务分布控制
CN111213401B (zh) * 2017-08-18 2023-08-29 诺基亚技术有限公司 用于无线混合接入网络的业务分布控制
US11258877B2 (en) 2018-07-26 2022-02-22 Netapp, Inc. Methods for managing workloads in a storage system and devices thereof
WO2022002215A1 (zh) * 2020-07-03 2022-01-06 华为技术有限公司 传输数据的方法和装置

Also Published As

Publication number Publication date
WO2015168909A9 (zh) 2016-03-17
CN105264845B (zh) 2019-04-12
US10511535B2 (en) 2019-12-17
EP3133784A4 (en) 2017-04-26
EP3133784A1 (en) 2017-02-22
CN105264845A (zh) 2016-01-20
US20170078206A1 (en) 2017-03-16
EP3133784B1 (en) 2019-07-03

Similar Documents

Publication Publication Date Title
WO2015168909A9 (zh) 数据传输控制节点、通信***及数据传输管理方法
US10841815B2 (en) Link control node and method, and communications system
US10397188B2 (en) Access control apparatus, system, and method
US20150237525A1 (en) Traffic Shaping and Steering for a Multipath Transmission Control Protocol Connection
US20210329487A1 (en) Data transmission method and apparatus, and service switching method and apparatus
US20190028929A1 (en) Data-interface flow-splitting method, apparatus, terminal device, and computer storage medium
US20160309534A1 (en) Multipath tcp subflow establishing on single ip connection
WO2015039449A1 (zh) 一种小小区切换方法、基站和计算机存储介质
JP2012253750A (ja) ＭｉＡＮ及びＭｉＡＮ帯域幅集約方法並びに集約システム
WO2017219355A1 (zh) 多连接通信方法和设备
WO2021232841A1 (zh) 通信方法和装置
EP4138443A1 (en) Communication method and apparatus
WO2018082597A1 (zh) 拥塞控制方法及装置、基站
WO2015070442A1 (zh) 业务分流方法、控制网元、网关路由器及用户面实体
Masek et al. User performance gains by data offloading of LTE mobile traffic onto unlicensed IEEE 802.11 links
US20180184332A1 (en) Wireless network function configuration method, wireless network node, and core network device
WO2015143763A1 (zh) 一种负荷信息传递方法、***、网元及计算机存储介质
CN109155956A (zh) 一种数据传输方法、装置及***
WO2016029409A1 (zh) 一种数据传输方法及装置
WO2014183276A1 (zh) 一种无线网络的分流方法及设备
Karimi et al. Achieving high-performance cellular data Services with multi-network access
Nakauchi et al. vBS on the move: Migrating a virtual network for nomadic mobility in WiFi networks
WO2022205251A1 (zh) 信息收发方法，数据发送方法以及装置
WO2016119451A1 (zh) 一种用户设备选择网络的方法和装置

Legal Events

Date Code Title Description
WWE Wipo information: entry into national phase

Ref document number: 201480029919.1

Country of ref document: CN

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 14891549

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

REEP Request for entry into the european phase

Ref document number: 2014891549

Country of ref document: EP

WWE Wipo information: entry into national phase

Ref document number: 2014891549

Country of ref document: EP