WO2007098676A1 - Procédé de réassemblage de données dans un système de communication sans fil et appareil associé - Google Patents

Procédé de réassemblage de données dans un système de communication sans fil et appareil associé Download PDF

Info

Publication number
WO2007098676A1
WO2007098676A1 PCT/CN2007/000406 CN2007000406W WO2007098676A1 WO 2007098676 A1 WO2007098676 A1 WO 2007098676A1 CN 2007000406 W CN2007000406 W CN 2007000406W WO 2007098676 A1 WO2007098676 A1 WO 2007098676A1
Authority
WO
WIPO (PCT)
Prior art keywords
received
sequence number
pdu
data unit
arq
Prior art date
Application number
PCT/CN2007/000406
Other languages
English (en)
French (fr)
Inventor
Fei Shao
Xiaoxiao Zheng
Original Assignee
Huawei Technologies Co., Ltd.
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
Priority claimed from CN2006100593426A external-priority patent/CN101030842B/zh
Priority claimed from CN2006100346359A external-priority patent/CN101043301B/zh
Application filed by Huawei Technologies Co., Ltd. filed Critical Huawei Technologies Co., Ltd.
Priority to EP07702289A priority Critical patent/EP1993241B1/en
Priority to DE602007011221T priority patent/DE602007011221D1/de
Priority to AT07702289T priority patent/ATE492100T1/de
Publication of WO2007098676A1 publication Critical patent/WO2007098676A1/zh
Priority to US12/203,799 priority patent/US20090046626A1/en

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00Traffic control in data switching networks
    • H04L47/10Flow control; Congestion control
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00Traffic control in data switching networks
    • H04L47/10Flow control; Congestion control
    • H04L47/34Flow control; Congestion control ensuring sequence integrity, e.g. using sequence numbers
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L49/00Packet switching elements
    • H04L49/90Buffering arrangements
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L49/00Packet switching elements
    • H04L49/90Buffering arrangements
    • H04L49/9084Reactions to storage capacity overflow
    • H04L49/9089Reactions to storage capacity overflow replacing packets in a storage arrangement, e.g. pushout
    • H04L49/9094Arrangements for simultaneous transmit and receive, e.g. simultaneous reading/writing from/to the storage element
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04WWIRELESS COMMUNICATION NETWORKS
    • H04W28/00Network traffic management; Network resource management
    • H04W28/02Traffic management, e.g. flow control or congestion control
    • H04W28/10Flow control between communication endpoints
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04WWIRELESS COMMUNICATION NETWORKS
    • H04W8/00Network data management
    • H04W8/02Processing of mobility data, e.g. registration information at HLR [Home Location Register] or VLR [Visitor Location Register]; Transfer of mobility data, e.g. between HLR, VLR or external networks
    • H04W8/04Registration at HLR or HSS [Home Subscriber Server]

Definitions

  • the present invention relates to the field of wireless communication technologies, and in particular, to a method and apparatus for rearranging data in a wireless communication system. Background technique
  • Universal Mobile Telecommunication Systems is the third generation mobile communication system using WCDMA air interface.
  • the UMTS system is also commonly referred to as a WCDMA communication system.
  • the network elements can be divided into a RAN (Radio Access Network) and a core network (CN, Core Network).
  • the wireless access network is used to handle all wireless related functions, while the core network handles all voice calls, data connections and external network switching and routing within the UMTS system.
  • the purpose of the (Long Term Evolution) study is to provide a low-cost network that reduces latency, increases user data rates, improved system capacity and coverage, using a new system framework and physical layer to provide higher rates and more Good performance service.
  • Evolved wireless land access network
  • E-UTRAN including an evolved access network and an evolved base station.
  • E-UTRAN including an evolved access network and an evolved base station.
  • the structural framework of the radio interface protocol in the pre-evolutionary base station is divided into three layers, layer 3 is the radio resource control (RRC) layer, and layer 2 is divided into packet data convergence protocol PDCP and broadcast/multicast control BMC layer, wireless.
  • the Link Control (RLC, Radio Link Control) layer and the Medium Access Control MAC layer, and Layer 1 are physical layer PHYs, as shown in FIG.
  • the uplink and downlink packet service data transmission process passes through the PDCP/BMC layer, the RLC layer, the MAC layer, and the physical layer, and then enters the physical layer PHY, the opposite MAC layer, the opposite RLC layer, and the opposite end of the opposite end.
  • the flow direction of the PDCP/BMC layer is transmitted, and the signaling is directly from the RRC layer to the RLC layer.
  • the data header of the layer is added to each layer of data, and the peer layer of the peer layer is removed.
  • the data header of the data is removed, and then the data header is removed.
  • the content of the data is passed to the top.
  • the RLC layer and the MAC layer divide/concatenate the data from the upper layer, wherein the RLC layer divides the data sent by the upper layer (RLC SDU) into equal-sized data blocks, and adds header information to form a wireless
  • the link control protocol data unit RLC PDU (Protocol Data Unit) is sent to the MAC layer; at the physical layer, the data layer from the MAC layer is divided/cascaded to form a physical frame of appropriate size.
  • the receiving end performs the inverse process of the splitting/cascading, and the MAC layer rearranges the MAC PDU according to the sequence of the transmission sequence number, for example, the serial number of the transmission sequence number is 1, 2, 3, etc., the serial number is received according to The order of the data blocks is sorted. And the ordered data is sent to the RLC layer in an orderly manner; at the RLC layer, the RLC PDU is rearranged according to the header information (sequence number) added by the transmitting end, and after all the data constituting one SDU is correctly received, the RLC SDU is received.
  • the header information sequence number
  • each MAC PDU may include multiple RLC PDUs, and the MAC PDUs are rearranged and sent to the RLC layer, and the RLC is easily discontinuous according to them (partially through HARQ)
  • the retransmission is still not transmitted correctly.
  • the data judges the loss of the RLC PDU and requests the AQ to retransmit. After the ARQ retransmission, the RLC PDUs will be out of order, so the RLC PDUs should also be sorted.
  • the entities of the RLC sublayer support three types of services, transparent mode (TM) service, unacknowledged mode (UM) service, and acknowledged mode service (AM).
  • TM transparent mode
  • UM unacknowledged mode
  • AM acknowledged mode service
  • the RLC sublayer entity model is shown in Figure 2.
  • Transparent Mode TM services are performed by independent transmit and receive transparent mode entities.
  • the sending entity receives the SDU from the upper layer and divides it into the appropriate RLC PDU, without any overhead to transmit to the MAC.
  • the receiving entity receives the PDUs from the MAC sublayer, and then reassembles the PDUs into RLC SDUs for transmission to the upper layer.
  • the acknowledgment mode UM service is completed by the independent transmitting and receiving unacknowledged mode entity 4.
  • the sending entity receives the SDU from the upper layer into the RLC PDU of the appropriate size or connects the different SDUs into one RLC PDU, attaches an RLC header and puts it into the sending buffer to send it to the MAC sublayer through a certain logical channel.
  • the receiving end receives the PDU, goes to the head and reassembles it into an SDU and sends it to the upper layer.
  • the transmitted and received PDUs include control PDUs and service PDUs.
  • all transmitted service PDUs require confirmation by the peer entity to determine whether to retransmit or not.
  • the Control PDU is a status report and a reset request for the received PDU generated by the RLC entity itself.
  • the receiving side entity receives the PDU from the MAC sublayer, extracts the piggyback status information, puts it into the receiving buffer, waits for the complete PDU reassembly SDU to be sent to the upper layer, or sends an erroneous receiving acknowledgement through its transmitting side to request the peer entity to resend the PDU. .
  • the receiver rearrangement scheme has the following mechanisms:
  • AM business rearrangement and reorganization mechanism 1.
  • the movement of the window is achieved by updating the lower bound of the window.
  • the PDUs received with different serial numbers from the window are buffered in the receive buffer, and the PDUs whose received sequence numbers are outside the window are deleted. As the receiving end, it can only passively wait for VR ( R ), and there is no timer to control the window movement.
  • the sender can limit the PDU transmission rate by timer and / or the maximum number of retransmissions. A timer is started for each SDU to be sent.
  • the SDU discard is discarded and the receiving end is notified of the window update.
  • the maximum number of retransmissions is no SDU discard configured ( The RLC entity's reset is triggered in the case of a loss of traffic data unit.
  • the reorganization mechanism can reorganize the complete SDU and send it to the upper layer according to the LI instruction according to the set order or out of order.
  • the MCCH receiver reorder reorganization mechanism is introduced due to the MCCH periodic retransmission feature. This mechanism also has a The window waits for the initial transmission to lose the retransmitted PDU. To ensure real-time performance, the out-of-order reassembly SDU guarantees the reassembly rate of the receiving end.
  • the window used is (VR(UOH)-OSD_Window_Size, VR(UOH)), where VR(UOH) is the largest sequence number of the received PDU, and the maximum value of the cache wait is configured by the upper layer.
  • VR(UOH) is the largest sequence number of the received PDU, and the maximum value of the cache wait is configured by the upper layer.
  • Timer Timer The OSD is used to control the update of the VR (UOH). Each time the VR (UOH) updates the timer reset, the timer expires and all PDUs in the cache are deleted.
  • the PDUs buffered in the window are reorganized, and the complete SDU is recovered according to the LI indication and sent to the upper layer regardless of the SDU order.
  • DAR is an abbreviation for repeated avoidance and reordering. Due to the MTCH selective merge mechanism of MBMS in WCDMA R6, MTCH reception will have problems of copy and out-of-order arrival. Therefore, the MTCH receiver introduces a window mechanism for reordering.
  • the window is (VR(UDH)-DAR_Window_Size, VR(UDH)] 5 where VR(UDH) indicates that the largest sequence number of the PDU has been received, and the maximum value of the buffer wait is configured by the upper layer.
  • the actual window is (VR(UDR), VR(UDH)], where VR(UDR) is the sequence number of the next PDU that is sent to the upper layer in order, that is, the PDUs smaller than this SN have been sequentially sent to the upper layer.
  • the PDU with the serial number VR (UDR) After the PDU with the serial number VR (UDR) is received, check the sequence number of the smallest PDU that is not correctly received in the window, update the VR (UDR) to this value, and all the serial numbers are smaller than the updated VR (UDR) The PDU is sent to the high-level reorganization. If the window advances to VR ( UDR ) ⁇ VR(UDH)-DAR—Window— Size+1, the update VR(UDR) is the smallest in the update window (VR(UDH)- DAR_Window_Size, VR(UDH)] The PDU that is sent updates the VR (UDR) to this SN, and sends all PDUs whose sequence number is less than the updated VR (UDR) to the high-level reorganization.
  • the timer Timer-DAR and the state variable VR(UDT) control the receiving window to not move for a long time.
  • VR ( UDT ) is initially set to the maximum sequence number in the window, and the timer Timer-DAR is started at the same time.
  • the timer restarts and VR (UTT) is reset to the maximum sequence number in the window. If the timer expires, all PDUs whose sequence number is less than or equal to VR (UDT) and consecutive to VR (UDT) are sent to the upper layer.
  • VR (UDR) is updated to the smallest number that is not received in the current buffer, and VR is reset. ) is the maximum serial number of the PDU in the window, restart the timer.
  • HSPA is an abbreviation for High Speed Packet Access
  • the rearrangement mechanism of HSDPA short for High Speed Downlink Packet Access
  • the purpose is to perform the reordering of HARQ (abbreviation of hybrid automatic repeat request) for different process retransmissions, and perform reordering according to the TSN sequence number to ensure that the received MAC-hs PDU is sequentially restored to a MAC-d PDU and sent to the MAC-d. entity.
  • LTE Long Term Evolution
  • UTRAN Long Term Evolution
  • the existing protocol layer structure causes many repetitive functions, such as retransmission of RLC and MAC sublayers, split cascading, and so on. In order to reduce latency and simplify the protocol, these repetitive functions must be considered for streamlining.
  • the all-IP requirement proposed by the LTE system that is, the network transmission is all based on the IP packet service, and this new requirement also needs a response mechanism to guarantee.
  • the MAC structure in the recently proposed LTE system is shown in FIGS. 5 and 6.
  • FIGS. 5 and 6 There are four kinds of three rearrangement mechanisms in the WCDMA R6 system discussed above, which can be adapted. Different rearrangement requirements, but complex structure, redundant functions, reduced response speed, can not be applied to existing LTE networks, for example, separate automatic retransmission request ARQ and hybrid automatic retransmission request HARQ, inefficient, increase The delay of ARQ retransmission.
  • both the MAC layer and the RLC layer need to perform data rearrangement. That is, in the RLC layer, the RLC SDU performs data segmentation according to the RLC PDU size configured by the upper layer, and uniformly numbers the segmented PDUs for use as receiving and retransmitting identifiers, and the data to be retransmitted is stored in the ARQ cache.
  • the number of the RLC PDU in the protocol is called the sequence number (SN, Sequence Number), the sequence number of the acknowledge mode is 12 bits, and the sequence number of the unacknowledged mode is 7 bits.
  • a service is configured with its corresponding RLC entity. The size of the RLC PDU is given.
  • the RLC layer rearranges and reassembles the RLC PDU according to the SN.
  • the RLC PDU is added to the MAC-d entity to form a MAC-d PDU
  • the MAC-hs/MAC-e entity is cascaded, and the data header is added, including the transmission sequence number. (TSN)
  • TSN transmission sequence number
  • a physical frame is formed and buffered in the HARQ buffer, and the MAC layer HARQ is rearranged according to the TSN at the receiving end MAC layer, and the ordered data is correctly received and sent to the RLC layer.
  • the protocol layer of the receiving end when the protocol layer of the receiving end is different, for example, the MAC layer and the RLC layer of the receiving end respectively rearrange the data of the receiving end according to the added sequence number, thereby increasing the overhead of the system, and the MAC The layer and the RLC layer are respectively located at the base station and the RNC. Through these two rearrangements, the delay of data transmission is increased.
  • Embodiments of the present invention provide a data rearrangement method and apparatus in a wireless communication system, which reduces data transmission delay and reduces data overhead by reducing the number of data reordering between protocol layers and simplifying data header information.
  • An embodiment of the present invention provides a method for rearranging data in a mobile communication system, including the steps of: receiving, by a media access control entity at a receiving end, a correctly decoded transport block, and sending the data block to a link control entity;
  • the link control entity at the receiving end decapsulates the received transport block, and rearranges and reassembles the decapsulated higher layer protocol data unit to obtain a service data unit.
  • the method further comprises:
  • the transport block When the received transport block has protocol data units of a plurality of link control entities, the transport block is decomposed and the decomposed sub-transport block is transmitted to the corresponding link control entity.
  • An embodiment of the present invention provides a data rearrangement apparatus in a wireless communication system, including: a first retransmission unit and a second retransmission unit,
  • the first retransmission unit is configured to send the received transport block to the second retransmission unit, where the second retransmission unit is configured to receive the transport block sent by the first retransmission unit, and the transport block Performing decapsulation, obtaining a high-level protocol data unit, and rearranging and reorganizing the high-level protocol data unit to obtain a service data unit, and sending the data unit to the upper layer.
  • the device further comprises:
  • a decomposing unit respectively connected to the first retransmission unit and the second retransmission unit, configured to decompose a transport block composed of protocol data units of the plurality of link control entities, and send the decomposed sub-transport block to the corresponding chain Road control entity.
  • the interaction delay of data between the two layers is greatly reduced, which is beneficial to the improvement of data transmission efficiency.
  • TSN transmission sequence number
  • the rearrangement delay improves data transfer efficiency and is especially advantageous for business data units with out-of-order delivery attributes.
  • the receiving end performs rearrangement/reassembly on the received demultiplexed data packet according to the sequence number SN, and sets a receiving buffer condition, and performs rearranged data according to the receiving buffer condition.
  • Each received SDU or PDU has a sequence number SN.
  • the HARQ sends the PDUs that need to be rearranged to each ARQ rearrangement unit for reordering according to the HARQ PDU header information.
  • DRAWINGS 1 is a schematic structural diagram of a radio interface protocol in the prior art
  • FIG. 2 is a schematic structural diagram of a radio link control RLC entity model
  • FIG. 3 is a schematic structural diagram of a MAC on a user equipment side
  • FIG. 4 is a schematic diagram of a MAC structure on the network side
  • FIG. 5 is a schematic diagram of a MAC structure of a UL in an eNB in an LTE network
  • FIG. 6 is a schematic diagram of a MAC structure of a user equipment DL in an LTE network
  • FIG. 7 is a schematic structural diagram of an evolved wireless land access system
  • FIG. 8 is a schematic diagram of a data rearrangement apparatus in an embodiment of the present invention.
  • FIG. 9 is a flowchart of a data rearrangement method in an embodiment of the present invention.
  • FIG. 11 is a schematic diagram of a protocol layer in a data rearrangement method according to an embodiment of the present invention.
  • Figure 12 is a diagram showing an embodiment of a rearrangement buffer of a receiving end in a data rearrangement method in an embodiment of the present invention. detailed description
  • E-UTRAN evolved wireless land access
  • the E-UTRAN includes: an evolved access gateway 1 (E-AGW) and an evolved base station 2 (E-NodeB) two-layer architecture, which in turn manages a plurality of user equipments UE3.
  • the E-Node and the E-AGW may establish a connection directly or may be connected through an IP network.
  • the E-AGW is located in the Evolved CN of the evolved core network evolution core network, and has the same function as the radio network controller RC in the R6 protocol, and is used to manage different E-NodeBs, and each E-NodeB directly
  • the E-AGW is connected to manage users in the cell.
  • the E-NodeB includes an evolved medium access control layer MAC and an evolved physical layer PHY
  • the E-AGW includes an evolved packet data convergence protocol layer.
  • FIG. 8 is a schematic structural diagram of a data rearrangement apparatus in a mobile communication system according to the present invention.
  • the apparatus may be located in an evolved base station or in a user equipment UE.
  • the apparatus includes: a first retransmission unit 11 and a second retransmission unit 12, and may further include a decomposition unit 13 (shown by a broken line in the figure).
  • the first retransmission unit 11 is configured to send the received transport block directly to the second retransmission unit, and the second retransmission unit 12 is connected to the first retransmission unit 11 and receives the
  • the transport block is decapsulated to obtain a higher layer protocol data unit, and the high layer protocol data unit is rearranged and reassembled to obtain a service data unit, which is sent to a higher layer.
  • the decomposing unit 13 is respectively connected to the first retransmission unit 11 and the second retransmission unit 12, and is configured to decompose a transport block composed of protocol data units of multiple link control entities, and after decomposing The sub-transport block is sent to the corresponding link control entity.
  • the first retransmission unit 11 includes: a hybrid automatic retransmission subunit 111 and a judging subunit 112, where the hybrid automatic retransmission subunit 111 is configured to retransmit the received transport block; the judging subunit 112, And determining whether the number of retransmissions of the transport block exceeds a preset maximum number of retransmissions, and if the maximum number of retransmissions is exceeded, notifying the second retransmission unit that the data in the transport block is lost.
  • the second retransmission unit 12 includes: at least one automatic request retransmission subunit 121, at least one reordering subunit 122, and at least one recombination subunit 123.
  • the automatic request retransmission sub-unit 121 is connected to the hybrid automatic retransmission sub-unit 111 or the decomposing unit 13 for decapsulating the received transport block or the sub-transport block and transmitting the high-layer protocol data obtained after decapsulation.
  • the rearrangement subunit 122 is connected to the automatic request retransmission subunit 121, rearranges the received higher layer protocol data unit according to the sequence number or the segmentation sequence number, and sends the rearranged higher layer protocol data unit.
  • the recombination sub-unit 123 is connected to the re-arrangement sub-unit 122, and is configured to re-reassemble the received high-level protocol data unit and obtain a service data unit.
  • the radio link control layer RLC and the medium access control layer MAC are combined, thereby reducing the transmission delay of the interface between the protocol layers, and even ignoring the interface delay.
  • the service data unit ie, IP data packet
  • the rearrangement device is located in the user equipment:
  • Another implementation manner is that the rearrangement device is located in the evolved base station. The following two implementation processes Describe:
  • An implementation manner is: in the evolved base station, the media access control layer of the eNodeB divides or/and cascades the packet, obtains a protocol data unit, and encapsulates the information header to form a transport block (TB), and then transmits the data to the user through the physical layer.
  • the receiving end of the device UE after receiving the transport block by the media access control layer MAC of the receiving end of the user equipment, after the retransmission of the transport block by the HARQ (Hybrid Automatic Repeat Request) layer
  • the correct decoding transport block is obtained and sent to the ARQ (Automatic Repeat Request) layer of the receiving end, and the automatic request retransmission ARQ layer de-encapsulates the received transport block to obtain an out-of-order protocol.
  • the data unit then rearranges and reassembles the protocol data unit according to the sequence number or the segmentation sequence number to obtain an ordered service data unit and sends it to the upper layer.
  • the ARQ refers to ensuring the reliability of the transmission by retransmission.
  • the HARQ protocol is an improvement of the ARQ protocol, which is to introduce a FEC subsystem in the ARQ system. This FEC subsystem is used to correct errors to reduce the number of retransmissions.
  • the transmitting end starts operating on the next data block only after the transmitted data block is correctly received.
  • the system uses the 1-bit serial number to distinguish the current data block from the next data block to be transmitted. A bit is used to confirm whether the transport block is correctly decoded (using a reply ACK or an error acknowledge NACK).
  • Another implementation manner is the reverse process of the foregoing implementation manner, and the implementation process is basically the same as the foregoing manner, where the difference is that the sending end device in the user equipment first splits or/and cascades the service data units that need to be transmitted. After the segmentation or/and the concatenated protocol data unit encapsulation information header is formed, the transport block is sent to the evolved base station through the physical layer, and the evolved base station is processed in the same manner as the receiving end of the user equipment, as described above. No longer.
  • the function of the physical layer is to perform operations such as modulation, coding, and time-frequency synchronization on the received data packet; the MAC layer performs media access control, dynamic scheduling, and processing priorities of the UE data stream.
  • the function of automatically requesting retransmission of ARQ is retransmission over hybrid automatic retransmission HARQ to prevent data errors caused by HARQ failure.
  • ARQ also has segmentation, reassembly, split cascading, and sequential delivery of out-of-order packets. , copy detection, flow control and other functions.
  • the present invention further provides a method for rearranging data in a communication system, and a flowchart of the method is shown in FIG. 9. The method includes the following steps:
  • Step S11 The media access control entity at the receiving end receives the correctly decoded transport block and sends it to the link control entity.
  • Step S12 The link control entity at the receiving end decapsulates the received transport block, and rearranges and reassembles the decapsulated higher layer protocol data unit to obtain a service data unit.
  • the method further includes the steps of: if the received transport block is composed of protocol data units of the plurality of link control entities, decomposing the transport block, and transmitting the decomposed sub-transport block to the corresponding Link control entity.
  • the present invention will be described below with the user equipment as the transmitting end and the evolved base station as the receiving end, and vice versa.
  • the segmentation and/or cascading function entity of the MAC layer of the sender media segment controls the segmentation and/or cascading of the upper service data unit SDU (ie, the IP packet), and sets the corresponding sequence number (such as a map).
  • SDU the upper service data unit
  • sequence number such as a map
  • the protocol data unit PDU is sent to automatically request retransmission of the ARQ entity; and the MAC layer's generated transport block function entity adds the necessary header information (TB header) to the received PDU (eg The service priority queue number is multiplexed to form a transport block (TB) and then sent to the hybrid automatic retransmission HARQ function entity.
  • the HARQ of the receiver receives the correctly decoded TB after retransmission, and delivers the correct reception to the ARQ. TB, no rearrangement.
  • the ARQ entity After the ARQ entity obtains the out-of-order protocol data unit PDU by decomposing and decapsulating, the status report is not sent, and only the PDU is placed in the rearrangement buffer, and the rearrangement function entity pairs the PDU according to the sequence number SNA segment sequence number SSN. Perform rearrangement.
  • the SN is the sequence number of the first transmitted PDU
  • the SSN is the sequence number of each segment in the first transmission PDU, so that the receiving end can perform retransmission, rearrangement, and reassembly through the SN and the SSN.
  • the reorganization function entity removes the information header in the PDU, it reorganizes to obtain the correct and ordered service data unit SDU, and sends it to the upper layer.
  • the HARQ at the receiving end can inform the ARQ that the transport block has been lost through the inter-layer source language, and the triggering of the status report can be triggered periodically and controlled by the link.
  • the entity is automatically generated or triggered by the maximum number of retransmissions of HARQ.
  • the sender has 2 UDPs (ie, IP packets)
  • the segmentation is performed. After cascading, it is divided into 3 PDUs, and the sequence number (SN) is 1, 2, 3 respectively.
  • the TB is formed, it is retransmitted in HARQ.
  • the receiving end sends the TB to the rearrangement entity after decapsulation.
  • the arranging entity rearranges according to the serial number or the segment serial number, and then reorganizes the protocol data unit after removing the information header.
  • the sequence in the rearrangement buffer of the receiving end is as shown in FIG.
  • the rearrangement of the HARQ of the receiving end is cancelled, that is, after the receiving end HARQ receives the correct TB, it is sent to the ARQ immediately, and the rearrangement is performed only in the ARQ, and for the HARQ transmission sequence number ( TSN can be eliminated, reducing the overhead of data. Therefore, the present invention not only reduces the number of reordering of data between protocol layers, but also simplifies the data header information, and also reduces the data transmission delay and reduces the data overhead.
  • the following takes the LTE network as an example to discuss the data rearrangement and reorganization method of the present invention, mainly for the AM and UM services.
  • the UM service only the HARQ service is used, and the reordering mechanism is only for the HARQ out of order.
  • each HARQ process at the receiving end receives the corresponding TB (transport block). If the check is correct, it is sent to the MUX unit for demultiplexing to each ARQ entity.
  • the ARQ rearrangement function unit performs PDU rearrangement according to the PDU sequence number and reassembles the SDU to the upper layer, triggering retransmission or deleting the corresponding retransmission buffer PDU.
  • the HARQ process at the receiving end receives the corresponding TB. If the check is correct, it is sent to the MU unit for demultiplexing to each ARQ entity.
  • the ARQ entity does not retransmit and only performs split cascading.
  • the rearrangement function unit sorts the HARQ retransmissions and reorganizes the SDUs to the upper layers. For some real-time services that are prohibited by HARQ, there is no need to reorder.
  • the method of the present invention sets a HARQ receive buffer and an ARQ receive buffer for buffering data waiting for HARQ and ARQ retransmission, respectively.
  • the present invention proposes a reordering mechanism, including a window control and a timer control mechanism, for distinguishing between HARQ retransmission out-of-order and ARQ retransmission out-of-order, since a cache is used to buffer the out-of-order caused by HARQ and ARQ retransmission. wait.
  • the caching and reordering mechanisms are discussed below. 1, HARQ cache
  • the SN indicates the highest sequence number of the PDU received in the cache
  • the SN indicates the next PDU sequence number sent to the upper layer or the next function unit in the cache
  • HARQ—RcvWindow—Size represents the maximum value of the HARQ cache waiting for the appropriate QoS of the upper layer configuration
  • the HARQ cache window is defined as (Highest_redforward—SN - HARQ—Window—Size, Highest_received_SN).
  • ARQ_RcvWindow_Size is the maximum ARQ cache wait time configured by the upper layer.
  • the definition of the ARQ cache window includes:
  • ARQ— RcvWindow— Size defines the ARQ cache window as (Next—expected—SN,
  • the HARQ receiving window reference boundary may also be set as the rearrangement mechanism. That is to say: HARQ—RcvWindow—Edge is the boundary between the actual HARQ buffer and the ARQ buffer, and numerically corresponds to the PDU sequence number received in sequence by the next HARQ.
  • HARQ_Window_ Size When the upper bound of the window is the lower bound of the window formed by the Highest-received-SN update, if the SN is Highest-accepted-SN- HARQ-Window-Size PDU already exists in the cache, the window lower bound HARQ-RcvWindow-Edge is updated to The smallest PDU sequence number that is not received in the window.
  • the ARQ window becomes (Next-expected_SN, HARQ_RcvWindow_Edge), and other ARQ related operations are the same as the above rearrangement mechanism.
  • the timer Timer_HARQ (T_SN) is defined to control the HARQ window movement.
  • the T-SN is initially set to the maximum sequence number in the buffer, Highest_received-SN, and each timer Timer_HARQ is started.
  • T_SN the timer restarts and the T-SN is reset to the maximum sequence number in the window.
  • T_SN>HARQ—RcvWindow_Edge HARQ—Rcv Window—Edge is updated to the lowest sequence number of the PDU that is not received in (Highest-received-SN, T-SN), and the T_SN is reset in the window.
  • the maximum number of the PDU restart the timer.
  • the status variable HARQ_RcvWindow_Edge_T indicates the HARQ receive window boundary, which is the maximum sequence number of the received PDU that the timer has timed out.
  • Highest_received—SN indicates the highest sequence number of the PDU received in the cache.
  • the SN indicates the next PDU sequence number sent to the upper or next functional unit in the cache.
  • the SN can be smaller than HARQ-RcvWindow—Edge—T.
  • ARQ—RcvWindow—Size is the maximum ARQ cache waiting for high-level configuration (large window) Small) upper limit.
  • the ARQ receive buffer is still controlled by the window, and the receive window is (Next—expected—SN, HARQ—RcvWindow—Edge—T.
  • the general cache wait maximum is less than or equal to ARQ—RcvWindow—Size, if HARQ_RcvWindow—Edge— ⁇ - Next—expected— SN > ARQ - RcvWindow - Size, ARQ window is set to (Next - expected - SN, Next - expected - SN + ARQ - Rev Windo w - Size)
  • the receiver sends a window overflow indication to the sender, and the sender sends the corresponding ARQ entity new data. The transmission will be suspended and only retransmitted will not allow new data transfers.
  • the HARQ and ARQ buffer wait maximum (ie, window size) can be configured by the upper layer according to the service.
  • the step number may indicate the execution order, or may not be the execution order
  • Step 1 is established within (Next - expected - SN, Highest - received - SN ). If the SN corresponding data is already in the cache, delete the data; if X is not ( Next - expected - SN, Highest Within the received- SN), the PDU is placed in the cache according to the sequence number;
  • Step 2 If x> Highest—received — if yes, if yes
  • ARQ Rev Window— Size, ⁇ 1 J ⁇
  • Trigger status report according to the status report trigger principle; update Highest—received—SN is ;
  • Trigger the status report according to the status report trigger principle; update Highest-received-SN to ⁇ ; Step 3: If x Next - expected - SN, then ⁇
  • the step number may indicate the execution order, or may not be the execution order
  • Step 1 If X is within (Next_expected_SN, Highest_ received_SN) ⁇ If the SN corresponding data is already in the cache, delete the data. Otherwise, put the PDU into the cache according to the serial number.
  • Step 2 If x> Highest— received—SN, Bay' J ⁇
  • Step 3 If SN is equal to HARQ - RcvWindow - Edge PDU is in the cache, then ⁇
  • Step 4 The corresponding status report is triggered according to the updated HARQ—RcvWindow—Edge and Status Report Trigger Principles;
  • Step 5 Check if HARQ - RcvWindow - Edge - Next - expected - SN > ARQ - RcvWindow - Size, then ⁇ send new data suspension message to the sender, update the ARQ window to (Next - expected - SN, Next - expected - SN + ARQ_RcvWindow_Size ); Delete SN from Next - expected - SN + ARQ_RcvWindow_Size to HARQ_RcvWindow_Edge;
  • the status report is triggered according to the status report trigger principle. ⁇
  • Step 1 If there is a PDU in the buffer and Timer_HARQ is not started, then ⁇ trigger timer Timer-HARQ;
  • T_SN is set to the largest PDU sequence number in the cache.
  • Step 3 If the timer expires, HARQ_RcvWindow—Edge ⁇ T—SN, then ⁇ HARQ_RcvWindow_Edge is updated to T—SN+1;
  • Step 4 Check, if HARQ_RcvWindow_Edge-Next-expected-SN>ARQ_RcvWindow_Size, then ⁇ send new data suspension message to the sender, update the ARQ window to (Next-expected_SN, Next-expected-SN+ARQ_Rcv Window- Size ); Delete SN from Next - expected - SN + ARQ - Rev Window - Size to HARQ_RcvWindow_Edge; Trigger status report according to the status report trigger principle.
  • the step number may indicate the execution order or may not be the execution order.
  • HARQ_RcvWindow_Edge_T y
  • the received transport block is composed of protocol data units of multiple link control entities
  • the correct transport block is first decomposed, and then the decomposed sub-transport block is sent to the corresponding link control entity.
  • the implementation process of the respective link control entity is the same as the implementation process of the protocol data unit that receives the single link control entity.
  • the receiving end performs rearrangement/reassembly on the received demultiplexed data packet according to the sequence number SN, and sets a receiving buffer condition, and performs rearranged data according to the receiving buffer condition.
  • Each received SDU or PDU has a sequence number SN.
  • the SN used for reordering and reassembling is the only SN of the ARQ, that is, the SDU SN of the high layer or the ARQ SN of the ARQ. If there is a segmentation subsequence after the retransmission split cascading, it is necessary to collect the multiple slice ends corresponding to each SN.
  • the ARQ unit is rearranged/reorganized according to its sequence number SN; the rearrangement is a reordering of the disorder between reception and transmission, and the reassembly is to reorganize the SDU by PDU. Recombination includes reordering of reordered submissions and reorganization of in-order submissions.
  • the HARQ sends the PDUs that need to be rearranged to each ARQ rearrangement unit according to the HARQ PDU header information for rearrangement.
  • the delay of data interaction between layers can be greatly reduced, and even the ignoring delay can be achieved.
  • the rearrangement retransmission function can be implemented at the base station, and can be completed at one time without splitting into two separate re-arrangements, thereby avoiding an increase in ARQ retransmission delay and reducing system complexity. Sex, increased speed.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Computer Security & Cryptography (AREA)
  • Databases & Information Systems (AREA)
  • Mobile Radio Communication Systems (AREA)
  • Detection And Prevention Of Errors In Transmission (AREA)
  • Communication Control (AREA)
  • Radar Systems Or Details Thereof (AREA)

Description

无线通信***中数据的重排方法及装置 本申请要求分别于 2006年 03月 03日、 2006年 03月 22日提交中国 专利局、 申请号分别为 200610059342.6、 200610034635.9、 发明名称分别 为"移动通信***中数据的重排方法及其装置" 、 "一种无线通信***中的 数据重排重組方法及其基站" 的中国专利申请的优先权, 其全部内容通 过引用结合在本申请中。 技术领域
本发明涉及无线通信技术领域, 特別是涉及一种无线通信***中数 据的重排方法及装置。 背景技术
通用移动通信*** ( UMTS , Universal Mobile Telecommunication Systems )是采用 WCDMA空中接口的第三代移动通信***。 通常也把 UMTS***称为 WCDMA通信***。 从功能上, 网络单元可以分为无线接 入网络(RAN, Radio Access Network )和核心网 ( CN, Core Network )。 其中无线接入网絡用于处理所有与无线有关的功能,而核心网处理 UMTS ***内所有话音呼叫、数据连接与外部网络的交换和路由。 长期演进 LTE
( Long Term Evolution )的研究的目的是提供一种能够降低时延、提高用 户数据速率、 改进的***容量和覆盖的低成本的网络采用全新的***框 架及物理层, 以提供更高速率, 更好性能的服务。 演进无线陆地接入网
( E-UTRAN ) , 包括演进接入网和演进基站。 但是, 针对目前的演进基 站中还没有提出对需要传输数据的进行怎样的重排, 才能减少协议层间 的重排次数, 以及传输时延等问题。
演进前基站中的无线接口协议的结构框架分为 3层, 层 3为无线资源 控制( RRC, Radio Resources Control)层,层 2分为分组数据会聚协议 PDCP 与广播 /多播控制 BMC层、 无线链路控制 (RLC, Radio Link Control)层 和媒体接入控制 MAC层, 层 1是物理层 PHY, 具体如图 1所示。 上下行的 分组业务数据传输过程依次通过 PDCP/BMC层、 RLC层、 MAC层和物理 层, 然后进入对端的物理层 PHY、 对端 MAC层、 对端 RLC层和对端 PDCP/BMC层的流向进行传输,其信令则是直接从 RRC层到 RLC层。在上 述数据从高层到低层的数据传递过程中, 每经过一层数据都会被加上本 层的数据头, 到了对端的对等层, 通过解析该数据, 除去该数据的数据 头, 再将该数据的内容传递给高层。
在数据传输过程中, RLC层和 MAC层会对高层过来的数据进行分割 / 级联, 其中 RLC层将高层送来的数据 ( RLC SDU )分割为大小相等的数 据块, 加上头信息, 组成无线链路控制协议数据单元 RLC PDU ( Protocol Data Unit)送往 MAC层; 在物理层, 将 MAC层过来的数据分割 /級联组成 大小合适的物理帧发送出去。 接收端执行分割 /级联的逆过程, MAC层根 据传输序列号的顺序对 MAC PDU进行重排, 比如, 传输序列号为 1、 2、 3之类的序列号, 该序列号是根据接收到数据块的先后顺序进行排序的。 并把有序接收正确的数据送 RLC层;在 RLC层,根据发送端加的头信息(序 列号)对 RLC PDU进行重排, 当组成一个 SDU的所有数据都接收正确之 后, 将此 RLC SDU向高层传递, 其中, MAC层重排的是 MAC PDU, 而 每个 MAC PDU可能包括多个 RLC PDU, MAC PDU重排好之后送到 RLC 层, RLC很容易根据其中不连续的(部分经过 HARQ重传还是没有传输正 确)数据判断 RLC PDU的丢失, 请求 A Q重传。 经过 ARQ重传之后 RLC PDU又会乱序, 因此还应对 RLC PDU进行排序。
RLC子层的实体共支持三种类型业务, 透明模式 (TM)业务、 无确认 模式 (UM)业务和确认模式业务 (AM)。 RLC子层实体模型如图 2所示。 透 明模式 TM业务是由独立的发送和接收透明模式实体完成的。 发送实体从 高层接收 SDU划分成适当的 RLC PDU, 无须加任何开销传送给 MAC。 接 收实体从 MAC子层接收到 PDU, 再将这些 PDU重组成 RLC SDU传送给上 层。
无确认模式 UM业务是由独立的发送和接收无确认模式实体 4完成 的。 发送实体从高层接收 SDU分割成适当大小的 RLC PDU或是将不同 SDU进行连接组成一个 RLC PDU, 附上一个 RLC头并放入发送緩存器通 过某一个逻辑信道将其发送到 MAC子层。 接收端接收 PDU, 去头并将其 重组成 SDU发送到高层。 确认模式 AM业务中, 发送和接收的 PDU包括控制 PDU和业务 PDU。 在确认模式下, 所有发送的业务 PDU都需要对等实体的确认来决定重传 与否。控制 PDU是由 RLC实体自身生成的一些针对接收 PDU的状态报告以 及复位请求等。接收侧实体从 MAC子层接收 PDU, 提取出捎带状态信息, 放入接收緩存器, 等待完整 PDU重组 SDU发送到上层, 或是再通过其发 送侧发送错误的接收确认要求对等实体重发 PDU。
对于 MAC子层, UE和 UTRAN侧的结构有所不同, 分别如图 3、 图 4 所示。 根据 WCDMA版本 R6, 接收端重排方案有如下几种机制:
1、 AM业务重排重组机制:
在 AM模式业务中, 为支持重传, 需要有一个窗口机制支持重排序功 能。 窗口为 ( VR(R), VR(MR) ) , 其中 VR ( R )为下一个按序接收的 PDU 序号, VR ( MR ) = VR ( R ) + Configured— Rx—Window_Size,緩存等待 最大值由高层配置。
实际的窗口为 ( VR(R), VR(H) ) , 其中 VR(H)为接收到的 PDU最大的 序号, VR ( H ) < = VR(MR)。 窗口的移动是通过更新由窗口下界实现的。 收到与窗口内不相同序号的 PDU被緩存在接收緩存内, 收到序号在窗口 外的 PDU被删除。 作为接收端只能被动等待 VR ( R ) , 没有计时器等方 法控制窗口移动。 发送端可以通过计时器和 /或最高重传次数限制 PDU传 输速率。 对于每一个要发送的 SDU都启动计时器, 当计时器超时或 PDU 达到最高重传次数进行 SDU discard (业务数据单元丢弃) 并通知接收端 进行窗口更新, 最高重传次数达到没有 SDU discard configured (业务数据 单元丟失配置) 的情况下触发 RLC实体的 reset (复位) 。
重组机制可以根据设定的按序还是乱序, 根据 LI指示将完整 SDU进 行重组并送到高层。
2. UM业务的 SDU乱序重排重组
普通的 UM业务由于没有重传, 不需要重排, 只要筒单地按序重组, 出现丟失的 PDU时, 将此 PDU所涉及的 SDU都删除。
但自从 R6进入 MBMS (广播组播业务的缩写)业务后由于 MCCH 周期重传特性, 引入 MCCH接收端乱序重组机制。 这个机制也是有一个 窗口等待初始传输丢失重传的 PDU, 为了保证实时性, 乱序重组 SDU保 证了接收端重组速率。
使用的窗口为 ( VR(UOH)-OSD— Window— Size, VR(UOH) ) , 其中 VR(UOH)为收到了 PDU最大的序号, 緩存等待最大值由高层配置。 对于 收到序号在窗口内的 PDU, 緩存起来, 将序号在窗口外的 PDU进行 VR ( UOH )更新, 窗口上界移动带动窗口的更新。 计时器 Timer— OSD用于 控制 VR(UOH)的更新, VR(UOH)每一次更新计时器重置, 计时器超时, 緩存内所有 PDU被删除。
对于缓存在窗口内的 PDU进行重组,根据 LI指示恢复出完整的 SDU 并送到高层而不管 SDU顺序。
3. DAR重排
DAR是重复避免和重排序的英文缩写,由于 WCDMA R6 中 MBMS 的 MTCH选择性合并机制, MTCH接收会出现副本和乱序到达的问题, 因此, 在 MTCH接收端引入窗口机制进行重排序。
窗口为( VR(UDH)-DAR— Window— Size, VR(UDH)]5 这里 VR( UDH ) 表示收到了 PDU最大的序号, 緩存等待最大值由高层配置。 实际窗口为 ( VR(UDR), VR(UDH)], 其中 VR(UDR)为下一个按序送到高层的 PDU 的序号, 即小于这个 SN 的 PDU都已经被按序送到高层。 VR(UDR)>= VR(UDH)-DAR_Window_Size„ 收到实际窗口内的 PDU进行緩存, 收到 序号大于窗口上界的 PDU, 进行窗口的更新, 也是由窗口上界移动带动 窗口更新。
当序号为 VR(UDR)的 PDU被接收后, 检查窗口内现在最小的没有 被正确接收的 PDU的序号为多少,将 VR(UDR)更新到此值,并将所有序 号小于更新后 VR ( UDR ) 的 PDU送到高层重组。 如果当窗口前移导致 VR ( UDR ) <VR(UDH)-DAR— Window— Size+1, 更新 VR(UDR)为更新窗 口( VR(UDH)- DAR_Window_Size, VR(UDH)]内最小的没有收到的 PDU, 将 VR ( UDR )更新到此 SN, 并将所有序号小于更新后 VR ( UDR ) 的 PDU送到高层重组。
计时器 Timer— DAR和状态变量 VR(UDT)控制接收窗长时间不移动。 VR ( UDT )初始设为窗口内最大序号, 同时启动计时器 Timer— DAR。 当 序号为 VR(UDT)的 PDU在计时器未超时被送到高层重组, 计时器重启, VR ( UDT )重置为窗口内最大序号。 如果定时器超时, 所有序号小于等 于 VR ( UDT )和与 VR ( UDT )连续的 PDU都被送入高层, VR ( UDR ) 更新为现在緩存中最小没有收到的序号, 同时重置 VR ( UDT )为窗口内 PDU最大序号, 重启定时器。
DAR重排和前两者不同在于, 这个功能单元只重排不做重组, 重组 功能在高层做。
4. HSPA重排
HSPA是高速分组接入的缩写, HSDPA (高速下行分组接入的缩写) 的重排机制和 DAR机制相同。 其目的是为了 HARQ (混合自动重传请求 的缩写)不同 process重传的乱序, 根据 TSN序号进行重排保证收到的 MAC-hs PDU按序进行恢复成 MAC-d PDU送到 MAC-d实体。
随着 HSDPA、 Enhanced Uplink等增强技术的引入, 而提出的长期演 进( LTE, Long Term Evolution )即' 'Evolved UTRA and UTRAN,,即 3GPP 无线接入技术的长期演进需要考虑降低的时延、 更高的用户数据速率、 改进的***容量和覆盖和对运营商而言降低的成本。为此,长期演进 LTE 网络的各种性能的需求, 网络结构、 无线接口、 协议栈功能都会发生相 应改进。 现有协议层结构造成很多重复的功能, 比如 RLC和 MAC子层 的重传、 分割级联等等。 为了减小延迟和简化协议, 这些重复的功能必 须考虑精简。 另外, LTE***提出的全 IP要求, 即网络传输全部是基于 IP包业务, 这一全新的需求也需要响应的机制去保证。
在众多网络改进方案中, RLC中很多功能被认为是重复冗余的。 目 前比较流行的观点认为 RLC中的很多功能在 MAC中都可以实现, 因此, RLC将可以被合并到 MAC实体中去。 这里我们也认为 RLC的重复功能 将带来不必要的延迟和复杂度, 简化合并到 MAC 中去是比较合理的方 案。
最近提出的 LTE***中 MAC结构如图 5和图 6所示。对现有技术而言, 上文已经讨论的 WCDMA R6***中共有以上 4个三种重排机制, 可以适 合不同的重排需求, 但结构复杂, 功能冗余, 降低了响应速度, 无法应 用于已有 LTE网絡, 例如, 将自动重传请求 ARQ和混合自动重传请求 HARQ分开处理, 效率低下, 增加了 ARQ重传的延迟。
由此可见, 在现有协议中 MAC层和 RLC层都需要进行数据的重排。 即: 在 RLC层, RLC SDU按照高层配置的 RLC PDU大小来进行数据分割, 对分割得到的 PDU统一编号, 用作接收和重传的标识, 需重传的数据存 于 ARQ緩存中,在目前协议中 RLC PDU的编号称为序列号( SN, Sequence Number ) , 确认模式的序列号长度为 12bit, 非确认模式的序列号为 7bit。 在目前协议中一种业务在配置其对应的 RLC实体时就给定了 RLC PDU的 大小, 在进行 RLC SDU分割时, 在数据不足 PDU大小时对 RLC PDU添加 填充 bit, 保证 PDU大小恒定。 在接收端 RLC层根据 SN对 RLC PDU进行重 排并重组。 再如对于 HSDPA/HSUPA***中, RLC PDU经过 MAC-d实体 添加 MAC数据头后形成 MAC-d PDU之后, 在 MAC- hs/MAC-e实体进行级 联, 添加数据头, 其中包括传输序列号(TSN ) , 形成物理帧, 并緩存于 HARQ緩存中, 在接收端 MAC层 HARQ根据 TSN对 MAC PDU进行重排, 并把有序接收正确的数据送往 RLC层。
因此, 在数据传输过程中, 当经过接收端不同的协议层时, 比如经 过接收端的 MAC层和 RLC层, 分别根据添加序列号对接收端的数据进 行重排, 从而增加了***的开销, 且 MAC层与 RLC层分别处于基站与 RNC, 通过这两次重排, 会增大数据传输的时延。 发明内容
本发明实施例提供一种无线通信***中数据的重排方法及其装置, 通过减少协议层间数据的重排次数, 简化数据头信息, 来降低数据的传 输时延及减少数据的开销。
本发明实施例提供一种移动通信***中数据的重排方法, 包括步骤: 接收端的媒质接入控制实体接收到正确解码的传输块, 并将所述数 据块送往链路控制实体;
接收端的链路控制实体对接收到的传输块进行解封装, 并将解封装 后的高层协议数据单元进行重排和重组, 得到业务数据单元。 更适宜地, 该方法还包括:
当接收到的传输块具有多个链路控制实体的协议数据单元, 则对所 述传输块进行分解, 并把分解后的子传输块发送到相应的链路控制实体。
本发明实施例提供一种无线通信***中数据的重排装置, 包括: 第 一重传单元和第二重传单元,
所述第一重传单元, 用于将接收到的传输块发送到第二重传单元; 所述第二重传单元, 用于接收第一重传单元发送的传输块, 并对该传输 块进行解封装, 得到高层协议数据单元, 以及对所述高层协议数据单元 进行重排和重组, 得到业务数据单元, 将其发送高层。
更适宜地, 所述装置还包括:
分解单元, 分别与第一重传单元和第二重传单元相连, 用于分解由 多个链路控制实体的协议数据单元组成的传输块, 并把分解后的子传输 块发送到相应的链路控制实体。
本发明实施例的技术方案中, 通过将无线链路控制层下移至基站或 合并到媒体接入控制层中, 将大大减少两层之间数据的交互时延, 有利 于数据传输效率的提高。 通过将两层的重排简化为一次重排, 不但简化 了协议层将数据层的封装, 取消了 HARQ传输序列号(TSN ), 减少数据 的开销, 同时也降低了在 HARQ时带来的数据重排的时延, 提高了数据 传输效率, 对于那些具有无序递交属性的业务数据单元尤其有利。
在本发明实施例的方案中, 接收端对收到解复用后的数据包根据其 序号 SN进行重排 /重组, 以及设置接收緩存条件, 并根据该接收緩存条 件对经过重排后的数据进行后续处理。 每个接收到的 SDU或 PDU均有 序号 SN, HARQ根据 HARQ PDU包头信息将需要重排的 PDU送到各 ARQ重排重组单元进行重排。
通过本发明实施例的技术方案, 可大大减少层与层之间的数据交互 的时延, 甚至可以达到忽略时延。 通过简化层间的重排的方式, 使得重 排重传功能可在基站实现, 且可以一次完成, 而无需分成两层单独进行 重排, 可避免 ARQ重传时延的增加, 降低了***复杂性, 提高了速度。 附图说明 图 1是现有技术中无线接口协议的结构框架示意图; 图 2是无线链路控制 RLC实体模型的结构示意图;
图 3是用户设备侧的 MAC结构示意图;
图 4是网絡侧的 MAC结构示意图;
图 5是 LTE网络中 eNB中 UL的 MAC结构示意图;
图 6是 LTE网络中用户设备 DL的 MAC结构示意图;
图 7是演进无线陆地接入***的架构示意图;
图 8是本发明实施例中数据重排装置的示意图;
图 9是本发明实施例中数据重排方法的流程图;
图 10是本发明实施例中数据重排方法的具体实现过程;
图 11是本发明实施例中数据重排方法中协议层的示意图;
图 12 是本发明实施例中数据重排方法中接收端的重排缓存的实施 例。 具体实施方式
为了使本技术领域的人员更好地理解本发明方案, 在描述本发明之 前先简要的介绍一下本发明所基于的演进无线陆地接入 ( E-UTRAN ) 系 统。
参照图 7, 为演进无线陆地接入(E-UTRAN )***的架构示意图。 所 述 E-UTRAN包括: 演进接入网关 1 ( E-AGW )和演进基站 2 ( E-NodeB ) 两层架构,所述 E-NodeB又管理多个用户设备 UE3。所述 E-Node与 E-AGW 可以直接建立连接, 也可以通过 IP网络相连。 其中, 所述 E-AGW位于演 进核心网演进核心网 Evolved CN中, 其功能与 R6协议中的无线网络控制 器 R C的功能类似, 用于管理不同的 E-NodeB, 而各个 E-NodeB直接与 E-AGW相连, 管理所述小区内的用户。 对于 E-UTRAN的协议栈, 所述 E-NodeB包括:演进媒体接入控制层 MAC和演进物理层 PHY,所述 E-AGW 包括演进分组数据会聚协议层。
通过对上面技术的了解, 下面结合附图和实施方式对本发明作进一 步详细的说明。 参照图 8, 为本发明所述移动通信***中数据重排装置的架构示意 图, 所述装置可以位于演进基站中也可以位于用户设备 UE中。所述装置 包括: 第一重传单元 11和第二重传单元 12, 还可以包括分解单元 13 (如 图中虚线所示)。 其中, 所述第一重传单元 11 , 用于将接收到的传输块直 接发送到第二重传单元;所述第二重传单元 12,与第一重传单元 11相连, 通过对接收到的传输块进行解封装, 得到高层协议数据单元, 并对所述 高层协议数据单元进行重排和重组, 得到业务数据单元, 将其发送到高 层。 所述分解单元 13 , 分别与所述第一重传单元 11和所述第二重传单元 12相连, 用于分解由多个链路控制实体的协议数据单元组成的传输块, 并把分解后的子传输块发送到相应的链路控制实体。
所述第一重传单元 11包括: 混合自动重传子单元 111和判断子单元 112, 所述混合自动重传子单元 111用于将接收的传输块进行重传; 所述 判断子单元 112,用于判断所述传输块的重传次数是否超过预设的最大重 传次数, 如果超过最大的重传次数, 则通知第二重传单元该传输块中数 据丟失。
所述第二重传单元 12包括: 至少一个自动请求重发子单元 121、 至 少一个重排子单元 122、 至少一个重组子单元 123。 其中, 所述自动请求 重发子单元 121 , 与混合自动重传子单元 111或分解单元 13相连, 用于 对接收到传输块或子传输块进行解封装并发送解封装后得到的高层协议 数据单元; 所述重排子单元 122, 与自动请求重发子单元 121相连, 根据 序列号或分段序列号对接收到的高层协议数据单元进行重排, 并发送重 排后的高层协议数据单元;所述重组子单元 123 ,与重排子单元 122相连, 用于将接收到的高层协议数据单元去掉信息头后进行重组, 得到业务数 据单元。
由上述装置可知, 针对演进后的移动通信***, 将无线链路控制层 RLC与媒体接入控制层 MAC合并, 从而减少了协议层之间接口的传输 时延, 甚至可以忽略接口时延。 对于需要传输的业务数据单元(即 IP数 据包), 有两种实现方式: 一种实现方式是重排装置位于用户设备中: 另 一种实现方式是重排装置位于演进基站中。 下面分别对这两种实现过程 进行描述:
一种实现方式: 演进基站中发送端媒体接入控制层对其进行分割或 / 和级联, 得到协议数据单元, 并对其封装信息头, 组成传输块(TB )后 经物理层传输到用户设备 UE的接收端 ,所述用户设备接收端的媒体接入 控制层 MAC接收到该传输块后, 通过接收端混合自动重传 (HARQ, Hybrid Automatic Repeat Request )层对所述传输块的重传后得到正确的解 码传输块, 并将其发送给接收端的自动请求重发 ( ARQ, Automatic Repeat Request )层, 所述自动请求重发 ARQ层对接收到的传输块进行解封装后 得到乱序的协议数据单元, 然后在根据序列号或分段序列号对所述协议 数据单元进行重排和重组, 得到有序的业务数据单元, 并将其发送给高 层。
其中, 所述 ARQ是指通过重传来保证传输的可靠性, 当前一次尝试 传输失败时, 就要求重传数据包的传输机制。 HARQ协议是 ARQ协议的 改进, 就是在 ARQ***中引入了一个 FEC子***, 这个 FEC子***是 用来纠正错误以减少重传的次数。 在 HARQ中, 发送端只在发送的数据 块被正确接收到之后才开始对下一个数据块进行操作, ***使用 1 比特 的序列号区分当前的数据块和下一个待传输的数据块, 接收端用于指示 传输块是否被正确解码的确认信息(使用应答 ACK或错误应答 NACK ) 使用一个比特。
另一种实现方式是上述实现方式的逆过程, 其实现过程基本与上述 方式相同, 其不同之处是, 用户设备中发送端设备先对需要传输的业务 数据单元进行分割或 /和级联, 并将分割或 /和级联后的协议数据单元封装 信息头后组成传输块通过物理层发送到演进基站中, 所述演进基站与上 述用户设备接收端的处理过程相同, 具体详见上述, 在此不再赘述。
其中, 所述物理层的功能是对接收到的数据包进行调制、 编码、 时 频同步等操作; MAC层进行媒体接入控制, 动态调度, 处理 UE数据流的 优先级等操作。自动请求重发 ARQ的功能是在混合自动重传 HARQ之上的 重传, 防止 HARQ失败造成数据差错, ARQ还具有将乱序的数据包进行分 段、 重组、 分割级联、 顺序递交到高层、 复制检测、 流控等功能。 另外, 本发明还提供一种通信***中数据的重排方法, 该方法的流 程图详见图 9, 所述方法包括步骤:
步骤 S11:接收端的媒质接入控制实体接收到正确解码的传输块后送 往链路控制实体;
步骤 S12: 接收端的链路控制实体对接收到的传输块进行解封装, 并 将解封装后的高层协议数据单元进行重排和重组, 得到业务数据单元。
在步骤 S11和步骤 S12之间还包括步骤: 如果接收到的传输块由多 个链路控制实体的协议数据单元组成, 则对传输块进行分解, 并把分解 后的子传输块发送到相应的链路控制实体。
为了便于理解, 下面以用户设备为发送端, 而演进基站为接收端为 例来说明本发明, 反之也成立。
再请参阅图 10和图 11 , 分别为本发明所述方法中数据传输具体过程 的实施例。 如图所示, 发送端媒体接入控制 MAC层的分割和 /或級联功能 实体对高层业务数据单元 SDU (即 IP包)进行分割和 /或级联, 并设置相 应的序列号(比如图中所示 1、 2和 3 ) , 組成协议数据单元 PDU后送自动 请求重发 ARQ实体; 而 MAC层的生成传输块功能实体对接收到的 PDU加 上必要的头信息 (TB头) (如业务优先级队列号等)后进行复用, 组成 传输块(TB )后送混合自动重发 HARQ功能实体, 所述接收端的 HARQ 经重传后收到正确解码的 TB, 并向 ARQ递交正确接收的 TB, 不作重排。 所述 ARQ实体通过分解及解封装得到乱序的协议数据单元 PDU后不发状 态报告, 只把所述 PDU放在重排緩存中, 重排功能实体根据序列号 SNA 分段序列号 SSN对 PDU进行重排。 其中, 所述 SN为首次传输的 PDU的序 列号, SSN为各分段在首次传输 PDU中的序列号,这样在接收端可以通过 SN及 SSN就可以进行重传、 重排及重组。 重组功能实体去掉 PDU中的信 息头后, 进行重组, 得到正确且有序业务数据单元 SDU, 并将其送往高 层。此外,接收端数据块的重传次数超过最大重传次数时,在接收端 HARQ 可以通过层间源语告知 ARQ该传输块已丢失, 所述状态报告的触发可以 周期性触发、 由链路控制实体自动产生或由 HARQ的最大重传次数触发。
本发明的又一实施例中, 如果发送端有 2个 UDP (即 IP包) , 经分割 级联后分成 3个 PDU, 序列号(SN )分別为 1 , 2, 3 , 形成 TB后在 HARQ 进行重传, 接收端收到 TB后通过解封装后, 将其送到重排实体, 重排实 体根据序列号或分段序列号进行重排, 然后去掉信息头后对所述协议数 据单元进行重组。 在该实现过程中, 如果传输块的重传次数超过预定的 最大重传次数, HARQ则通知 ARQ进行重传, 再次重传 PDU时, 可以再次 对其进行分割级联,比如 1号 PDU可以再次分成 1.1及 1.2号 PDU,即 SN = 1, SSN - 1及 SN = 1, SSN = 2这2个PDU。这样在接收端的重排緩存中顺序如 图 12所示。
由此可见, 本发明实施例的方案中取消了接收端 HARQ的重排, 即在 接收端 HARQ收到正确的 TB后立即送往 ARQ, 重排只在 ARQ做, 而对于 HARQ传输序列号(TSN )可以取消, 减少数据的开销。 因此, 本发明不 但減少协议层间数据的重排次数, 简化数据头信息, 同时也降低了数据 的传输时延和减少数据的开销。
下面再以 LTE网絡为例讨论本发明数据重排重组方法,主要针对 AM 和 UM两种业务。 对 UM业务只用 HARQ业务, 重排机制只针对 HARQ 乱序。
对于 AM业务, 接收端各 HARQ进程收到相应 TB (传输块) , 如 果校验正确, 送到 MUX单元进行解复用到各 ARQ实体。 ARQ重排功能 单元根据 PDU序号进行 PDU重排并重组 SDU送到高层, 触发重传或删 除相应重传緩存 PDU。对于 UM业务,接收端各 HARQ进程收到相应的 TB, 如果校验正确, 送到 MU 单元进行解复用到各 ARQ实体, 对于 UM, ARQ实体不做重传, 只做分割级联。 重排功能单元针对 HARQ重 传乱序进行排序, 并重组 SDU送到高层。 对于某些 HARQ禁止的实时 业务, 无需重排序。
为实现重排,本发明的方法设置了 HARQ接收緩存和 ARQ接收緩存, 分别用于緩存等待 HARQ和 ARQ重传的数据。由于采用一个缓存来緩存 等待 HARQ和 ARQ重传带来的乱序,本发明提出了重排机制, 包括窗口 控制和定时器控制机制,用于区分 HARQ重传乱序等待和 ARQ重传乱序 等待。 下面讨论緩存及重排机制。 1、 HARQ緩存
一些参数和状态变量定义如下:
Highest— received— SN表示缓存内收到的 PDU最高序号;
Next一 expected— SN表示緩存内下一个按序送到高层或下一个功能单 元的 PDU序号;
HARQ— RcvWindow— Size表示高层配置的合适 QoS的 HARQ緩存等 待最大值;
将 HARQ 緩 存 窗 口 定 义 为 ( Highest— received— SN - HARQ— Window— Size, Highest_received_SN ) 。
2、 ARQ緩存窗口
ARQ_RcvWindow_Size是高层配置的 ARQ緩存等待最大值。
ARQ緩存窗口的定义包括:
若 Next—expected— SN< Highest_received_SN - HARQ— Window— Size , 定义 ARQ緩存窗口为:
(Next— expected— SN, Highest— received—SN - HARQ_Window_Size); 若 Highest— received— SN - HARQ— Window— Size - Next— expected— SN
>ARQ— RcvWindow— Size , 定义 ARQ 緩存窗口为(Next— expected—SN,
Next—expected— SN + ARQ_Rcv Window— Size);
3、 ARQ緩存与 HARQ緩存的关系
3-1 ) 合并检测: 如果 Next— expected— SN< Highest—received—SN - HARQ— Window— Size , HARQ 和 ARQ 緩存 重合成 一 个 。 Highest—received—SN - HARQ— Window— Size - Next—expected— SN <ARQ_Rcv Window— Size。
3-2 ) 溢出检测: 如果 Highest—received—SN - HARQ_Window_Size - Next—expected— SN >ARQ— RcvWindow— Size , 发生窗口溢出, 接收端发送 窗口溢出指示给发送端, 发送端相应 ARQ实体新数据传输将会被挂起即 只能进行重传不能进行新数据传输。
3-3 )脱节检测: 如果 Highest—received—SN - HARQ— Window— Size - Next— expected— SN <ARQ— RcvWindow— Size这时会出现 HARQ緩存窗口 和 ARQ缓存窗口脱节, 原来两个窗口是重叠的, 可以由接收端发送挂起 结束指示给发送端, 也可以由发送端根据重传緩存的状态进行配置。
4、 HARQ接收窗口参考边界
本发明实施例中还可设定 HARQ接收窗口参考边界作为重排机制。 即定义: HARQ—RcvWindow—Edge为实际的 HARQ緩存和 ARQ緩存的 边界, 数值上对应于下一个 HARQ按序收到的 PDU序号。
HARQ—RcvWindow— Edge <= Highest— received—SN -
HARQ_Window— Size。 当窗口上界即 Highest— received— SN更新形成的窗 口下界更新, 如果 SN为 Highest— received— SN - HARQ— Window— Size的 PDU在緩存中已经存在,将窗口实际下界 HARQ— RcvWindow— Edge更新 到现在窗口内最小的没有收到的 PDU序号。
ARQ窗口则变为 ( Next— expected_SN, HARQ_RcvWindow_Edge ) , 其他 ARQ相关操作同上述重排机制。
定义定时器 Timer_HARQ ( T_SN ) ,用于控制 HARQ窗口移动 , T—SN 初始设为緩存内最大序号 Highest— received—SN , 同时启动各计时器 Timer_HARQ。 当序号为 T— SN 的 PDU 在计时器未超时即 T— SN< =HARQ— Rev Window— Edge收到时, 计时器重启, T—SN重置为窗口内最 大序 号 。 如 果定 时 器超时 , T— SN>HARQ—RcvWindow_Edge HARQ—Rcv Window— Edge被更新到 (Highest— received— SN, T—SN)中没有收 到的 PDU最小序号, 同时重置 T—SN为窗口内 PDU最大序号, 重启定时 器。
5、 采用定时器作为重排机制
每次从 MUX单元收到一个 ARQ PDU就启动定时器 Timer— Rev, 状 态变量 HARQ—RcvWindow— Edge— T表示 HARQ接收窗口边界,即为定时 器已经超时的接收到 PDU最大的序号。 Highest_received—SN表示緩存内 收到的 PDU最高序号。 Next— expected— SN表示緩存内下一个按序送到高 层或下一个功能单元的 PDU 序号。 Next—expected— SN 可以小于 HARQ一 RcvWindow— Edge—T。
ARQ—RcvWindow— Size是高层配置的 ARQ缓存等待最大值(窗口大 小)上限。 ARQ接收緩存仍用窗口控制,接收窗口为( Next— expected— SN, HARQ— RcvWindow— Edge— T ) 一般緩存等待最大值小 于等于 ARQ— RcvWindow— Size , 如 果 HARQ_RcvWindow— Edge— Τ - Next— expected— SN > ARQ— RcvWindow— Size, ARQ 窗 口 定 为 ( Next— expected—SN, Next— expected— SN + ARQ— Rev Windo w— Size )接收 端发送窗口溢出指示给发送端,发送端相应 ARQ实体新数据传输将会被 挂起即只能进行重传不能进行新数据传输。
HARQ和 ARQ緩存等待最大值(即窗口大小)根据业务可以由高层 进行配置。
实施例一
当收到 SN = X的 PDU, 进行以下处理(步骤号可以表示执行顺序, 也可以不是执行顺序) :
步據 1 口果 X在 ( Next— expected— SN, Highest— received— SN )之内成 立, 如果此 SN对应数据已经在緩存中存在则删除此数据; 如果 X不在 ( Next— expected— SN, Highest— received— SN )之内, 才艮据序号将 PDU放 入緩存;
步骤 2: 如果 x> Highest— received—成立, 如果是
x-HARQ— RcvWindow— Size-Next— expected— SN 大 于
ARQ— Rev Window— Size, 贝1 J {
发送新数据挂起消 息给发送端 , 更新 ARQ 窗 口 为 ( Next— expected— SN, Next— expected—SN + ARQ— Rev Window— Size ) ; 删 除 SN 从 Next— expected— SN + ARQ— RcvWindow— Size 到 Highest—received— SN- HARQ_Rcv Window— Size;
根据状态报告触发原则触发状态报告; 更新 Highest— received— SN为 ; }
否则 { 检查从 Highest— received— SN-HARQ_RcvWindow— Size 到 x-HARQ— RcvWindow— Size中的 missing PDU;
根据状态报告触发原则触发状态报告; 更新 Highest—received— SN为 } ; 步骤 3 : 如果 x = Next— expected— SN, 则{
检查是否有与 Next_expected—SN序号连续的 PDU;
将 Next_expected—SN更新为从 x开始第一个没有收到的 PDU序号 }。 实施例二
当收到 SN = x的 PDU, 进行以下处理(步骤号可以表示执行顺序, 也可以不是执行顺序) :
步骤 1 : 如果 X在( Next—expected— SN, Highest— received— SN )之内 { 如果此 SN对应数据已经在緩存中存在则删除此数据, 否则,根据序 号将 PDU放入緩存
步骤 2: 如果 x> Highest— received— SN, 贝' J {
更新 Highest— received一 SN为 x;
检查 x-HARQ— RcvWindow— Edge是否大于 HARQ— RcvWindow— Size, 如果是则更新 HARQ— RcvWindow— Edge为 x - HARQ_RcvWindow_Size; } 步驟 3: 如果 SN等于 HARQ— RcvWindow— Edge的 PDU在缓存中, 则{
检查是否有和这个 PDU 序号连续的 PDU 在缓存中, 将 HARQ— RcvWindow— Edge重新更新到最小的不连续的 PDU序号; }
步骤 4: 才艮据更新的 HARQ— RcvWindow— Edge和状态报告触发原则 触发相应状态报告;
步骤 5: 检查如果 HARQ— RcvWindow— Edge-Next— expected— SN > ARQ— RcvWindow— Size, 则{发送新数据挂起消息给发送端, 更新 ARQ 窗口为( Next— expected— SN, Next— expected— SN + ARQ_RcvWindow_Size ); 删 除 SN 从 Next— expected— SN + ARQ_RcvWindow_Size 到 HARQ_RcvWindow_Edge;
根据状态报告触发原则触发状态报告。 }
步骤 6: 如果收到 x = Next— expected— SN, 则{
检查是否有与 Next— expected— SN序号连续的 PDU;
将 Next_expected— SN更新为从 X开始第一个没有收到的 PDU序号; } 在此同时, 有以下定时器操作 (步骤号可以表示执行顺序, 也可以 不是执行顺序) :
步骤 1: 如果緩存中有 PDU, Timer_HARQ没有启动, 则{ 触发定时器 Timer—HARQ;
T_SN设置为緩存中最大的 PDU序号。 }
步骤 2: 如果计时器超时之前, HARQ_RcvWindow— Edge > =T—SN, 则停止计时器;
步骤 3: 如果计时器超时, HARQ_RcvWindow— Edge < T—SN, 则{ HARQ_RcvWindow_Edge更新到 T— SN+1;
检查 T_SN+1 是否在緩存中, 有和序号连续的 PDU在緩存中, 将 HARQ_RcvWindow_Edge重新更新到最小的不连续的 PDU序号。 }
步骤 4: 检查, 如果 HARQ_RcvWindow— Edge-Next— expected— SN > ARQ_RcvWindow_Size , 则{ 发送新数据挂起消息给发送端, 更新 ARQ 窗口为 ( Next— expected— SN, Next— expected— SN + ARQ_Rcv Window— Size ); 删 除 SN 从 Next— expected— SN + ARQ— Rev Window— Size 到 HARQ_RcvWindow_Edge; 根据状态报告触发原则触发状态报告。 } 实施例三
当收到 SN = x的 PDU, 则进行以下步驟, 步骤编号可以表示执行顺 序, 也可以不是执行顺序。
1、 触发定时器 Timer— Rev;
2、 如果定时器超时, 定时器对应的序号为 y, 则{
如果没有激活状态变量 HARQ_RcvWindow— Edge— T 则设 HARQ_RcvWindow_Edge_T = y;
否 则 如 果 HARQ— RcvWindow— Edge— T < y 则 更 新 HARQ_Rcv Window— Edge— T = y; }
3、 如果收到 PDU序号 y = Next— expected— SN, 贝 'J {
检查是否有与 Next—expected— SN序号连续的 PDU;
将 Next—expected— SN更新为从 x开始第一个没有收到的 PDU序号; 停止所有 SN < Next— expected—SN对应的定时器; }
4、 HARQ— RcvWindow— Edge— T更新,根据状态报告触发机制触发相 应状态报告;
5、 检查如果 HARQ_RcvWindow_Edge_T - Next_expected_SN > ARQ_RcvWindow_Size , 则{发送新数据挂起消息给发送端, 更新 ARQ 窗口为( Next— expected— SN, Next_expected_SN + ARQ_Rcv Window— Size ); 删除 SN从 Next— expected— SN + ARQ_RcvWindow— Size到
HARQ— RcvWindow—Edge—T的 PDU; 根据状态报告触发原则触发状态报 告。 }
此外, 如果接收到的传输块由多个链路控制实体的协议数据单元组 成, 则先对接收到正确的传输块进行分解, 然后把分解后的子传输块发 送到相应的链路控制实体, 所述各个链路控制实体的实现过程与上述接 收到单一链路控制实体的协议数据单元的实现过程相同 , 具体详见前述, 在此不再赘述。
在本发明实施例的方案中, 接收端对收到解复用后的数据包根据其 序号 SN进行重排 /重组, 以及设置接收緩存条件, 并根据该接收緩存条 件对经过重排后的数据进行后续处理。 每个接收到的 SDU或 PDU均有 序号 SN, 这里用于重排重组的 SN是 ARQ唯一的 SN, 即可以是重用高 层的 SDU SN, 也可以是发送端 ARQ加的 ARQ SN。 如果存在重传分割 级联后的分段子序号, 需要将每个 SN对应的多块片端都收齐。 ARQ单 元就是根据其序号 SN进行重排 /重组; 重排是对在接收与发送之间的乱 序而进行的重新排序, 重组是按 PDU重组 SDU。 重组包括乱序递交的重 组和按序递交的重组。 HARQ根据 HARQ PDU包头信息将需要重排的 PDU送到各 ARQ重排重组单元进行重排。
通过本发明实施例的技术方案, 可大大减少层与层之间的数据交互 的时延, 甚至可以达到忽略时延。 通过简化层间的重排的方式, 使得重 排重传功能可在基站实现, 且可以一次完成, 而无需分成两层单独进行 重排, 可避免 ARQ重传时延的增加, 降低了***复杂性, 提高了速度。
以上所述仅是本发明的优选实施方式, 应当指出, 对于本技术领域 的普通技术人员来说, 在不脱离本发明原理的前提下, 还可以做出若干 改进和润饰, 这些改进和润饰也应视为本发明的保护范围。

Claims

- 19- 权 利 要 求
1、 一种无线通信***中数据的重排方法, 其特征在于, 包括步骤: 接收端的媒质接入控制实体接收到正确解码的传输块, 并将所述数 据块送往链路控制实体;
接收端的链路控制实体对接收到的传输块进行解封装, 并将解封装 后的高层协议数据单元进行重排和重组, 得到业务数据单元。 '
2、 根据权利要求 1所述的方法, 其特征在于, 还包括:
当接收到的传输块具有多个链路控制实体的协议数据单元, 则对所 述传输块进行分解, 并把分解后的子传输块发送到相应的链路控制实体。
3、 根据权利要求 1或 2所述的方法, 其特征在于, 所述媒质接入控 制实体收到乱序的正确解码的传输块不作重排, 直接将其送往上层。
4、 根据权利要求 1所述的方法, 其特征在于, 当传输块的重传次数 超过预设的最大重传次数, 则媒质接入控制实体通过层间源语告知链路 控制实体该传输块中数据已丟失。
5、 根据权利要求 1或 2所述的方法, 其特征在于, 所述链路控制实体 得到高层协议数据单元后, 将其存放在重排緩存中, 按照预先制定好的 重排机制, 根据 PDU序号进行 PDU重排;
基于所述重排后的 PDU, 根据预先制定好的重组策略对业务数据单 元 SDU进行重组并将重组后的 SDU送到高层。
6、 根据权利要求 5所述方法, 其特征在于, 状态报告的产生由周期 触发、 媒质接入控制实体中由最大重传次数引起的触发或者由自动重传 实体自动产生。
7、 根据权利要求 1或 2所述的方法, 其特征在于, 所述对解封装后 的高层协议数据单元进行重排和重组, 得到业务数据单元的具体实现过 程包括:
接收端的重排实体^^据序列号及分段序列号对接收到的高层协议数 据单元进行重排, 并去掉重排后高层协议数据单元的信息头后进行重组, 得到业务数据单元。
8、 根据权利要求 1或 2所述的方法, 其特征在于, 所述媒质接入控 制实体接收到正确解码的传输块的过程包括对混合自动重传 HARQ进程 接收的数据块进行校验的步骤。
9. 根据权利要求 4所述的方法, 其特征在于, 所述预先制定的重组策 略包括, 设置按序递交 SDU和乱序递交 SDU, 如果配置为按序递交 SDU, SDU必须按照 PDU序号标识的先后顺序进行 SDU的恢复和向高层递交; 如果配置为乱序递交 SDU, SDU可以按照任意顺序进行递交。
10. 根据权利要求 5所述的方法, 其特征在于, 所述预先制定的重排 机制通过预先定义的滑动窗口控制 HARQ緩存等待范围和 ARQ緩存等待 范围, 对接收到的新协议数据进行处理, 并检测需要重组的协议数据包 或 /和协议数据单元丢失, 进行数据包的重组或 /和生成反馈。
11. 根据权利要求 10所述的方法, 其特征在于: 所述的预先定义的滑 动窗口控制的 HARQ缓存等待范围是由高层配置的固定窗口大小参数表 示的, 其起始范围分别为緩存内收到的 PDU最大的序号减去由高层配置 的固定窗口大小参数以及緩存内收到的 PDU最大的序号。
12. 根据权利要求 11所述的方法, 其特征在于, 所述的预先定义的滑 动窗口控制的 ARQ緩存等待范围由 HARQ緩存等待下限, ARQ按序接收 PDU序号以及 ARQ缓存等待最大值共同决定。
13. 根据权利要求 12所述的方法, 其特征在于, 当 ARQ按序接收的下 一个 PDU序号到 HARQ允许緩存等待的最小值小于等于 ARQ緩存等待最 大值, ARQ緩存等待范围从 ARQ按序接收的下一个 PDU序号到 HARQ允 许緩存等待的最小值;
当 ARQ按序接收的下一个 PDU序号到 HARQ允许緩存等待的最小值 大于 ARQ緩存等待最大值, ARQ緩存等待范围从 ARQ按序接收的下一个 PDU序号到这个序号加上 ARQ缓存等待的最大值。
14.根据权利要求 13所述的方法, 其特征在于, 所述的 ARQ緩存等待 最大值是由高层配置的。
15.根据权利要求 11所述的方法, 其特征在于, 所述的緩存内收到的 PDU最大的序号, HARQ緩存等待下界, ARQ按序接收的下一个 PDU序 号是由状态变量表示的。
16.根据权利要求 11所述的方法, 其特征在于, 所述 ARQ按序接收的 下一个 PDU是指 ARQ緩存下一个按序希望接收的 PDU, 所有序号小于此 PDU的协议数据单元都已经移出重排緩存。
17.根据权利要求 10所述的方法, 其特征在于, 所述对接收到的新协 议数据进行处理方法包括:
若所述新协议数据单元序号大于 HARQ緩存等待范围, 对此数据进 行緩存并更新最高接收序号的状态变量,根据窗口机制进行其他状态变 量更新, 进行溢出检测;
若所述新协议数据单元序号位于 HARQ緩存等待范围内且没有重复 序号数据单元, 緩存此 PDU;
若所述新协议数据单元序号位于 ARQ緩存等待范围内且没有重复 序号数据单元, 緩存此 PDU, 并根据接收反馈策略进行接收反馈;
若所述新协议数据单元序号等于 ARQ按序接收的下一个 PDU序号, 进行重组 SDU的检测。
18.根据权利要求 10所述的方法, 其特征在于, 所述的检测需要重組 的协议数据包的步骤包括:
如果配置为按序递交 SDU,如果接收到序号等于 A Q按序接收的下 一个 PDU序号,检查从这个 PDU开始的连续 PDU直到出现不连续的协议 数据单元, 将这些连续 PDU中含有的 SDU重组出来并送到高层, 并更新 相应按序接收 PDU的状态变量;
如果配置为无序递交 SDU,将接收到的协议数据单元中指示完整的 SDU发送到高层。
19.根据权利要求 10所述的方法, 其特征在于, 所述的检测丢失协议 数据包的步骤包括:
在 ARQ緩存等待范围内, 如果存在序号大于 SN的协议数据单元已 经被收到, 则判断序号为 SN的协议数据单元为丢失数据包。
20、 根据权利要求 10所述的方法, 其特征在于, 所述的预先定义的 滑动窗口控制的 HARQ緩存等待范围根据接收协议数据单元情况、 高层 配置的 HARQ緩存范围的最大值和预先设定的记时器确定; 该范围为从 緩存内接收协议数据单元最高的序号, 到下一个 HARQ按序收到的协议 数据单元序号。
21、 根据权利要求 10所述的方法, 其特征在于, 所述的预先定义的 滑动窗口控制的 HARQ緩存等待范围由接收协议数据单元情况和定时器 控制确定; 该范围的上界为收到协议数据单元最大的序号, 下界为超时 定时器所对应的 PDU中最大的序号。
22、 一种无线通信***中的数据重排装置, 其特征在于, 包括: 第 一重传单元和第二重传单元, 其中,
所述第一重传单元, 用于将接收到的传输块发送到第二重传单元; 所述第二重传单元, 用于接收第一重传单元发送的传输块, 并对该 传输块进行解封装, 得到高层协议数据单元, 以及对所述高层协议数据 单元进行重排和重组, 得到业务数据单元, 将其发送高层。
23、 根据权利要求 22所述的装置, 其特征在于, 所述装置还包括: 分解单元, 分别与第一重传单元和第二重传单元相连, 用于分解由 多个链路控制实体的协议数据单元组成的传输块, 并把分解后的子传输 块发送到相应的链路控制实体。
24、 根据权利要求 22或 23所述的装置, 其特征在于, 所述第一重 传单元包括:
混合自动重传子单元, 用于将接收的传输块进行重传;
判断子单元, 用于判断所述传输块的重传次数是否超过预设的最大 重传次数, 如果超过最大的重传次数, 则通知第二重传单元该传输块中 数据丢失。
25、 根据权利要求 22或 23所述的装置, 其特征在于, 所述第二重 传单元包括:
至少一个自动请求重发子单元, 与混合自动重传子单元或分解单元 相连, 用于对接收到传输块或子传输块进行解封装并发送解封装后得到 的高层协议数据单元;
至少一个重排子单元, 与自动请求重发子单元相连, 根据序列号或 分段序列号对接收到的高层协议数据单元进行重排 , 并发送重排后的高 层协议数据单元;
至少一个重组子单元, 与重排子单元相连, 用于将接收到的高层协 议数据单元去掉信息头后进行重组, 得到业务数据单元。
26、 4艮据权利要求 22或 23所述的装置, 其特征在于, 所述重排装 置位于演进基站或用户设备中。
PCT/CN2007/000406 2006-03-03 2007-02-06 Procédé de réassemblage de données dans un système de communication sans fil et appareil associé WO2007098676A1 (fr)

Priority Applications (4)

Application Number Priority Date Filing Date Title
EP07702289A EP1993241B1 (en) 2006-03-03 2007-02-06 A method for reassembling data in wireless communication system and an apparatus thereof
DE602007011221T DE602007011221D1 (de) 2006-03-03 2007-02-06 Verfahren zur neuzusammenstellung von daten in einem drahtlosen kommunikationssystem und vorrichtung dafür
AT07702289T ATE492100T1 (de) 2006-03-03 2007-02-06 Verfahren zur neuzusammenstellung von daten in einem drahtlosen kommunikationssystem und vorrichtung dafür
US12/203,799 US20090046626A1 (en) 2006-03-03 2008-09-03 Method and device for reordering data in wireless communication system

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
CN200610059342.6 2006-03-03
CN2006100593426A CN101030842B (zh) 2006-03-03 2006-03-03 移动通信***中数据的重排方法及其装置
CN200610034635.9 2006-03-22
CN2006100346359A CN101043301B (zh) 2006-03-22 2006-03-22 一种无线通信***中的数据重排重组方法及其基站

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US12/203,799 Continuation US20090046626A1 (en) 2006-03-03 2008-09-03 Method and device for reordering data in wireless communication system

Publications (1)

Publication Number Publication Date
WO2007098676A1 true WO2007098676A1 (fr) 2007-09-07

Family

ID=38458653

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2007/000406 WO2007098676A1 (fr) 2006-03-03 2007-02-06 Procédé de réassemblage de données dans un système de communication sans fil et appareil associé

Country Status (5)

Country Link
US (1) US20090046626A1 (zh)
EP (1) EP1993241B1 (zh)
AT (1) ATE492100T1 (zh)
DE (1) DE602007011221D1 (zh)
WO (1) WO2007098676A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2232732A4 (en) * 2007-12-18 2014-10-22 Korea Electronics Telecomm MULTI-FA PERSONAL SUBSCRIBER END DEVICE AND METHOD FOR ORDERING A LOG DATA UNIT THEREOF

Families Citing this family (31)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TWI364964B (en) * 2007-02-14 2012-05-21 Innovative Sonic Ltd Method and apparatus for enhancing transmission efficiency in a wireless communications system
MX2009009505A (es) 2007-03-07 2010-01-15 Interdigital Tech Corp Metodo y aparato para generar y procesar una undad de datos de protocolo mac-ehs.
US8351380B2 (en) * 2007-05-09 2013-01-08 Samsung Electronics Co., Ltd Method and apparatus for layer 2 ARQ for packets
US7949012B2 (en) * 2007-08-01 2011-05-24 Broadcom Corporation High-speed uplink packet access (HSUPA) cipher multiplexing engine
JP4447028B2 (ja) * 2007-08-10 2010-04-07 富士通株式会社 通信制御方法、送信装置、およびコンピュータプログラム
US8576767B2 (en) * 2008-05-19 2013-11-05 Hughes Network Systems, Llc Method and system for providing a satellite interface to support mobile communication services
EP2136501B1 (en) * 2008-06-20 2019-12-04 LG Electronics Inc. Method of delivering a PDCP data unit to an upper layer
FI20085924A0 (fi) * 2008-09-30 2008-09-30 Nokia Corp Palvelun tietoyksiköiden uudelleenkokoaminen viestintäjärjestelmässä
KR100972166B1 (ko) 2008-10-13 2010-07-26 한국전자통신연구원 무선통신 시스템에서의 프로토콜데이터유닛 순서정렬 방법 및 이를 수행하는 단말
US8638699B2 (en) * 2008-11-10 2014-01-28 Qualcomm Incorporated Method and apparatus for supporting the large service data unit (SDU)
KR20100070113A (ko) * 2008-12-17 2010-06-25 한국전자통신연구원 센서 네트워크에서의 대용량 데이터 전송을 위한 매체접근제어 방법
US8711881B2 (en) * 2009-01-07 2014-04-29 Qualcomm Incorporated Packet bundling at the PDCP layer
US8644338B2 (en) * 2009-01-07 2014-02-04 Qualcomm Incorporated Unbundling packets received in wireless communications
JP5052549B2 (ja) * 2009-03-03 2012-10-17 株式会社東芝 無線通信装置及び無線通信方法
KR101232599B1 (ko) 2009-12-15 2013-02-12 한국전자통신연구원 매체 접근 제어 프로토콜 데이터 유닛 재조립 방법 및 이를 수행하는 수신기
CN102209311B (zh) * 2010-03-29 2013-11-06 电信科学技术研究院 优先级队列的维护方法和设备
JP5643434B2 (ja) * 2010-10-01 2014-12-17 インターデイジタル パテント ホールディングス インコーポレイテッド 複数の送信点からの受信を可能にするmacおよびrlcアーキテクチャならびに手順
US8792601B2 (en) 2011-10-04 2014-07-29 Qualcomm Incorporated Non-coherent combining detection with reduced buffering requirements
CN108847919B (zh) 2013-04-26 2021-09-21 华为技术有限公司 一种数据传输的方法、基站和无线通信设备
WO2015066923A1 (zh) * 2013-11-11 2015-05-14 华为技术有限公司 数据传输方法及装置
CN104753629B (zh) * 2013-12-26 2018-03-30 联芯科技有限公司 一种数据递交方法和装置
KR20150090804A (ko) * 2014-01-29 2015-08-06 삼성전자주식회사 이동 통신 시스템에서 복수의 캐리어를 이용하여 데이터를 송수신하는 방법 및 장치
WO2015168951A1 (zh) * 2014-05-09 2015-11-12 华为技术有限公司 数据包的处理方法、终端、基站及***
KR102195069B1 (ko) 2014-08-06 2020-12-24 삼성전자주식회사 패킷 송신기 및 이를 포함하는 인터페이스 장치
KR102298991B1 (ko) * 2015-05-22 2021-09-07 삼성전자 주식회사 무선 통신 시스템에서 버퍼 관리 방법 및 장치
DE102015116221B4 (de) * 2015-09-25 2021-05-12 Apple Inc. Kommunikationsendgerät und Verfahren zum Übertragen einer Signalisierungsnachricht
US11201780B2 (en) * 2016-01-29 2021-12-14 Qualcomm Incorporated Configurations associated with segmentation of one or more packets for wireless communication
WO2018189882A1 (ja) * 2017-04-14 2018-10-18 富士通株式会社 無線通信装置、無線通信方法、及び無線通信システム
KR102614500B1 (ko) * 2017-04-25 2023-12-18 엘지전자 주식회사 데이터 유닛을 전송하는 방법 및 장치
CN117295112A (zh) 2017-08-10 2023-12-26 三星电子株式会社 无线通信***中数据处理的方法和装置
EP3918739A4 (en) 2019-02-02 2022-04-27 ZTE Corporation METHOD AND APPARATUS FOR PERFORMING DATA PACKET RETRANSMISSION IN A WIRELESS COMMUNICATION SYSTEM

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1379557A (zh) * 2002-05-16 2002-11-13 武汉汉网高技术有限公司 多拒绝自动请求重传arq机制
WO2002096006A2 (en) 2001-05-18 2002-11-28 Nokia Corporation Hybrid automatic repeat request (harq) scheme with in-sequence delivery of packets
CN1505416A (zh) * 2002-11-29 2004-06-16 华为技术有限公司 第三代无线通信***中的***信息接收重组方法
CN1716835A (zh) * 2004-07-01 2006-01-04 日本电气株式会社 移动无线通信终端设备

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR100776084B1 (ko) * 2001-09-17 2007-11-15 엘지노텔 주식회사 티씨피/아이피 프로토콜 계층 2에서 리라이어블 서비스제공장치 및 그 방법
KR100747464B1 (ko) * 2002-01-05 2007-08-09 엘지전자 주식회사 고속하향링크패킷접속(hsdpa)시스템을 위한타이머를 이용한 교착상황 회피방법
US6901063B2 (en) * 2002-05-13 2005-05-31 Qualcomm, Incorporated Data delivery in conjunction with a hybrid automatic retransmission mechanism in CDMA communication systems
DE10252536A1 (de) * 2002-11-08 2004-05-27 Philips Intellectual Property & Standards Gmbh Verfahren und Vorrichtung zur Übertragung von Datenpaketen
US20050185609A1 (en) * 2004-02-16 2005-08-25 Esa Malkamaki Communication method, user terminal, network element and computer program
KR100921241B1 (ko) * 2004-06-16 2009-10-12 엘지전자 주식회사 통신 시스템의 데이터 유닛 처리 시스템
JP4474217B2 (ja) * 2004-07-09 2010-06-02 富士通株式会社 パケット通信装置

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2002096006A2 (en) 2001-05-18 2002-11-28 Nokia Corporation Hybrid automatic repeat request (harq) scheme with in-sequence delivery of packets
CN1379557A (zh) * 2002-05-16 2002-11-13 武汉汉网高技术有限公司 多拒绝自动请求重传arq机制
CN1505416A (zh) * 2002-11-29 2004-06-16 华为技术有限公司 第三代无线通信***中的***信息接收重组方法
CN1716835A (zh) * 2004-07-01 2006-01-04 日本电气株式会社 移动无线通信终端设备

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2232732A4 (en) * 2007-12-18 2014-10-22 Korea Electronics Telecomm MULTI-FA PERSONAL SUBSCRIBER END DEVICE AND METHOD FOR ORDERING A LOG DATA UNIT THEREOF
US8891530B2 (en) 2007-12-18 2014-11-18 Samsung Electronics Co., Ltd. Multi-FA personal subscriber terminal and method of ordering protocol data unit thereof

Also Published As

Publication number Publication date
US20090046626A1 (en) 2009-02-19
DE602007011221D1 (de) 2011-01-27
EP1993241A1 (en) 2008-11-19
EP1993241A4 (en) 2009-06-24
EP1993241B1 (en) 2010-12-15
ATE492100T1 (de) 2011-01-15

Similar Documents

Publication Publication Date Title
WO2007098676A1 (fr) Procédé de réassemblage de données dans un système de communication sans fil et appareil associé
JP4981904B2 (ja) 媒体アクセス制御破棄通知
KR101084135B1 (ko) 무선 통신 시스템의 송수신 단에서의 상태 pdu송수신방법
EP2290866B1 (en) Method for moving a receive window in a radio access network
JP4188774B2 (ja) フレーム送受信システム、フレーム送信装置、フレーム受信装置、及びフレーム送受信方法
JP5342012B2 (ja) ハンドオーバ性能を改良する装置および方法
US8285330B2 (en) HARQ reordering method for WCDMA enhanced uplink dedicated channel
EP2469750A1 (en) Method and apparatus for downlink data transmission control in multi-hop relay communication system
ES2357045T3 (es) Método y dispositivo para reensamblaje de datos en un sistema de comunicación inalámbrica.
CN101151872B (zh) 通过信道接收数据单元的接收装置和方法
JP2006523042A (ja) 再送信の方法とシステム
JP5081900B2 (ja) 再送要求送信方法及び受信側装置
US8300583B2 (en) Method for transmitting control information in a mobile communication system
KR20100053625A (ko) 무선 통신 시스템에서의 핸드오버 동안 데이터의 계층 2 터널링
KR20070077798A (ko) 이동통신 시스템에서 재전송 제어를 위한 상태보고의요청/전송 방법 및 장치
JP2004530364A (ja) 再シーケンシングエンティティと再送エンティティ間のインターレイヤー・コントロールのための方法及びシステム
KR101379408B1 (ko) 재송요구 송신방법 및 수신측장치
KR20090031239A (ko) 성공적으로 수신했으나 헤더 압축 복원에 실패한 패킷의 처리 방법
WO2011047621A1 (zh) 基于重传机制的对头压缩数据包进行传输的方法和装置
WO2013139249A1 (zh) 数据包发送方法、模式转换方法及装置
WO2007022694A1 (fr) Pile de protocoles utilisateur et procede de transfert sans perte
WO2008007176A1 (fr) Procédé et dispositif d&#39;émission-réception de données en mode d&#39;accès par paquets sur une liaison descendante à haut débit
KR101084136B1 (ko) 무선 통신 시스템의 송수신 단에서 상태정보를 포함하는pdu를 송수신하는 방법
WO2009152726A1 (zh) 一种无线链路控制重传处理方法和装置
WO2008133577A1 (en) Method for selectively discarding data units in a radio communication system

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application
NENP Non-entry into the national phase

Ref country code: DE

WWE Wipo information: entry into national phase

Ref document number: 2007702289

Country of ref document: EP