WO2022176172A1 - タイムスロット化した受信を使用する光ネットワークおよびノード - Google Patents
タイムスロット化した受信を使用する光ネットワークおよびノード Download PDFInfo
- Publication number
- WO2022176172A1 WO2022176172A1 PCT/JP2021/006454 JP2021006454W WO2022176172A1 WO 2022176172 A1 WO2022176172 A1 WO 2022176172A1 JP 2021006454 W JP2021006454 W JP 2021006454W WO 2022176172 A1 WO2022176172 A1 WO 2022176172A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- nodes
- switch
- node
- bandwidth
- network
- Prior art date
Links
- 230000003287 optical effect Effects 0.000 title claims abstract description 115
- 238000012545 processing Methods 0.000 claims abstract description 10
- 238000003860 storage Methods 0.000 claims abstract description 10
- 238000000034 method Methods 0.000 claims description 9
- 230000008569 process Effects 0.000 claims description 6
- 230000005540 biological transmission Effects 0.000 abstract description 22
- 230000009286 beneficial effect Effects 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 30
- 238000004891 communication Methods 0.000 description 10
- 238000005516 engineering process Methods 0.000 description 6
- 230000007246 mechanism Effects 0.000 description 6
- 238000013459 approach Methods 0.000 description 5
- 230000007423 decrease Effects 0.000 description 5
- 230000002093 peripheral effect Effects 0.000 description 5
- 238000006243 chemical reaction Methods 0.000 description 3
- 238000001816 cooling Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 230000009467 reduction Effects 0.000 description 3
- 230000015556 catabolic process Effects 0.000 description 2
- 238000006731 degradation reaction Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 2
- 235000008694 Humulus lupulus Nutrition 0.000 description 1
- 241000283973 Oryctolagus cuniculus Species 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 230000000593 degrading effect Effects 0.000 description 1
- 230000001934 delay Effects 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 239000004744 fabric Substances 0.000 description 1
- 230000017525 heat dissipation Effects 0.000 description 1
- 230000020169 heat generation Effects 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 229910044991 metal oxide Inorganic materials 0.000 description 1
- 150000004706 metal oxides Chemical class 0.000 description 1
- 238000003825 pressing Methods 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04J—MULTIPLEX COMMUNICATION
- H04J14/00—Optical multiplex systems
- H04J14/02—Wavelength-division multiplex systems
- H04J14/0278—WDM optical network architectures
- H04J14/02862—WDM data centre network [DCN] architectures
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04Q—SELECTING
- H04Q11/00—Selecting arrangements for multiplex systems
- H04Q11/04—Selecting arrangements for multiplex systems for time-division multiplexing
- H04Q11/08—Time only switching
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B10/00—Transmission systems employing electromagnetic waves other than radio-waves, e.g. infrared, visible or ultraviolet light, or employing corpuscular radiation, e.g. quantum communication
- H04B10/27—Arrangements for networking
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04Q—SELECTING
- H04Q11/00—Selecting arrangements for multiplex systems
- H04Q11/0001—Selecting arrangements for multiplex systems using optical switching
- H04Q11/0005—Switch and router aspects
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04Q—SELECTING
- H04Q11/00—Selecting arrangements for multiplex systems
- H04Q11/04—Selecting arrangements for multiplex systems for time-division multiplexing
- H04Q11/0428—Integrated services digital network, i.e. systems for transmission of different types of digitised signals, e.g. speech, data, telecentral, television signals
- H04Q11/0435—Details
- H04Q11/0442—Exchange access circuits
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04Q—SELECTING
- H04Q11/00—Selecting arrangements for multiplex systems
- H04Q11/0001—Selecting arrangements for multiplex systems using optical switching
- H04Q11/0005—Switch and router aspects
- H04Q2011/0007—Construction
- H04Q2011/0033—Construction using time division switching
Definitions
- the present invention relates to optical networks, optical nodes and optical transmission systems used therein.
- a data center is a general term for facilities that specialize in the installation and operation of computers (mainframes, minicomputers, servers, etc.) and data communication equipment.
- Data center (DC) networks have grown rapidly to provide a wide range of services while accommodating a significant increase in traffic volume. Despite this significant increase in traffic volume, the basic structure of networks has not changed significantly.
- FIG. 1 is a diagram showing the configuration of a conventional DC network.
- a typical DC network consists of multiple layers, for example, it has a three-layer DC network configuration as shown in FIG.
- a three-layer DC network 100 includes a switch 101, an optical link 102, a switch 103, an optical link 104, a ToR (TopOfRack) switch 105, and a server 106 from the upper layer side.
- the switches 101 and 103 are each composed of upper and lower electrical switches and optical links 107 and 108 connecting them. Therefore, the DC network of FIG. 1 would have 8 hops between the two ToRs.
- Optical-Electrical-Optical (OEO) conversion is performed at a large extent in each electrical switch.
- data is transmitted as optical signals on links connecting network nodes consisting of one or more servers.
- the optical signal is converted into an electrical signal and switched by an electrical switch, ie, an ASIC (Application Specific Integrated Circuit) switch.
- ASIC Application Specific Integrated Circuit
- the switching capacity of ASIC switches continues to increase significantly and is now reported with a capacity of 12.8 Tb/s.
- CMOS Complementary Metal Oxide Semiconductor
- a more pressing problem is the requirement for very high data rates for optical signals entering and exiting ASIC switches.
- co-packaged implementations of transceivers and ASIC switches have received increasing attention in recent years as an alternative to traditional pluggable transceivers.
- significant power consumption and heat generation problems remain.
- the present invention has been made in view of such problems, and its purpose is to propose a new DC network structure that addresses various limitations of ASIC switches.
- one embodiment is an optical core section having a full mesh network configuration, and a plurality of nodes connected to the optical core section, which are divided into a plurality of groups. , a plurality of nodes including up to m nodes in a group, each of said plurality of nodes receiving optical signals from said at most m nodes; and routing to a plurality of servers, the ASIC switch having a switching capacity corresponding to the average incoming traffic of the plurality of nodes, the receiving cycle comprising a plurality of time slots.
- idle time slots addressed by any node in said group to which said source node belongs and following said plurality of time slots, only in time slots associated with the group to which said source node belongs, for a period of time; It is characterized by not receiving optical signals from any of said source nodes and processing traffic in excess of said average incoming traffic.
- An optical network of another embodiment includes an optical core section having a full mesh network configuration, and a plurality of nodes connected to the optical core section, divided into a plurality of groups, and a maximum of and a plurality of nodes, including m nodes, each of said plurality of nodes being associated with a group to which a source node belongs during a receive cycle including a plurality of time slots. switching and routing electrical signals corresponding to optical signals received from the up to m nodes addressed by any node in the group to which the source node belongs, only in the time slots selected by the source node, and routed to a plurality of servers.
- An ASIC switch a main switch having a switching capacity corresponding to the average incoming traffic of the plurality of nodes and operating synchronously with the time slots, and a switching capacity capable of processing traffic exceeding the average incoming traffic. and an auxiliary switch that operates regardless of the time slot.
- optical network described above also has aspects as an invention of network nodes within the optical network.
- the optical network of the present disclosure simplifies the node configuration and reduces the capacity and power consumption of ASIC switches. It also supports large-scale optical networks and low power consumption.
- FIG. 1 illustrates a basic configuration of a DC network according to the present disclosure
- FIG. 1 is a diagram illustrating a problem of a node in a conventional optical network periphery
- FIG. 1 is a diagram conceptually showing the configuration of an optical network and a receiving node of the present disclosure
- FIG. 4 is a diagram illustrating the effect of reducing receivers obtained in the optical network of the present disclosure
- FIG. 2 is a diagram for explaining data reception operation in a receiving node of an optical network
- FIG. 10 is a diagram showing another example of the receiving operation of networks with different group configurations
- 1 illustrates exemplary network and node specifications according to this disclosure
- FIG. 3 illustrates different concepts of receive bandwidth required at a receiving node
- FIG. 4 is a diagram for explaining possible combinations of a plurality of nodes with different average numbers of connections
- FIG. 4 is a diagram for explaining effective BW degradation between node pairs due to network division
- Fig. 3 illustrates different bandwidths of traffic in a DC network
- FIG. 2 illustrates the reduced bandwidth concept of a receiving ASIC switch
- FIG. 10 is a diagram showing the relationship between average bandwidth BW switch_avg and reception cycle period T
- FIG. 10 is a diagram illustrating an extended reception cycle period associated with the introduction of factor F
- FIG. 4 is a block diagram of a node including modified switches with the introduction of factor F
- Fig. 2 is a block diagram of a node that mitigates delay in a modified switch;
- the following disclosure includes a DC network architecture that achieves end-to-end optical transmission between desired node pairs with optical switching at the core and electrical switching only at the network perimeter.
- the inventors have proposed a DC network structure that more efficiently utilizes electrical switches facing performance limitations.
- BW capacity
- FIG. 2 is a diagram showing the basic configuration of a DC network according to the present disclosure.
- the DC network 1 includes a flat optical network 2 that is the core part of the entire network, and a SW unit 3, ToR 4, and server 5 that are the peripheral parts of the entire network. These peripheral elements form part of the node.
- Two nodes are shown in FIG. 2, with one node including four ToRs for simplicity, but it should be understood that many other nodes are located around the flat optical network 2 .
- optical switching is used in the flat optical network 2, which is the core portion, and electrical switching is used only in the SW portion 3, which is the peripheral portion, and its peripheral side.
- the optical network structure of FIG. 2 is scalable and supports highly dynamic connections between arbitrary pairs of nodes.
- the core part of the network can be implemented as a physical full-mesh network or as a full-mesh-like network.
- the full-mesh network or full-mesh-like network premised in the DC network of the present disclosure shown in FIG. are different.
- only optical switching is used without performing OEO conversion in the SWs 101 and 103 of each hierarchy of FIG.
- the following disclosure reveals a novel and practical data reception mechanism and its associated hardware at each node. They replace some of the previously required numerous optical receiver units and complex bulk switching arrangements.
- the basic approach of configuring the optical network of the present disclosure is to introduce a small time-domain constraint on transmissions from network nodes to the same destination node.
- the network of the present disclosure operates according to a timeslot scheme [Problem of reception in optical network periphery] When realizing a physical full-mesh connection in the conventional N-node DC network shown in FIG. Requires enormous resources. Such full-mesh or full-mesh-like networks require extensive switching fabric at each node.
- any destination node in the network can be addressed simultaneously by any number of source nodes, up to N-1.
- "addressing" is not limited to specifying and designating the destination node of the other party as a communication destination by a source node in order to set up a communication link between nodes, but also actually setting up a communication link, It shall also include conducting communications.
- FIG. 3 is a diagram for explaining problems in conventional optical network peripheral nodes.
- FIG. 3 conceptually shows a full-mesh optical network 140, representing the flat optical network 1 shown in FIG. 2 as a set of nodes.
- the optical network 140 has many interconnected nodes.
- FIG. 3 depicts four adjacent nodes interconnected to one node, this symbolically indicates that all nodes in the optical network are connected. there is
- the receiving node 141 has an interface section 144 with the optical network side and is further connected to the ToR 145 .
- the interface section 144 comprises a receiver 143 for receiving multiple optical signals 142 arriving simultaneously. If the optical network 140 consists of N nodes, then the receiving node 141 would require a number of ports and corresponding (N-1) receivers 143 to handle all simultaneously incoming data communications. Become. Further, the interface unit 144 has a large-scale receiving ASIC switch (not shown). As mentioned earlier, very high data rates are required for optical signals entering and exiting ASIC switches. Also from FIG. 3, it can be understood that a large number of receivers and large-scale ASIC switches are required at all receiving nodes.
- the optical network of the present disclosure proposes to introduce a slight time-domain restriction on data transmission from a network node to the same destination node.
- FIG. 4 is a diagram conceptually illustrating the basic configuration of the optical network and receiving nodes of the present disclosure. Similar to FIG. 3, FIG. 4 conceptually illustrates a full-mesh or full-mesh-like optical network 10.
- the optical network 10 consists of N (eg, 36) nodes, and as an example consider data transmission between a source node or source node 11 and a destination node or receiving node 13 .
- N eg, 36
- the source node can also be the receiving node, and the names are simply distinguished according to the operations and functions to be described. Therefore, it should be noted that the source node also has the functions and configurations described for the receiving node.
- N network nodes that are source nodes are divided into a large number of groups (d), and groups to which each source node belongs are defined.
- groups to which each source node belongs are defined.
- six groups, a first group (G1) to a sixth group (G6), are defined, and the source node 11 belongs to the first group 12.
- Reception of data communication at any node is performed separately in time slots of fixed duration T for each group to which the source node belongs, as will be described later. That is, a receiving node can be addressed by any source node belonging to a group of source nodes only during the time slots assigned to that group. Specifically, the receiving node 13 receives data from the source node 11 during the time slot assigned to G1 to which the source node 11 belongs. It is also addressed during the same timeslot by the other five source nodes belonging to G1.
- the nodes belonging to one group in FIG. 4 are shown here as being close together for the sake of explanation, but they are generally spatially distributed throughout the network.
- a receiving node sequentially receives data from different groups for each time slot, and the cycle time for completing reception from all groups is d (number of divisions) ⁇ T (TS time).
- d number of divisions
- T TS time
- B out be the maximum transmission bandwidth (BW) of each node.
- BW transmission bandwidth
- bandwidth means the transmission bandwidth that can be received or transmitted by a node, and may be understood as transmission speed (transmission rate). It should be noted that the term “bandwidth” is a broad concept meaning the capacity of communication resources for data transmission determined according to the modulation scheme and signal configuration of optical communication.
- the bandwidth that can be sent from a node and the bandwidth that can be received at a node are the same, and the maximum outgoing bandwidth, B_out , is also the maximum incoming bandwidth.
- the effective BW which is the effective bandwidth between two nodes, is B out /d.
- increasing the number d of node groups in the network will decrease the effective BW between any pair of nodes.
- a new mechanism for increasing the value of this effective BW is described below as it addresses the performance limitations of electrical switches in the optical network and receiving nodes of the present disclosure of FIG.
- the nodes in the optical network of the present disclosure of FIG. Compared to nodes, the number of receivers can be greatly reduced.
- an arrayed waveguide grating (AWG: Arrayed Waveguide Grating) 14 has been added.
- Each of the AWGs 14 has at least d groups of input ports so that data from N nodes can be received simultaneously, and the front part of the node 13 is provided with AWGs of m nodes in one group. .
- the received data multiplexed through the m AWGs 14 is supplied to the interface section 15 similar to the conventional technology and input to the receiver 16 .
- the number of receivers 16 in the interface section 15 is significantly reduced to 1/m compared to the prior art configuration of FIG.
- FIG. 5 is a diagram showing the effect of reducing receivers obtained by dividing the nodes in the optical network of the present disclosure into groups.
- the relationship between the number of receivers (Rx units) in a node and the number of all nodes is shown using the number of divisions, that is, the number of groups, as a parameter.
- (b) shows linear representation. Regardless of the number of nodes, the number of required receivers can be greatly reduced as the number of divisions increases.
- the optical network of the present disclosure includes an optical core section 2 having a full-mesh network configuration, and a plurality of nodes connected to the optical core section, divided into a plurality of groups. a plurality of nodes, each of which includes m nodes, each of said plurality of nodes 13 being in the group to which said source node belongs only in the time slots associated with the group to which said source node belongs; It can be implemented as being addressed by any node. Further, each of the plurality of nodes is m arrayed waveguide gratings (AWG) 14, and the plurality of input ports of the AWG are the corresponding optical signals from the maximum m nodes belonging to the same group.
- AWG arrayed waveguide gratings
- the wavelengths used by one or more source nodes among the maximum m nodes are set to match the operating wavelengths of the plurality of input ports, and an output combination of the AWG. It can be implemented as having m receivers 16 connected to wave ports and an ASIC switch 17 for switching and routing electrical signals from the m receivers to a plurality of servers 18 .
- a node connected to an optical core section having a full mesh network configuration the node is divided into a plurality of groups together with a plurality of other nodes connected to the core section, and within one group contains up to m nodes, each of said up to m nodes being addressed by any node in said group to which said source node belongs only in the time slot associated with the group to which said source node belongs.
- a node connected to an optical core section having a full mesh network configuration the node is divided into a plurality of groups together with a plurality of other nodes connected to the core section, and within one group contains up to m nodes, each of said up to m nodes being addressed by any node in said group to which said source node belongs only in the time slot associated with the group to which said source node belongs.
- the interface section 15 of the receiving node 13 has a receiving ASIC switch 17 as in the configuration of the prior art.
- the configuration of the interface section at the receiving node is greatly simplified compared to the prior art in terms of the number of receivers and the capacity of the receiving ASIC switch. Further details of the operation of the receiving node and reducing the bandwidth of the receiving ASIC switch are provided.
- B. Simplification of reception switching Using the maximum transmission bandwidth of one node as described above, B out , and the number of nodes in one group, m, the maximum BW that a node in the basic configuration of FIG.
- FIG. 6 is a diagram for explaining the data receiving operation of the receiving node of the basic configuration in the optical network of the present disclosure.
- FIG. 6(a) shows a time slot configuration
- FIG. 6(b) is a diagram for explaining the switching operation when data is being received in TS1.
- the ASIC switch 17 of the interface unit 15 of (b) has m ports on the input side facing the network and the number of ports corresponding to the number of the ToR switches 18 on the other output side. . If multiple links are used for each ToR switch, there are ports corresponding to the multiple links.
- the time slot structure of FIG. 6(a) shows an example of dividing all the nodes of the optical network 10 shown in FIG. 4 into six groups. Receives only in slots. For example, data from 6 source nodes belonging to the first group are received at the corresponding ports of each AWG only in TS1 as shown in (b). Data from m different source nodes are received in m AWGs 14-1 to 14-6, the same number as the number of nodes in one group. Thus, in a particular time slot, each of the m input ports of ASIC switch 17 can capture data received from different source nodes within the same group. That is, data can be received simultaneously from all m source nodes in the same group.
- an AWG is connected to each port on the input side of the ASIC switch 17, and nodes of the same group are connected to separate AWGs. It is connected. During the assigned time slot, only nodes in the permitted group are active while all nodes in other groups do not transmit to the destination node.
- the same wavelength ⁇ 1 is used by all nodes of a group (eg, G1) active in TS1.
- An input signal of wavelength ⁇ 1 will be passively routed towards the ASIC switch 17 by multiple (m) AWGs. That is, routing is performed by connecting different nodes within the same group to the ports of the corresponding wavelengths of the m AWGs.
- G2 next active group
- the ⁇ 2 input signal from each node belonging to G2 is passively routed towards the ASIC switch 17 by multiple (m) AWGs.
- nodes in the same group do not need to transmit on the same wavelength.
- Multiple wavelengths used by different groups of nodes connected to the same AWG may be coordinated to different wavelengths so as not to cause conflicts.
- the receiving nodes can share the AWG configuration.
- FIG. 7 is a diagram showing another operation example of reception switching in networks with different group configurations.
- the number of groups (divided number) d 16
- node numbers (1 to 512) are shown at the inputs of the AWG.
- 32 nodes in the same group are shown using the same wavelength, but the wavelength may be different for each AWG.
- FIG. 8 illustrates exemplary network and node specifications according to this disclosure.
- An example of parameters for a configuration with a network divided into 16 and corresponding 16 timeslots is shown.
- one time slot has a length of 40 ns, which is 30 ns corresponding to a 400 Gb/s packet and a margin of 10 ns.
- One cycle is divided into 16 time slots and the 512 network nodes are divided into 16 groups. The number of nodes in one group is 32.
- the destination (receiving) node for each timeslot is only addressed by the authorized group.
- the same source node can address the same destination node every 16 timeslots.
- C. Reduction of BW of Receiving ASIC Switch As mentioned in the explanation of FIG.
- the node of the basic configuration in the optical network of the present disclosure is configured to greatly reduce the capacity of the receiving ASIC switch compared to the conventional technology.
- the receiving ASIC switch capacity is set to a value that matches or slightly exceeds the average incoming traffic.
- a memory associated with the receiving ASIC switch is provided to handle traffic in excess of this switching capacity with some delay.
- FIGS. 9 and 10 it will be described below that the basic node of the flat optical network of the present disclosure can handle substantially sufficient traffic even with reduced receiving ASIC switch capacity compared to prior art nodes. do.
- the inventors classify the traffic input to the receiving node, in other words, the input bandwidth BW of the receiving node, into two categories with different requirements. did.
- Two categories of traffic are: (a) traffic that needs to be switched in real-time without additional delay; All traffic that does not need to be switched. Assume that the average received BW is assigned to the traffic of (a), and the maximum received BW is assigned to the traffic of (b).
- the switching capacity of the receiving ASIC switch ie the transmission bandwidth of the receiving ASIC switch, can be allocated to different BWs corresponding to the two categories of traffic.
- introducing a storage medium (memory) for partial storage of received data at the receiving node avoids the need for received data to be lost or retransmitted. Guaranteed not to.
- the main advantage of introducing a storage medium is that the switching capacity of the receiving ASIC switch can be reduced at the expense of introducing extra delay.
- FIG. 9 is a diagram illustrating different concepts of reception bandwidth required at a receiving node.
- FIG. 9 is a diagram illustrating different concepts of reception bandwidth required at a receiving node.
- 9(b) is a diagram for explaining the two reception bandwidths mentioned above.
- the reception ASIC switch in the interface unit is equipped with a memory to provide an average reception BW for 8 nodes, which means that real-time switching processing for 8 nodes is sufficient.
- the average received BW can vary from node to node, but except for a limited number of specific nodes, the average received BW can usually be set significantly lower than the maximum received BW. This is due to the natural balance of traffic in the flat optical network 2 assumed in the DC network of the present disclosure shown in FIG. This balancing mechanism can be explained as follows for the nature of traffic in flat optical networks.
- FIG. 10 is a diagram explaining possible combinations of multiple nodes with different average numbers of connections.
- the properties of traffic that are prerequisites for the flat optical network will be described.
- the average number of incoming connections per node is 1. From this state, the average number of connections received begins to deviate from 1 as the node receives more connections on average. In flat optical networks this means that traffic that was being received by one node is directed to the new busy node. As more nodes become busy, the average number of connections on the remaining nodes naturally decreases.
- the value of the average number of connections Q of busy nodes of category (a) is used as a parameter, and the node ratio (horizontal axis) of category (b) that satisfies the condition of 100% traffic load and the ratio (horizontal axis) of category (a). This indicates the relationship with the node ratio (vertical axis).
- the receiving node of the basic configuration in the optical network of the present disclosure includes multiple AWGs 14 and interface units 15 that passively route received data from different source nodes.
- a receiver and a reception ASIC switch 17 are provided on the front of the interface section 15 , and a storage means (memory) 19 interlocking with the reception ASIC switch 17 is provided.
- the capacity of the receiving ASIC switch 17 should match or slightly exceed the average received traffic, and the storage means 19 can be used to handle traffic exceeding this switching capacity with a small delay. .
- the optical network of the basic configuration of the present disclosure operates in a time-slotted manner for transmission from a source node to the same destination node in order to limit data reception at the receiving node.
- the source node must wait until one full receive cycle has passed before it can again address the same destination node. Restrictions on data reception will result in a lower effective transmission rate, degrading the effective BW of data transmission between node pairs.
- FIG. 11 is a diagram for explaining degradation of effective BW between arbitrary node pairs due to network division.
- the horizontal axis indicates the network division number (group number) d
- the vertical axis indicates the reception BW normalized by Bout of the maximum transmission BW at the reception node.
- the reception BW decreases as the network division number d increases. For example, when the division number d increases from 8 to 32, the reception BW decreases from 12.5% to 3.12%.
- a mechanism is needed to compensate for this reduction in effective BW between node pairs caused by network partitioning.
- the inventors have proposed two solutions: modifying the operation of the source node and modifying the transmission path between arbitrary nodes.
- One solution is to allow the source node to reach the desired destination node in one or more time slots, and another solution is to set up another optical link between any pair of nodes. and increased the effective BW.
- this specification proposes to modify the configuration of the time slots or the configuration of the electrical switches at the receiving nodes in the optical network with the basic configuration described above. do. From a new perspective different from the previous solutions, we propose a method to deal with performance limitations in ASIC switches. [Introduction of factor F for maximum incoming bandwidth BW in_max of receiving node]
- bandwidth BW switch_avg of the receiving ASIC switch in the network node is a fraction F of the maximum receiving bandwidth BW in_max (0 ⁇ F ⁇ 1), if maximum traffic handling can be maintained with some compromise. It is possible to In order to cope with rapidly increasing traffic volumes and the limitations of switching technology as first mentioned, it will be necessary to reduce the switching capacity (bandwidth) in ASIC switches to a realistically available value. First, different concepts of bandwidth in each part of the DC network of the basic configuration shown in FIG. 4 will be explained.
- FIG. 12 is a diagram illustrating different bandwidths of traffic in the DC network.
- FIG. 12 shows the same basic configuration of the DC network of the present disclosure as shown in FIG. Here we present various definitions of bandwidth from different perspectives in relation to network nodes.
- BW out Total output bandwidth from each node
- BW in_max Maximum incoming bandwidth of each node (to receiving switch)
- the DC network of the present disclosure may comprise storage means 19 associated with the receiving ASIC switch 17, as already explained in FIGS.
- the average bandwidth BW switch_avg for the receiving ASIC switch is represented by arrow 25 .
- the inventors believe that it is difficult to flexibly cope with a sudden increase in traffic and the restrictive situation of ASIC technology progress by allocating the maximum available bandwidth to the receiving ASIC switch at any given time.
- the receiving switch bandwidth BW switch_avg preconfigured to match or slightly exceed the average received traffic, but also a fraction F of the maximum receiving bandwidth BW in_max (Fraction: 0 ⁇ F ⁇ 1)
- F the maximum receiving bandwidth BW in_max
- BW switch_avg Reduced bandwidth of the main ASIC switch at the node (Definition 5)
- Factor F Ratio of the reduced bandwidth BW switch_avg to the ASIC switch to the maximum incoming bandwidth BW in_max to the node Above From the definition of "reduced bandwidth" in (5), the following equation is obtained.
- BW switch_avg F x BW in_max formula (2)
- the ASIC switch has difficulty meeting the bandwidth required for the maximum received bandwidth BW in_max , it is reasonable to introduce a “reduced bandwidth” It can also be said.
- By introducing the factor F it is possible to flexibly cope with the bandwidth of the ASIC switch that is practically available.
- the total bandwidth of the network BW network can be increased by the same factor of F.
- limiting the bandwidth of the receiving ASIC switch generally allows the power consumption of the node to be reduced. Introducing a factor F in the bandwidth of the receiving ASIC switch in this manner is beneficial for network scalability.
- FIG. 13 is a diagram illustrating the concept of reduced bandwidth of ASIC switches in the DC network of this embodiment.
- FIG. 13(a) simply shows the relationship between the average bandwidth BW switch_avg and the maximum incoming bandwidth BW in_max at the receiving node.
- the average bandwidth BW switch_avg 25 of the receiving ASIC switch 17 is set to a fraction of the maximum incoming bandwidth BW in_max 24 to the node, and the maximum incoming bandwidth BW in_max 24 is is bisected.
- the factor F represents the ratio between the maximum incoming bandwidth BW in_max 24 and the average bandwidth BW switch_avg 25 allocated to the ASIC switch and is in the range 0 ⁇ F ⁇ 1.
- the reception cycle in which reception from all node groups is completed is T, and the maximum incoming bandwidth BW in_max of each node is expressed by the following equation from the configuration of the reception cycle T having a time slot structure corresponding to d groups. be able to.
- the reception cycle period T has the following relationship with the number of groups d.
- T t TS ⁇ BW network /BW in_max formula (5)
- T ⁇ BW switch_avg t TS ⁇ F ⁇ BW network formula (6)
- the relationship between the average bandwidth BW switch_avg for the ASIC switch and the total bandwidth BW network of the network in equation (6) via the factor F describes the scalability of the aforementioned network.
- the product of the coefficient F and the BW network (right hand side) is fixed for the receive cycle period T and the given value of BW switch_avg (left hand side).
- the relationship of equation (6) can be maintained by increasing or decreasing the coefficient F corresponding to this fluctuation.
- the coefficient F by varying the set value of the coefficient F, the configuration of the same reception cycle period T and average bandwidth BW switch_avg of the reception switch can be maintained.
- FIG. 14 is a diagram showing the relationship between the average bandwidth BW switch_avg of the receiving ASIC switch and the receiving cycle period T.
- the horizontal axis is the average bandwidth BW switch_avg (Tbps) of the receiving switch
- the vertical axis is the receiving cycle period T (nsec)
- the total network bandwidth BW network is used as a parameter.
- the relationship between is calculated and shown.
- 6 curves are shown with a total output bandwidth BW out per node of 400 Gbps, a time slot time t TS of 40 ns, and a total network bandwidth BW network of 51.2 Tbps to 1.64 Pbps as parameters.
- the plot on the left side corresponds to the current level of ASIC technology
- the plot on the right side corresponds to the level of ASIC technology expected in the future. It can be seen that the reception cycle period T increases as the total network bandwidth BW network increases.
- the receive cycle period T is preferably shorter because it is the period until the next data is received in the DC network timeslot scheme of the present disclosure.
- the graph of FIG. 14 shows the scalable relationship between the ASIC switch average bandwidth BW switch_avg and the network total bandwidth BW network by introducing the factor F in equation (6), including the set value of the reception cycle period T. showing.
- FIG. 13(b) illustrates two concepts of how to handle above-average input traffic in an ASIC switch bandwidth limited by a factor of F.
- FIG. 13(b) shows the configuration of the extended reception cycle period, which adds an idling time slot 26 that does not receive data from the node to the reception cycle period T on the time axis. .
- This idle time slot 26 allows the receiving ASIC switch to switch for an additional portion of the input traffic beyond the temporarily reserved average (average bandwidth BW switch_avg ).
- This idle time slot does not actually receive data from any node, is not addressed, and remains unassigned.
- FIG. 15 is a diagram explaining the configuration of the extended reception cycle period accompanying the introduction of the factor F.
- the timeslot structure 31 for the extended receive cycle is shown.
- the normal time slot structure 30 consists of 16 time slots (TS1 to TS16), one of the 16 groups corresponding to each time slot.
- the reception cycle period T33 is TS time*number of groups (16).
- the extended receive cycle period T EXTEND 35 comprises an idle time 34 following the receive cycle period T33.
- idle time 34 it is neither addressed nor receives data from any network group. Additional (excess) traffic that could not be processed due to above-average input traffic in each time slot can be processed by the receiving ASIC switch during idle time 34 .
- the time in the extended receive cycle period T EXTEND 35 divides the switching process by the ASIC switch into two. One is a process performed within the range of the average bandwidth BW switch_avg for each node group with a time limit during the normal reception cycle period T33. The other is processing that is performed during idle time 34 without time restrictions depending on the node group to which it belongs. This solution does not require any changes in the configuration of the receiving node and the storage means 19 shown in FIG.
- the optical network of the present disclosure includes an optical core section having a full mesh network configuration, and a plurality of nodes connected to the optical core section, divided into a plurality of groups, and a maximum of m nodes in one group. nodes, each of said plurality of nodes switching electrical signals corresponding to optical signals received from said at most m nodes. , an ASIC switch routing to a plurality of servers, the ASIC switch having a switching capacity corresponding to the average incoming traffic of the plurality of nodes, the group to which the source node belongs during a receive cycle period comprising a plurality of time slots.
- any node in the group to which the source node belongs only in the time slots 33 associated with and in one or more idle time slots 34 following the plurality of time slots from any of the source nodes It can also be implemented as not receiving optical signals and handling traffic in excess of said average incoming traffic.
- BW i be the input bandwidth in each time slot of one receive cycle period T, where i is the time slot index.
- the amount of traffic arriving at the ASIC switch varies from time slot to time slot and the difference between BW i and the average bandwidth BW switch_avg is denoted ⁇ BW i .
- the maximum additional traffic that the receiving ASIC switch can handle is limited by the following equation. The following equation averages additional traffic over one receive cycle period T.
- FIG. 13(b) shows the concept of yet another second solution for handling above-average incoming traffic with an ASIC switch bandwidth limited by a factor F.
- the main An auxiliary switch 27 is provided in addition to the ASIC switch. It is not the time-based approach as in the first solution, but the approach of having main and auxiliary switches 27 with different functions and modifying the switching behavior.
- the main switch may be a high speed switch with state of the art performance at the time, while the auxiliary switch 27 may utilize a lower speed switch with less bandwidth than the main switch, as will be described below.
- FIG. 16 is a diagram for explaining the configuration of an ASIC switch modified with the introduction of the coefficient F.
- FIG. FIG. 16 shows a receiving node 40 with ASIC switches in a modified configuration, identical to the receiving node configuration in the basic configuration of the DC network according to the present disclosure shown in FIGS. That is, the AWG 43 of m nodes in one group are provided in the front part of the node, and the corresponding receivers 44 are provided.
- the node 40 has an auxiliary switch 42 in addition to the main switch 41 .
- the main switch 41 like the receiving node in the basic configuration of FIGS. 4 and 9, is a high speed, wideband ASIC switch responsible for processing the average bandwidth BW switch_avg scaled by a factor F.
- Auxiliary switch 42 is a low speed switch with a narrow bandwidth set to 1/q with respect to the bandwidth of main switch 41 .
- the main switch 41 operates under time restrictions according to the time slot method described in FIGS. can.
- the main switch 41 and the auxiliary switch 42 are controlled by a node controller not shown in FIG. 16, and two switches with different bandwidths work together.
- the optical network of the present disclosure includes an optical core section having a full mesh network configuration, and a plurality of nodes connected to the optical core section, divided into a plurality of groups, and a maximum of m nodes in one group. nodes, each of said plurality of nodes being associated with a group to which a source node belongs during a receive cycle period comprising a plurality of time slots An ASIC for switching and routing electrical signals corresponding to optical signals received from said up to m nodes addressed by any node in said group to which said source node belongs, only in time slots, to a plurality of servers.
- a main switch 41 which has a switching capacity corresponding to the average incoming traffic of the plurality of nodes and operates in synchronization with the time slots, and a switching capacity capable of processing traffic exceeding the average incoming traffic. and has an auxiliary switch 42 that operates independently of the time slot.
- the maximum additional traffic that can be handled by the auxiliary switch is determined as follows. Let the input bandwidth in each time slot of one receive cycle period T be BW i . The traffic arriving at the ASIC switch varies from time slot to time slot and the difference between BW i and the average bandwidth BW switch_avg is denoted ⁇ BW i . Let d be the number of time slots in the receiving cycle period T (the number of time slots d also corresponds to the number of groups of nodes). The maximum additional traffic through the ASIC switch in the modified configuration of FIG.
- the causes of latency (delay) encountered by traffic passing through the path of the auxiliary switch can be divided into two.
- the first is latency due to long physical switching times due to the narrow bandwidth of auxiliary switch 42 .
- a second cause is latency due to extra latency due to queuing until the transmission to the top-of-rack (ToR) switch 45 is complete.
- further modifications to the second solution of FIG. 16 are proposed to address these delays due to the auxiliary switch 42 .
- FIG. 17 is a block diagram of a node that mitigates the delay of a modified ASIC switch. Since the receiving node 50 having the modified ASIC switch of FIG. 17 has substantially the same configuration as the receiving node 40 described with reference to FIG. 16, only differences will be described.
- the queue 55 in the auxiliary switch 52 is also included in the receiving node 40 of FIG. 16, and corresponds to the queue in the delay of the second reason mentioned above.
- the auxiliary switch 52 and the ToR switch 57 are connected via a link 56 with a large transmission capacity that can be realized as an optical link.
- the configuration of FIG. 17 can support sufficient capacity to transmit the accumulated data in one time slot to the destination ToR switch, reducing delay in the modified ASIC switch.
- the optical network of the present disclosure can simplify the configuration of nodes in the periphery of the DC network and reduce power consumption. It solves or at least alleviates the problems of ASIC switches, and can cope with the large-scale and low power consumption of optical networks. In addition, it realizes scalability that flexibly adapts to network traffic demands and technological progress of realistic ASIC switches in response to various limitations of ASIC switches.
- the present invention can be generally used in optical communication systems.
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Computing Systems (AREA)
- Physics & Mathematics (AREA)
- Electromagnetism (AREA)
- Optical Communication System (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
Description
White Rabbit project webpage, インターネット<URL: https://www.ohwr.org/project/white-rabbit/wikis/home > K. Clark et al., "Sub-Nanosecond Clock and Data Recovery in an Optically-Switched Data Centre Network,"2018年, post-deadline paper in ECOC 2018, Italy
[光ネットワークのコア部構成]
図2に示した本開示のDCネットワークで前提とするフルメッシュネットワークまたはフルメッシュ様ネットワークは、図1の従来技術のDCネットワーク100と比べて、コア部において光スイッチングのみを利用している点で相違している。図2のフラット光ネットワーク2では、図1の各階層のSW101、103におけるOEO変換を行わずに、光スイッチングのみを利用している。
[光ネットワーク周辺部の受信の問題]
図1に示した従来技術のNノードのDCネットワークで物理的なフルメッシュ接続を実現する場合、送信機、受信機、双方向光リンクの数がN×(N-1)個となるため、膨大なリソースが要求される。このようなフルメッシュネットワークまたはフルメッシュ様ネットワークでは、各ノードにおいて、大規模なスイッチング構成が必要となる。N個のノードを持つフルメッシュまたはフルメッシュ様ネットワークでは、ネットワーク内の任意の宛先ノードは、N-1個まで可能な任意の数のソースノードによって同時にアドレス指定(addressing)することができる。ここで「アドレス指定」とは、ノード間において通信リンクを設定するために、ソースノードが相手方の宛先ノードを通信先として特定し、指定することだけに限らず、実際に通信リンクを設定し、通信を実施することも含むものとする。
[提案するネットワーク方式]
図4は、本開示の光ネットワークおよび受信ノードの基本構成を概念的に示す図である。図3と同様に、図4はフルメッシュまたはフルメッシュ様の光ネットワーク10を概念的に示している。簡単のため光ネットワーク10は、N個(例えば36)のノードからなっており、一例として、送信元ノードすなわちソースノード11および宛先ノードすなわち受信ノート13の間のデータ伝送を考える。当然のことであるが、以後の説明において、ソースノードは受信ノードともなり得るのであって、説明する動作・機能に応じて呼称を区別しているに過ぎない。したがって、受信ノードで説明している機能・構成を、ソースノードも同様に有していることに留意されたい。
B.受信切り替えの簡素化
前述の1つのノードの最大発信帯域幅をBoutおよび1つのグループ内のノード数mを使えば、図4の基本構成のノードが任意のタイムスロットで受信する最大BWは、m×Boutとなる。この最大BWは、タイムスロットを使用しないで完全な非同期動作(例えば光パケットスイッチング)でデータを受信する場合のBWと比較して、約1/dに減少する。図4の受信ノードの構成で示したように、受信ノードへの入力データは、所望のToRスイッチ18へ切り替える必要がある。この切り替えに、受信ASICスイッチ17が使用される。
C.受信ASICスイッチのBWの削減
図4の説明でも言及したように、本開示の光ネットワークにおける基本構成のノードでは、従来技術と比べて、受信ASICスイッチの容量を大幅に抑える構成とした。具体的には、受信ASICスイッチの容量を、平均着信トラフィックに一致するか、または平均着信トラフィックをわずかに超える値に設定する。このスイッチング容量を超えるトラフィックを若干の遅延を伴って処理するために、受信ASICスイッチに関連付けられたメモリを備える。以下、図9および図10とともに、本開示のフラット光ネットワークの基本的なノードにおいて、受信ASICスイッチの容量を従来技術のノードよりも削減しても、実質的に十分なトラフィックを扱えることを説明する。
[受信ノードの最大着信帯域幅BWin_maxに対する係数Fの導入]
(定義2)BWin_max: 各ノードの最大着信帯域幅(受信スイッチへ)
(定義3)BWnetwork:ネットワークの総帯域幅
図12を参照すると、ネットワーク10全体の総帯域幅BWoutは矢印22で表され、個々のノードから総出力帯域幅BWoutは矢印23で表される。ネットワーク10のノード数をNとすれば、次の関係が成り立つ。
BWnetwork =BWout×N 式(1)
図12において、性能限界が問題になっている受信ASICスイッチを含む受信ノードの最大着信帯域幅BWin_maxは、矢印24で表されている。本開示のDCネットワークでは、既に図9および図10においても説明したように、受信ASICスイッチ17と連動する記憶手段19を備えることができる。これによって、受信ASICスイッチ17のスイッチング容量を、平均受信トラフィックに一致するか、または平均受信トラフィックをわずかに超えるものに設定し、記憶手段19を利用して、このスイッチング容量を超えるトラフィックをわずかの遅延を伴って処理できる。図12において、受信ASICスイッチに対する平均帯域幅BWswitch_avgは、矢印25で表わされている。
(定義5)係数F:ノードへの最大着信帯域幅BWin_maxに対する、ASICスイッチへの縮小された帯域幅BWswitch_avgの比
上記の「縮小された帯域幅」の定義(5)より次式が得られる。
BWswitch_avg=F×BWin_max 式(2)
図4に示したDCネットワーク構成において、ASICスイッチが、要求される最大の受信帯域幅BWin_maxに見合った帯域幅を満たすことが難しければ、「縮小された帯域幅」を導入することは合理的とも言える。係数Fを導入することで、現実的に利用可能なASICスイッチの帯域幅に柔軟に対応できる。また、同時にネットワークの総帯域幅BWnetworkを同じ係数Fのファクタで増加させることができる。さらに受信ASICスイッチの帯域幅を制限することで、一般にノードの消費電力を減らすことも可能となる。このように受信ASICスイッチの帯域幅に係数Fを導入することは、ネットワークのスケーラビリティのために有益である。
BWin_max=BWout×m=BWout ×(N/d) 式(3)
式(3)は、式(1)および式(2)を使って変形され、帯域幅の視点から、グループ数dを次式によっても表すことができる。
BWin_max=(BWnetwork /N)×(N/d)
d=BWnetwork /BWin_max 式(4)
図6および図8に示したように、1つのTS時間をtTSとすれば、受信サイクル期間Tはグループ数dと次式の関係を有する。
T=tTS×BWnetwork /BWin_max 式(5)
式(2)~(5)から、さらに次の関係が得られる。
T×BWswitch_avg=tTS×F×BWnetwork 式(6)
式(6)における、ASICスイッチに対する平均帯域幅BWswitch_avgおよびネットワークの総帯域幅BWnetworkの間の、係数Fを介した関係は、前述のネットワークのスケーラビリティを説明している。式(6)では、係数FとBWnetworkの積(右辺)は、受信サイクル期間TおよびBWswitch_avgの所与の値(左辺)に対して固定される。例えば、ネットワークの総帯域幅BWnetworkが変動すれば、この変動に対応して係数Fを増減して設定することで、式(6)の関係は維持される。言い換えると、係数Fの設定値を可変することで、同一の受信サイクル期間Tおよび受信スイッチの平均帯域幅BWswitch_avgの構成を維持できる。
[拡張した受信サイクル期間による平均を超えた入力トラフィックの処理]
受信ノードの最大着信帯域幅BWin_maxに対し係数Fを導入して、ASICスイッチの平均帯域幅BWswitch_avgを「縮小された帯域幅」とする場合、ASICスイッチで平均を超える入力トラフィックを処理する必要がある。平均を超える入力トラフィックを処理するための第1のアイデアはシンプルであり、いずれのノードからも実際にデータを受信することのない1つ以上のタイムスロットを追加することに基づく。この追加のタイムスロットは、アイドルタイムスロットであって、拡張した受信サイクル期間が構成される。
[メインスイッチに補助スイッチを追加した修正された構成]
再び図13の(b)の下側の図を参照すると、係数Fで帯域幅を制限されたASICスイッチで、平均を超える入力トラフィックを処理するもう第2の解決法の概念を示している。第2の解決法では、係数Fを導入して、ASICスイッチの平均帯域幅BWswitch_avgを「縮小された帯域幅」とする場合、ASICスイッチで平均を超える入力トラフィックを処理するために、メインのASICスイッチに加えて補助スイッチ27を備える。第1の解決法のような時間軸へのアプローチではなくて、異なる機能のメインスイッチおよび補助スイッチ27を備え、スイッチングの動作を修正するアプローチである。メインスイッチは、その時々の最高技術水準の性能を持つ高速のスイッチとする一方、補助スイッチ27は、後述するように帯域幅がメインスイッチより狭く、より低速のスイッチを利用することができる。
式(8)から、受信サイクル期間T内の追加のトラフィックΔBWiは、右辺の係数d/qによってむしろ強化され、例えばq=4およびd=16とすると、d/q=4の拡張係数が実現される。アイドル時間の追加によって受信サイクル期間Tを拡張した第1の解決法における式(7)と比較すれば、アイドルタイムスロットを追加することなしに、平均帯域幅を超える入力トラフィックを処理できる。係数Fで「縮小された帯域幅」で動作するASICスイッチの構成をさらに修正することで、第1の解決法での遅延問題を生じさせずに、平均を超える入力トラフィックを処理可能となり、超過帯域幅の量を強化することさえ可能である。
[補助スイッチでスイッチング後のキューイング時間最小化]
Claims (6)
- フルメッシュネットワーク構成を有する光コア部と、
前記光コア部に接続された複数のノードであって、複数のグループに分割されており、1つのグループ内に最大m個のノードが含まれている、複数のノードと
を備えた光ネットワークであって、
前記複数のノードの各々は、
前記最大m個のノードから受信された光信号に対応する電気信号をスイッチングし、複数のサーバーへルーティングするASICスイッチであって、前記複数のノードの平均着信トラフィックに対応するスイッチング容量を有するASICスイッチを有し、
複数のタイムスロットを含む受信サイクル期間において、ソースノードが属するグループに関連付けられたタイムスロットにおいてのみ、前記ソースノードが属する前記グループ内の任意のノードによってアドレス指定され、
前記複数のタイムスロットに引き続く1つ以上のアイドルタイムスロットにおいて、前記ソースノードのいずれからも光信号を受信せず、前記平均着信トラフィックを越えるトラフィックを処理すること
を特徴とする光ネットワーク。 - フルメッシュネットワーク構成を有する光コア部と、
前記光コア部に接続された複数のノードであって、複数のグループに分割されており、1つのグループ内に最大m個のノードが含まれている、複数のノードと
を備えた光ネットワークであって、
前記複数のノードの各々は、
複数のタイムスロットを含む受信サイクル期間において、ソースノードが属するグループに関連付けられたタイムスロットにおいてのみ、前記ソースノードが属する前記グループ内の任意のノードによってアドレス指定され、
前記最大m個のノードから受信された光信号に対応する電気信号をスイッチングし、複数のサーバーへルーティングするASICスイッチであって、
前記複数のノードの平均着信トラフィックに対応するスイッチング容量を有し、前記タイムスロットに同期して動作するメインスイッチ、および、
前記平均着信トラフィックを越えるトラフィックを処理可能なスイッチング容量を有し、前記タイムスロットに関係なく動作する補助スイッチ
を有することを特徴とする光ネットワーク。 - 前記複数のノードの各々は、
前記複数のタイムスロットに引き続く1つ以上のアイドルタイムスロットにおいて、いずれの前記ソースノードからも光信号を受信せず、前記平均着信トラフィックを越えるトラフィックを処理し、
前記補助スイッチは、前記複数のタイムスロットおよび前記1つ以上のアイドルタイムスロットの全期間を通して、前記平均着信トラフィックを越えるトラフィックを処理することを特徴とする請求項2に記載の光ネットワーク。 - 前記複数のノードの各々は、前記補助スイッチと、前記複数のサーバーを含むトップオブラックスイッチとを接続する追加の光リンクをさらに有することを特徴とする請求項2または3に記載の光ネットワーク。
- 前記複数のノードの各々は、
m個のアレイ導波路回折格子(AWG)であって、当該AWGの複数の入力ポートは同一のグループに属する前記最大m個のノードから対応する光信号を受信し、前記複数の入力ポートの動作波長に適合するように、前記最大m個のノードの内の1つ以上のソースノードの使用波長が設定されている、AWGと、
前記AWGの出力合波ポートに接続されたm個の受信機と、
前記スイッチング容量を超えるトラフィックを記憶して処理する記憶媒体と
をさらに有することを特徴とする請求項1乃至4いずれかに記載の光ネットワーク。 - 前記複数のノードの各々への最大着信帯域幅BWin_maxとして、
前記ASICスイッチは、係数F(0<F<1)によって縮小された帯域幅であって、前記平均着信トラフィックに対応する平均帯域幅BWswitch_avg=F×BWin_maxが割り当てられることを特徴とする請求項1乃至5いずれかに記載の光ネットワーク。
Priority Applications (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/JP2021/006454 WO2022176172A1 (ja) | 2021-02-19 | 2021-02-19 | タイムスロット化した受信を使用する光ネットワークおよびノード |
US18/262,046 US20240089641A1 (en) | 2021-02-19 | 2021-02-19 | Optical Network and Node Using Timeslotted Reception |
JP2023500471A JPWO2022176172A1 (ja) | 2021-02-19 | 2021-02-19 |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/JP2021/006454 WO2022176172A1 (ja) | 2021-02-19 | 2021-02-19 | タイムスロット化した受信を使用する光ネットワークおよびノード |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2022176172A1 true WO2022176172A1 (ja) | 2022-08-25 |
Family
ID=82930408
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2021/006454 WO2022176172A1 (ja) | 2021-02-19 | 2021-02-19 | タイムスロット化した受信を使用する光ネットワークおよびノード |
Country Status (3)
Country | Link |
---|---|
US (1) | US20240089641A1 (ja) |
JP (1) | JPWO2022176172A1 (ja) |
WO (1) | WO2022176172A1 (ja) |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20150181317A1 (en) * | 2013-12-24 | 2015-06-25 | Nec Laboratories America, Inc. | Scalable hybrid packet/circuit switching network architecture |
US20150312659A1 (en) * | 2014-04-25 | 2015-10-29 | Futurewei Technologies, Inc. | System and Method for Photonic Switching |
US20160277816A1 (en) * | 2015-03-20 | 2016-09-22 | National Chiao Tung University | Optical data center network system and optical switch |
JP2019216417A (ja) * | 2018-06-11 | 2019-12-19 | 台達電子工業股▲ふん▼有限公司Delta Electronics,Inc. | 知恵定義光トンネルネットワークシステムとネットワークシステム制御方法 |
-
2021
- 2021-02-19 JP JP2023500471A patent/JPWO2022176172A1/ja active Pending
- 2021-02-19 US US18/262,046 patent/US20240089641A1/en active Pending
- 2021-02-19 WO PCT/JP2021/006454 patent/WO2022176172A1/ja active Application Filing
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20150181317A1 (en) * | 2013-12-24 | 2015-06-25 | Nec Laboratories America, Inc. | Scalable hybrid packet/circuit switching network architecture |
US20150312659A1 (en) * | 2014-04-25 | 2015-10-29 | Futurewei Technologies, Inc. | System and Method for Photonic Switching |
US20160277816A1 (en) * | 2015-03-20 | 2016-09-22 | National Chiao Tung University | Optical data center network system and optical switch |
JP2019216417A (ja) * | 2018-06-11 | 2019-12-19 | 台達電子工業股▲ふん▼有限公司Delta Electronics,Inc. | 知恵定義光トンネルネットワークシステムとネットワークシステム制御方法 |
Non-Patent Citations (1)
Title |
---|
CLARK KARI; BALLANI HITESH; BAYVEL POLINA; CLETHEROE DANIEL; GERARD THOMAS; HALLER ISTVAN; JOZWIK KRZYSZTOF; SHI KAI; THOMSEN BENN: "Sub-Nanosecond Clock and Data Recovery in an Optically-Switched Data Centre Network", 2018 EUROPEAN CONFERENCE ON OPTICAL COMMUNICATION (ECOC), IEEE, 23 September 2018 (2018-09-23), pages 1 - 3, XP033447614, DOI: 10.1109/ECOC.2018.8535333 * |
Also Published As
Publication number | Publication date |
---|---|
JPWO2022176172A1 (ja) | 2022-08-25 |
US20240089641A1 (en) | 2024-03-14 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Lu et al. | Flexible and scalable optical interconnects for data centers: Trends and challenges | |
US9848254B2 (en) | Efficient transport network architecture for content delivery network | |
Kamchevska et al. | Experimental demonstration of multidimensional switching nodes for all-optical data center networks | |
US11102110B2 (en) | Data center network (DCN) architecture and communication | |
Xue et al. | Flexibility assessment of the reconfigurable OPSquare for virtualized data center networks under realistic traffics | |
US8358932B2 (en) | All-optical data center network | |
JP7307388B2 (ja) | 光コア部を有する光ネットワーク、タイムスロット化した受信を使用するノード | |
CN100428660C (zh) | 一种具有内部加速的光突发交换节点 | |
Liu et al. | Space-frequency joint contention scheduling algorithm based on AoD in SDM-EONs | |
Wu et al. | Joint scheduling and routing for QoS guaranteed packet transmission in energy efficient reconfigurable WDM mesh networks | |
Christodoulopoulos et al. | Efficient bandwidth allocation in the NEPHELE optical/electrical datacenter interconnect | |
WO2022176172A1 (ja) | タイムスロット化した受信を使用する光ネットワークおよびノード | |
Chen et al. | Optimized parallel transmission in otn/wdm networks to support high-speed ethernet with multiple lane distribution | |
Yang et al. | An optical interconnect network design for dynamically composable data centers | |
Stavdas et al. | Dynamic CANON: A scalable multidomain core network | |
Drainakis et al. | Scalable and low server-to-server latency data center network architecture based on optical packet inter-rack and intra-rack switching | |
Shukla et al. | Optical switching in next-generation data centers: Architectures based on optical switching | |
Karthi et al. | WMRD net: An optical data center interconnect | |
Baziana | Asynchronous Transmission Strategy and Optical IntraCluster Connectivity for Hyperscale Data Centers | |
Jin et al. | Sandwich Tree: A new datacenter network based on passive optical devices | |
Yu et al. | Enhancing performance of cloud computing data center networks by hybrid switching architecture | |
Zhang et al. | PROPH: A Contention-free and Elastic Bandwidth Scheduling Scheme for AWGR-based Optical DCN | |
Baziana et al. | A Scalable Optical Intra-Cluster Data Center Network with an Efficient WDMA protocol: Performance and Power Consumption Study | |
Zhao et al. | Minimizing delay by multi-hop routing and scheduling in clos networks for DCNs | |
Drainakis et al. | Optical Intra-and Inter-Rack Switching Architecture for Scalable, Low-Latency Data Center Networks |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 21926605 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 2023500471 Country of ref document: JP Kind code of ref document: A |
|
WWE | Wipo information: entry into national phase |
Ref document number: 18262046 Country of ref document: US |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 21926605 Country of ref document: EP Kind code of ref document: A1 |