CN113794956B - Method and device for optical packet switching - Google Patents
Method and device for optical packet switching Download PDFInfo
- Publication number
- CN113794956B CN113794956B CN202110847054.1A CN202110847054A CN113794956B CN 113794956 B CN113794956 B CN 113794956B CN 202110847054 A CN202110847054 A CN 202110847054A CN 113794956 B CN113794956 B CN 113794956B
- Authority
- CN
- China
- Prior art keywords
- optical
- switching
- node
- message
- authorization
- Prior art date
- Legal status (The legal status 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 status listed.)
- Active
Links
- 230000003287 optical effect Effects 0.000 title claims abstract description 257
- 238000000034 method Methods 0.000 title claims abstract description 139
- 238000013475 authorization Methods 0.000 claims abstract description 128
- 230000008569 process Effects 0.000 claims abstract description 82
- 238000012545 processing Methods 0.000 claims abstract description 23
- 238000004891 communication Methods 0.000 claims abstract description 17
- 230000005540 biological transmission Effects 0.000 claims description 36
- 239000013307 optical fiber Substances 0.000 claims description 30
- 238000005259 measurement Methods 0.000 claims description 10
- 238000012423 maintenance Methods 0.000 claims description 8
- 230000003111 delayed effect Effects 0.000 claims description 6
- 238000006243 chemical reaction Methods 0.000 claims description 2
- 238000005516 engineering process Methods 0.000 description 15
- 230000000903 blocking effect Effects 0.000 description 8
- 230000001360 synchronised effect Effects 0.000 description 7
- 238000010586 diagram Methods 0.000 description 6
- 239000000835 fiber Substances 0.000 description 5
- 230000032683 aging Effects 0.000 description 3
- 230000003139 buffering effect Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 230000002776 aggregation Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001934 delay Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000011010 flushing procedure Methods 0.000 description 1
- 230000009191 jumping Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 230000000704 physical effect Effects 0.000 description 1
- 230000000644 propagated effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 230000008054 signal transmission Effects 0.000 description 1
Images
Classifications
-
- 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
- H04Q11/0001—Selecting arrangements for multiplex systems using optical switching
- H04Q11/0062—Network aspects
-
- 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
- H04Q11/0062—Network aspects
- H04Q2011/0075—Wavelength grouping or hierarchical aspects
-
- 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
- H04Q11/0062—Network aspects
- H04Q2011/0079—Operation or maintenance aspects
Abstract
The invention relates to a method and a device for optical packet switching, belonging to the technical field of optical packet switching of optical communication. The device comprises an optical switching node and an edge node; the method uses optical packet switching and virtual output queues, and comprises the processes of timing positioning, reporting, authorization list updating and authorization, and specifically comprises the following steps: the optical switching node establishes a link with the edge node mutual handshake message, and then mutually sends a positioning message to measure the distance of the edge node; in each time slot of the communication stage, the optical switching node receives a queue length report message sent by the edge node, processes the port conflict again and adds the processing result into an authorization list; adjusting the sending time according to the distance of the edge node; and transmitting the authorization message to the edge node, and then converging and transmitting the data packet. The method avoids a large amount of intervals caused by data packets, avoids the conflict between the output ports of the optical switch, and realizes the optical packet switching between long-distance edge nodes.
Description
Technical Field
The invention relates to a method and a device for optical packet switching, belonging to the technical field of optical packet switching of optical communication.
Background
Optical communication is an important cornerstone of modern communication, and more than 90% of data traffic needs to be propagated through optical fibers. The two most important directions in the field of optical communications include optical transmission technology and optical switching technology. The optical transmission technology is used for realizing the efficient transmission of information in the optical fiber channel; while optical switching technology is used to achieve flexible forwarding of signals. At present, an optical network uses one wavelength as the minimum channel granularity, and the number of wavelengths supported by optical fibers is limited, so that one node cannot establish a transmission channel with all other nodes. Therefore, optical switching technology is relied upon to enable information exchange between any two nodes in a limited fiber connection. Optical switching technology determines the flexibility of optical signal transmission and the granularity of switching.
Optical switching technologies are classified into Optical Circuit Switching (OCS), Optical Burst Switching (OBS), and Optical Packet Switching (OPS) according to the granularity of optical switching. The OCS has the minimum bandwidth distribution granularity of wavelength, is usually used in the fields of backbone networks and the like, and is characterized in that after a transmission channel is configured, the transmission channel cannot be dismantled in a short time, the service time is long, and the switching frequency is low. The switching granularity of OBS and OPS is small, respectively optical burst/packet. One optical burst is maintained in the microsecond-millisecond order, the length of the optical burst is not fixed and is determined by the arrival time of the flow; the optical packet is fixed in length, and the information data content included in one optical packet is fixed. The difference between the two is that the OBS adopts asynchronous switching, and different output ports of optical switching can be asynchronous; the edge nodes of the OPS must be synchronized to ensure that data packets arrive at the optical switching node at the same time.
In the current OPS optical switching node-edge node related technology, if an edge node uses a queue to buffer an optical packet. Packets Of different destination addresses will create a Head-Of-Block (HOB) problem. The HOB problem is described as follows, for a buffer with only one queue, a packet at the head of the queue has only one destination port, and since the packet is not sent in port contention, the packet behind the queue cannot occupy an available port, and finally, the throughput of each port can only reach 58% of the linear speed, which seriously reduces the transmission efficiency of the optical switching apparatus.
For head-blocking problems: both edge nodes are destined to port 2 and if the packet is not successfully forwarded, the data packet behind the queue cannot get to the port and can only be in a blocking state. The header data determines the port that the edge node must acquire, and if the port's forwarding authority cannot be obtained, the queue is always blocked. In order to solve the problem of head blocking HOB, a cache organization manner of a Virtual Output Queue (VOQ Queue) is generally adopted. The output buffer is divided into a plurality of queues by the VOQ queues, and each output port of the optical switch is provided with an independent queue to ensure that the data packet does not have the problem of head blocking. However, since only one VOQ queue of data forming a data packet can be selected per slot and sent to the corresponding port of the switching node, how to select the VOQ queue becomes an important issue for the apparatus. Selecting different VOQ queues will cause different numbers of port collisions in the optical switch array, resulting in different switching performance. Therefore, in the VOQ queue, how to select the buffer to reduce the blocking rate becomes a core idea for solving the port conflict. The occurrence of the VOQ queue selection problem requires reasonable cache selection by the edge node to reduce the probability of port collision and improve the throughput.
Port collision problem for OPS systems: the OPS technology is different from the conventional switching technology of the electrical domain, the optical switch does not have any buffer, and the data from the edge node transparently passes through the optical switch without being buffered. Therefore, the greatest technical difficulty of optical switching technology compared to electrical switches is how to export an input optical signal to an output port without buffering data packets; second, there may be multiple inputs with the same output port at the same time, which is called port collision (or port contention). Such as: both port 1 and port 3 packets are destined for port 2, and port collisions occur between the two, resulting in one of the packets being blocked. Another major problem of optical switches is how to handle port contention without data loss due to lack of optical buffering, which is called the port collision problem. Because the distance difference between the edge node and the optical switching node is large, how to ensure that the optical packets of all the edge nodes reach the optical switching node at the same time and realize synchronous switching becomes an important bottleneck of the OPS technology.
Timing problem for non-equidistant edge nodes: the conventional method for solving the port conflict has a large problem in the case of large distance difference between the edge node and the optical switching node in the OPS system. First, because the optical packet has a short duration, there may be a time difference between the data packets sent by different edge nodes at the same time and arriving at the optical switching node after passing through optical fiber channels with different lengths. The time difference causes that the optical switching node cannot perform synchronous port switching, so that the port collision algorithm of the conventional synchronous processing system cannot be applied to an OPS system with large distance difference. Secondly, how to ensure the accurate synchronization of all non-equidistant edge nodes is a difficult problem facing at present. Not only the optical switching node needs to sense the distance of the edge node, but also the edge node needs to sense the distance reaching the optical switching node at the same time, thereby ensuring the accuracy of information reported by different edge nodes. In the case of a longer distance, there may be multiple "report messages" simultaneously in the edge node to switching node link. The calculation of the grant message is based on the received "report message", however, due to the distance difference, the optical switching node needs to take the hysteresis factor of the message into account when calculating the port grant. Therefore, the prior art faces many problems in optical OPS systems for long distance transmission.
Positioning problem of OPS fiber aging: in long-haul OPS systems, there are hundreds of meters to kilometers of fiber channels from edge nodes to optical switching nodes. The slot granularity of OPS is small, so the requirement for timing positioning is high. The distance from the edge node to the switching node changes gradually due to the aging of the optical fiber and the maintenance and replacement of the optical fiber. Therefore, how to continuously and accurately measure and track the distance of the edge node is a main problem at present. In the ranging scheme of the switching system in the existing scheme, the ranging of the optical fiber channel is generally carried out in the initialization stage, the existing scheme cannot track the distance and the optical path well, and the existing scheme has a great defect when being applied to an OPS system needing precise timing. Furthermore, the OPS system must ensure that the label and the optical packet data are received at the ports of the optical switching node at the same time. Therefore, the edge nodes must be precisely located and tracked in real time.
The invention mainly aims at the problems of port conflict, non-equidistant timing synchronization, positioning calibration and the like of an OPS technology, and provides a method and a device for solving optical packet switching, which are used for realizing the problems of port conflict and data synchronous transmission of an OPS system under the non-distance condition.
In the aspect of port collision of the OPS, there are many ways to solve the collision in the existing research, wherein the most typical way is to add an optical buffer device such as an optical fiber delay line to an optical switching node to store an optical packet with a port contention failure, forward the optical packet in a later cycle, and increase the capability of solving the port collision through the optical fiber delay line; another similar method may employ a wavelength converter to convert multiple same-wavelength signals of the colliding ports to different wavelengths, and then converge and transmit the signals to the destination port. Both of the above two methods for solving the conflict require additional hardware structures to be added to the optical switch to handle the packet conflict. Another way is to solve the collision by means of Flow control (Flow control), where the edge node sends data to the optical switching node every timeslot, and then the optical switching node feeds back a message to the edge node according to whether the data packet is successfully forwarded, and the edge node decides whether to retransmit the data packet according to the state of the feedback message. The method does not need to add extra hardware, and avoids the loss of the data packet by means of retransmission. However, this method still has a drawback in that a large number of retransmissions causes a drop in throughput. Therefore, at present, there is no method for solving the problem of port conflict without adding extra hardware of the system and reducing the performance of the system.
Disclosure of Invention
The invention aims to solve the problem of port conflict without adding extra hardware of a system and reducing the performance of the system, and provides a method and a device for optical packet switching.
In order to achieve the above object, the present invention adopts the following technical solutions.
The method and the device for optical packet switching comprise an optical packet switching device and an optical packet switching method;
the optical packet switching device comprises an optical switching node and an edge node;
the optical switching node comprises a switching controller, an optical switch array, a switching demultiplexer at each input end and a switching multiplexer at each output end;
the switching controller is provided with a logic unit with a high-speed signal processing function, and comprises a port conflict processing module which can process a control packet and control the optical switch array; the optical switch array comprises N optical switch modules which are switched rapidly;
each fast switching optical switch module is connected with one input end and N output ends, the input end and any one output end can be switched, and the switching time delay range of the optical switch is 3-5 nanoseconds;
the edge node comprises a packet header processor, a VOQ queue, a controller, a control cache, two pairs of optical transceivers, an edge multiplexer of a transmitting end and an edge demultiplexer of a receiving end; the edge node has an OPS signal processing function;
wherein, the packet header processor comprises an electric switch, and the two pairs of optical transceivers comprise transmitters Tx1 and Tx2 and receivers Rx1 and Rx 2; the packet header processor processes all data traffic reaching the edge node, specifically including uplink traffic from a traffic source node and downlink traffic of the optical switch; the electric switch forwards the flow switched inside the edge node, and the flow needing to pass through the optical switch is sent to the VOQ queue; in the VOQ queues, a corresponding queue is arranged aiming at the output port of each optical switching node; the controller processes the authorization message, selects a VOQ queue, controls a cache and is responsible for the timing transmission of two transmitters, and the control cache is used for storing the report message of the edge node; the edge demultiplexer separates the authorization message from the optical switching node from the data packet, the authorization message is sent to the controller, and the data packet is sent to the packet header processor; the controller comprises a port conflict processing module for processing the authorization information and controlling the timing transmission of the transmitter; the control packet and the data packet are respectively sent to an input port of an edge multiplexer, and the edge multiplexer converges the control packet and the data packet to an output port of the edge multiplexer;
the connection relationship of the optical packet switching device is as follows: the switching controller in the optical switching node is respectively connected with the switching multiplexer and the switching demultiplexer, the switching demultiplexer is connected with the optical switch array, and the optical switch array is connected with the switching multiplexer; the packet header processor in the edge node is connected with the VOQ queue, and the VOQ queue is connected with the Tx1 and the controller; the multiplexer is respectively connected with Tx1 and Tx2, and Tx2 is connected with the control buffer; the controller is connected with Rx 2; rx1 and Rx2 are connected to the demultiplexer respectively; rx1 is connected to the packet header processor and demultiplexer;
the working process of the optical packet switching device is as follows: after the control message and the data packet are respectively sent to the switching demultiplexer through the control channel and the data channel, the control message is sent to the switching controller through the switching demultiplexer, and the data packet is sent to the optical switch array through the switching demultiplexer; after the port conflict processing module in the exchange controller resolves the port conflict, the authorization result is put into an authorization list; the data packet and the authorization message are sent to an edge multiplexer of the edge node through two independent channels after passing through the switch multiplexer; the authorization message is sent to the edge receiver Rx2 after passing through the edge demultiplexer, and the data packet is sent to the edge receiver Rx1 after passing through the edge demultiplexer;
wherein the control channel occupies one independent wavelength in the optical fiber communication, and the data channel occupies another independent wavelength in the optical fiber communication.
The method of the optical packet switching comprises the processes of timing positioning, reporting, authorization list updating and authorization;
the timing positioning process comprises a coarse timing positioning process and a timing calibration process, and comprises the following steps:
step 1.1 the edge node first sends out handshake message to the optical switching node;
the handshake message comprises protocols supported by the node and MAC address physical attributes;
step 1.2 after receiving the handshake message, the optical switching node also sends a handshake feedback message to the edge node, thus completing the initialization phase and establishing the communication connection between the optical switching node and the edge node;
the clock calibration and positioning measurement of one edge node comprises the following substeps:
step 2.1, the edge node firstly sends a timing message to the optical switching node, wherein the message contains the current time;
step 2.2, after receiving the timing message, the optical switching node writes the current time into the message and updates the timing message into a two-time timing message;
step 2.3, starting timing from the current time, and sending a two-time timing message after a fixed time, wherein the fixed time is known;
wherein the fixed time ranges from 2 to 20 ns;
step 2.3, after receiving the timing messages of the two moments, the edge node records the current receiving moment and calculates to obtain the optical fiber transmission delay;
wherein, the optical fiber transmission time delay is equal to the current receiving time minus the current time in the step 2.1, and then the time obtained by subtracting the fixed time is divided by 2;
step 2.4, calibrating the clock of the edge node into the clock of the optical switching node based on the optical fiber transmission delay calculated in the step 2.3;
so far, through the steps 2.1 to 2.4, clock calibration and positioning measurement of an edge node are completed;
dividing the optical fiber transmission time delay by the device time slice to obtain a remainder, namely the phase delay of the edge node; dividing the optical fiber transmission time delay by the device time slice and rounding down to obtain an integer as the time slot delay of the edge node;
the time slot delay of the edge node, that is, the number of device time slices included in the optical fiber transmission delay;
step 5, after step 4, every other fixed period, the optical switching node records the current recording time and adds fixed time, time slot delay and phase delay to form a timing calibration message, and then the timing calibration message is sent to the edge node after fixed time based on the current recording time of the timing calibration message;
the edge node finishes the accurate calibration of the clock after receiving the timing calibration message;
wherein the fixed period ranges from 5 milliseconds to 10 seconds;
so far, from step 1 to step 5, the timing positioning process of the optical packet switching method is completed;
after the timing positioning process is finished, the device realizes clock synchronization and distance measurement of all non-equidistant edge nodes, so that data exchange is started, and the device enters a reporting process;
each time slot of the device triggers a reporting process, comprising the steps of:
step 6, the edge node writes the size of the VOQ queue into a report message, and waits for the end of the time slot;
step 7, entering a second time slot, continuing to wait for a waiting time, and then the Tx2 sending the report message to the optical switching node through the control channel, and simultaneously the Tx1 sending the data packet to the optical switching node through the data channel;
the waiting time is the time slot minus the phase delay of the edge node;
step 8, after the switching node receives the report message, the switching demultiplexer sends the report message to the switching controller, and the switching demultiplexer sends the data packet to the optical switch array to complete data switching; to master
So far, the reporting process of the edge node is completed from step 6 to step 8, the switching controller completes the reporting process after receiving the reporting messages of all the edge nodes, the reporting process ensures that the reporting messages of all the edge nodes can reach the optical switching node synchronously, and then the device enters the maintenance process of the authorization list;
the maintenance process of the authorization list comprises the following steps:
wherein, N is the time slot delay of the edge node;
recording the difference value from the time slot delay of each edge node to the maximum time slot delay of all edge nodes as D;
step 12, inserting the port authorization result after the transmission timing sequence is adjusted into a time slot position corresponding to the transmission time in the authorization list, and then waiting for entering the next time slot;
so far, from step 9 to step 12, the conversion from the report message to the port authorization result is realized, the maintenance process of the authorization list is completed, and then the authorization process is entered;
an authorization process comprising the steps of:
step 13, when the next time slot starts, the switching controller configures the current optical switch array as the port authorization result calculated before the 2D +2 nd time slot; meanwhile, when the next time slot starts, whether the authorization list has a complete port authorization result is judged, and whether the authorization message is sent to the edge node is determined according to whether the complete port authorization result exists, which specifically comprises the following steps:
if the complete port authorization result exists, the exchange controller takes out the complete port authorization result with the most advanced time position from the authorization list to form an authorization message, then sends the authorization message to the edge node, and skips to the step 14;
if not, not sending an authorization message, and jumping to the step 13;
step 14, the edge multiplexer executes a waiting time after receiving the authorization message;
step 15, the edge node enters a second time slot of the authorization process, the edge node of the time slot gathers the data traffic into optical packets, and then waits for entering a third time slot of the authorization process;
step 16, after the third time slot of the authorization process, the edge node continues to execute a waiting time;
step 17, sending the data packet and the report message to the optical switching node at the same time;
so far, the authorization process of the device is completed from step 13 to step 17, and data packet exchange between non-equidistant edge nodes is realized;
from step 1 to step 17, the apparatus completes the optical packet switching method between non-equidistant edge nodes through the processes of timing positioning, reporting, authorization list updating and authorization.
Advantageous effects
The invention relates to a method and a device for solving optical packet switching, which have the following beneficial effects compared with the existing optical packet switching method and device:
1. aiming at the problems of port conflict, non-equidistant timing synchronization, positioning calibration and the like of optical packet switching, the method avoids the port conflict of an optical switch array through the process of report-authorization, calculates the port matching in advance, authorizes the port to the edge node to avoid the reduction of throughput caused by the port conflict, and realizes the packet switching between the non-equidistant edge nodes through the process of positioning and tracking and the process of synchronous logic, so the device and the method can realize the data packet switching of the edge nodes at any distance within 10 km;
2. the device and the method only change the control logics of data packets and report messages, no additional hardware module is added to store the conflicted optical packets, the information exchange of the optical packets is realized through a report-authorization logic process, and the device solves the problem of port conflict under the conditions of not increasing additional hardware of a system and not reducing the performance of the system;
3. the device reduces idle time among optical packets, in the existing disclosed scheme, the distance among the optical packets is about 200 nanoseconds, the device and the method can reduce the distance among the optical packets to about 50 nanoseconds, and an optical switch array can not generate port flushing, so that the service efficiency of a link is greatly improved, and compared with the existing optical packet switching device and method, for example, a rapid optical packet switching technology with 1 microsecond time slot is adopted, and the throughput can be improved by 11.9% under the condition of the same data rate;
3. the method and the device solve the problems that the existing switching device or method cannot track the distance and the optical path well, the optical fiber distance changes due to factors such as line aging and the like, the prior art is lack of accurate tracking of the optical fiber distance, the optical path change has large influence on the timing positioning of an OPS system, the device and the method monitor and adjust the change of the optical path of the line in real time through two processes of coarse timing positioning and accurate timing tracking, and the accurate positioning and real-time tracking of the edge node are realized.
Drawings
Fig. 1 is a schematic process diagram of a head blocking problem addressed by the method and apparatus for optical packet switching according to the present invention;
fig. 2 is a schematic diagram illustrating the generation of VOQ queue port collision that is addressed by the method and apparatus for optical packet switching according to the present invention;
fig. 3 is a general flowchart of an optical packet switching method and apparatus according to the present invention;
fig. 4 is a diagram of a switching node apparatus for solving optical node collisions according to an optical packet switching method and apparatus of the present invention;
fig. 5 is a diagram of an edge node apparatus for solving optical node collision according to an optical packet switching method and apparatus of the present invention;
fig. 6 is a schematic diagram of a timing alignment stage process of an optical packet switching method and apparatus according to the present invention, which is directed to solving the timing alignment problem;
fig. 7 is a data exchange flow chart between an edge node and an optical switching node according to an optical packet switching method and apparatus of the present invention;
fig. 8 is a diagram of an authorization process of a four-node optical switching apparatus according to an optical packet switching method and apparatus of the present invention.
Detailed Description
An optical packet switching method and apparatus according to the present invention are described in detail below with reference to the accompanying drawings and embodiments.
Example 1
The device and the method can be applied to scenes with higher requirements on data exchange rate, such as a data center network, a metropolitan area network and the like, particularly to a switching device under a medium-long distance optical network, such as a fine-grained exchange scene between a large-scale data center internet and the metropolitan area network; the apparatus and method can provide rate transparent, modulation format transparent high speed information exchange for non-equidistant edge nodes. Under the condition of a plurality of non-equidistant edge nodes, the device can provide the information exchange function of the optical grouping level for the metropolitan area network, thereby greatly improving the utilization rate and the flexibility of the channel resources of the metropolitan area network.
As shown in fig. 1, which is a process of the head blocking problem, two edge nodes are both addressed to port 2, and if the packet is not successfully forwarded, the data packet behind the queue cannot obtain the port, and only in the blocking state. The head data determines the port which must be obtained by the edge node, if the forwarding authorization of the port cannot be obtained, the queue is always in a blocking state;
fig. 2 shows the occurrence of port collision in VOQ queues, where data packets at port 1 and port 3 are both sent to port 2, and port collision occurs between the two, resulting in one of the data packets being blocked. Another major problem of optical switches is how to handle port contention without data loss due to lack of optical buffering, which is called the port collision problem. Because the distance difference between the edge node and the optical switching node is large, how to ensure that the optical packets of all the edge nodes reach the optical switching node at the same time and realize synchronous switching becomes an important bottleneck of the OPS technology.
In order to solve the above problems, an optical packet switching method and apparatus are provided.
When the device is implemented, a typical example of each fast-switching Optical switch is a Semiconductor Optical Amplifier (SOA), and the switching time delay of the switch is about 3-5 nanoseconds; the multiplexer and demultiplexer are general purpose optical devices; the multiplexer and the demultiplexer are Arrayed Waveguide Gratings (AWG); a typical example of the switching controller is a Field Programmable Gate Array (FPGA). In specific implementation, the edge node is an ethernet switch including an optical transceiver, and a typical example of the packet header processor is a universal switching chip;
a typical example of the controller is an FPGA, or an Application-Specific Integrated Circuit (ASIC).
When the step 1 is implemented, the method comprises the following substeps:
step 1.1 edge node (A) first sends out handshake message (H)A) To an optical switching node (B), the message content comprises protocols supported by the node, port number and MAC address physical attributes;
step 1.2 optical switching node receives HAThereafter, a handshake message (H) is also sentB) The edge node is reached, thus completing the initialization phase, and the two nodes establish the communication connection;
2.2 when the method is implemented specifically, because the processing process of the message needs a certain time, the existing method ignores the processing time of the message, which results in poor algorithm precision;
fig. 3 shows a general flow of an optical packet switching method and apparatus, in which an optical switching node and an edge node first send handshake messages to each other to establish a link, and then send positioning messages to each other to measure a distance between the edge nodes; then entering a communication stage, wherein in each time slot, the optical switching node receives a report message of the queue length sent by the edge node, processes port conflict and adds a processing result into an authorization list; the authorization list adjusts the sending time according to the distance of the edge node; the optical switching node sends the authorization message in the authorization list to the edge node for processing, and the edge node receives the authorization message and then gathers and sends the data packet.
Due to the phase offset from each edge node to the optical switching node, the transmission time of each edge node needs to be adjusted to ensure that the report message and the data packet of all the optical switching nodes can arrive at the optical switching node at the same time. Assuming that the phase deviation from the edge node to the optical switching node is Dt, Dt is 0 to 3 microseconds, and a time slot ts is 1 microsecond to 3 microseconds, the sending time of the edge node needs to be delayed by (ts-Dt), thereby ensuring that the report message and the data packet of the edge node arrive at the optical switching node at the same time; after completing the synchronization of the reporting procedure, the edge node starts to send report messages and data packets to the optical switching node; after receiving the report message, the optical switching node processes the port conflict, that is, the matching of the input-output ports needs to meet the requirement that each output port is matched with at most one input port, and each input port is matched with at most one output port; then inserting the matching result into an authorization list; after each time slot is finished, the exchange controller resets the exchange switch, then takes out the authorization message from the authorization list and sends the authorization message to each edge node; if the current optical switching node receives the report message R0If there are n complete time slots on the round-trip link, n is an integer from 0 to 30 (each time slot contains a control message and a data packet, and n is the same as the time slot delay Ds); suppose the grant message n slots before the slot is G-nThe report message generated m time slots later is Rm。
For a 4-port optical switch, R may be a 4-dimensional vector [0, 30, 10, 50 ]]Indicating the flow of the VOQ queue, wherein 10 indicates the flow of 10kb, and the value is 0-300; g-nCan be a 4-dimensional vector, such as [0,0,10,0 ]]Indicating that the authorized port is 3 and 10 indicating that traffic can be sent 10kb in the authorized slot.
The optical packet switching method comprises the following steps:
step 2.1, the optical switching node carries out clock calibration and positioning distance measurement on the edge node according to the received 'timing message', and calculates time slot delay and phase deviation of the optical path according to the distance;
then entering a data packet exchange stage, wherein each time slot is subjected to a reporting process and an authorization process;
step 5, in each time slot, the optical switch sends an authorization message according to the authorization list "
Step 5.1, after receiving the authorization message, the edge node aggregates the traffic of the corresponding VOQ queue into an optical packet format;
and 5.2, after the optical packet format is aggregated, waiting for the next time slot to send.
Thus, optical packet switching is completed through step 1 to step 5.2.
Fig. 4 shows a switching node apparatus of an optical packet collision solution, where a control message and a data packet are sent to an optical switching node demultiplexer through two independent channels, the control message is sent to a switching controller through the demultiplexer, and the data packet is sent to a port collision processing module of an optical switch array through the demultiplexer; and the port conflict processing module is used for solving the port conflict and then putting the authorization result into an authorization list.
Fig. 5 shows an edge node apparatus of an optical packet collision solution, where in the edge node structure, all data traffic arriving at an edge node needs to be processed by a packet header processor, including uplink traffic from a traffic source node and downlink traffic of an optical switch; the flow exchanged in the edge node is forwarded through the electric switch, and the flow needing to pass through the optical switch is sent to the VOQ queue; the controller processes the authorization message, selects the VOQ queue, controls the buffer and is responsible for the timing transmission of the two transmitters; the control cache is used for storing the report message of the edge node; the demultiplexer separates the authorization message from the data packet from the optical switching node, the authorization message is sent to the controller, and the data packet is sent to the packet header processor; the processor comprises a port conflict processing module for processing the authorization information and controlling the timing transmission of the transmitter; in the VOQ queues, there is a corresponding queue for each output port of the optical switching node.
FIG. 6 shows a timed alignment phase procedure where the edge node (A) first sends out a handshake message (H)A) To the optical switching node (B), the message content contains the physical properties of protocols, port numbers, MAC addresses, etc. supported by the node. Optical switching node receives HAThereafter, a handshake message (H) is also sentB) To the edge node. This completes the initialization phase and both establish a communication connection. And after finishing the handshake process, entering a timing positioning process. Node A first sends a timing message (L (t))1) To node B) the message contains the time t at that time1(ii) a Optical switched reception of L (t)1) Then, the current time t is compared2Write to the message, i.e. become L (t)1，t2) (ii) a Thereafter, the A node receives L (t)1，t2) And will receive the time t3And (7) recording. The final available fiber transmission delay Δ t is: Δ t ═ t (t)3-t1-t')/2, Δ t may take a value of 0 to 33 microseconds; the a-node clock (T1) is then calibrated to the B-node clock, i.e.: t1 ═ T2+(t3-t1+t’)/2。
Assuming that a time slice ts of the device takes a value from 1 microsecond to 3 microseconds, the time slot delay Ds of the edge node may be an integer from 0 to 30, that is, the number of the whole time slot slices included in Δ t, and Δ t is divided by ts and rounded down; the phase deviation Dt is: Dt-Ds ts, which may be 0 to 1 microsecond; based on the obtained Ds and Dt, the edge node inserts the Ds and Dt into a message to form L (Ds, Dt) and then sends the L (Ds, Dt) to the optical switching node to report the measurement result, and the rough calibration of the device is completed; then, the device adjusts the sending time of the report message and the authorization message to complete the long-distance system exchange;
in the process of data forwarding, every other fixed period, the optical switching node adds the current time with the fixed time (t', the value of which is 3 to 10 nanoseconds) and the optical fiber time delay (delta t, the value of which is 0 to 33 microseconds), and then puts in a fourth optical signal; then after a time t', the message is sent to the edge node; and after receiving the fourth optical signal, the edge node finely calibrates the clock of the edge node. And completing accurate timing calibration of the clock and the distance under the condition of optical path time variation.
Fig. 7 shows a data exchange process between an edge node and an optical switch node, where each start, the edge node sends a data packet and sends a report message to the optical switch via a control channel, for example, the phase offset Dt from the edge node to the optical switch node is 200 ns, the delay offset is 2, the time ts of one time slot is 1 ms, the sending time of the edge node needs to be delayed by 800 ns, after the synchronization of the report process is completed, the edge node starts to send the report message and the data packet to the optical switch node, the optical switch node processes the port collision after receiving the report message, the system adopts a conventional bipartite graph matching algorithm, such as LPF and MWM, to solve the port collision, then inserts the matching result into an "authorization list", and after the system enters a new time slot, the switch controller resets the switch to a port authorization result before 6 time slots, meanwhile, the authorization message is taken out from the authorization list and then issued to each edge node, after the edge nodes receive the authorization message, the edge nodes wait for 800 nanoseconds to start processing the labels and aggregate the data packets, and after the aggregation of the data packets is completed, the edge nodes wait for a third time slot to send the data packets; after entering a third time slot, firstly waiting for 800 nanoseconds, and then sending a data packet;
fig. 8 shows an authorization process of a four-node optical switching apparatus, assuming that an optical switching apparatus has four edge nodes (denoted by 1, 2, 3, and 4), and the time slot delays to the optical switching nodes are 2, 1, 4, and 3 time slots, respectively, after receiving a report message, because there is an authorized flow but an unrewarded flow in a link, it is necessary to process the existing report message. G-1To G-nHas authorized the forwarding of part of the traffic, but has not forwarded in time due to the fiber delay. If the report message is used for calculating the port authorization, repeated authorization is easily caused, and the bandwidth is wasted. Therefore, the proposed apparatus of this scheme first processes the report message. Assuming that the current time slot index is 0, the report message of node 1 is R0(1)＝{qi,j，i∈[1,n]}，qi,jIndicating the length of the ith VOQ queue in the edge node 1. The authorization message is noted as:
G0＝{g0(j)},j∈[1,n]
authorization message, g, indicating time 00(1) Representing the traffic granted to the edge node 1, then R':
is processed as R 'according to the process report message shown in FIG. 8'0(1) Then obtaining an original authorization list through a matching algorithm; because the authorization list needs to adjust the time sequence according to the time slot delay, a group of authorization results G are obtained through a bipartite graph matching algorithm0Due to delay of edge nodes in time slotAlso, the switch controller adjusts the transmission time slot of the grant message before inserting the grant list. First, the device finds Ds (i), i ∈ [1, n ]]The maximum value of the medium delay, namely node 3, Ds is maximum 4; then, the sending time of all authorization messages is delayed by 2 x (Ds (i) -max { Ds (i) }), namely 6 time slots, and then the messages are inserted into the authorization list. The grant sequence for the four nodes is delayed by 4, 6, 0, 2, respectively. Therefore, as shown in the figure, the grant message sent in the next time slot is g-4(1)、g-6(2)、g0(3)、g-2(4). The dotted line element in the authorization list represents the authorization message which has been sent, and after the authorization message is sent next time, a completed authorization message does not exist in the list, that is, the queue length of the edge node which is farthest away in the list is 0.
While the foregoing is directed to the preferred embodiment of the present invention, it is not intended that the invention be limited to the embodiment and the drawings disclosed herein. It is intended that all equivalents and modifications which do not depart from the spirit of the invention disclosed herein are deemed to be within the scope of the invention.
Claims (7)
1. An optical packet switching method comprises the processes of timing positioning, reporting, authorization list updating and authorization;
the timing positioning process comprises a coarse timing positioning process and a timing calibration process, and comprises the following steps:
step 1, initializing the optical packet switching device, specifically: an edge node is just accessed to an optical switching node, the edge node firstly needs to send handshake messages to the optical switching node, and communication connection is established, and the method specifically comprises the following substeps:
step 1.1, the edge node firstly sends out handshake information to the optical switching node;
step 1.2 after receiving the handshake message, the optical switching node also sends a handshake feedback message to the edge node, thus completing the initialization phase and establishing the communication connection between the optical switching node and the edge node;
step 2, after the initialization is completed in the step 1, coarse timing positioning is carried out, and clock calibration and positioning measurement are carried out by traversing edge nodes;
step 3, acquiring the time slot delay and the phase delay of the edge node based on the optical fiber transmission delay, specifically:
dividing the optical fiber transmission delay by the device time slice to obtain a remainder, namely the phase delay of the edge node; dividing the optical fiber transmission time delay by the time slice of the device and rounding down to obtain an integer as the time slot delay of the edge node;
the time slot delay of the edge node, that is, the number of device time slices included in the optical fiber transmission delay;
step 4, the controller of the edge node writes the time slot delay and the phase delay into a timing calibration message and sends the timing calibration message to the optical switching node, and the optical switching node obtains and records the time slot delay and the phase delay;
step 5, after step 4, every other fixed period, the optical switching node records the current recording time and adds fixed time, time slot delay and phase delay to form a timing calibration message, and then the timing calibration message is sent to the edge node after fixed time based on the current recording time of the timing calibration message; the edge node finishes the accurate calibration of the clock after receiving the timing calibration message;
wherein the fixed period ranges from 5 milliseconds to 10 seconds;
so far, from step 1 to step 5, the timing positioning process of the optical packet switching method is completed;
after the timing positioning process is finished, the device realizes clock synchronization and distance measurement of all non-equidistant edge nodes, so that data exchange is started, and the device enters a reporting process;
each time slot of the device triggers a reporting process, comprising the steps of:
step 6, the edge node writes the size of the VOQ queue into a report message, and waits for the end of the time slot;
step 7, entering a second time slot, continuing to wait for a waiting time, and then the Tx2 sending the report message to the optical switching node through the control channel, and simultaneously the Tx1 sending the data packet to the optical switching node through the data channel;
step 8, after the switching node receives the report message, the switching demultiplexer sends the report message to the switching controller, and the switching demultiplexer sends the data packet to the optical switch array to complete data switching;
so far, the reporting process of the edge node is completed from step 6 to step 8, the switching controller completes the reporting process after receiving the reporting messages of all the edge nodes, the reporting process ensures that the reporting messages of all the edge nodes can reach the optical switching node synchronously, and then the device enters the maintenance process of the authorization list;
the maintenance process of the authorization list comprises the following steps:
step 9, the exchange controller firstly processes the report message, and subtracts the sum of the flow of the VOQ queue authorized by the edge node in the previous N time slots from the report message of the edge node at the current moment to obtain the processed report message;
step 10, matching bipartite graphs of the processed report message to obtain a port authorization result;
step 11, adjusting the sending time sequence of the port authorization result, firstly calculating the difference value between the time slot delay of each edge node and the time slot delay maximum value of all edge nodes, and then determining the sending time of the port authorization result as the time slot delayed by two times of the difference value;
recording the difference value from the time slot delay of each edge node to the maximum time slot delay of all edge nodes as D;
step 12, inserting the port authorization result after the transmission timing sequence is adjusted into a time slot position corresponding to the transmission time in an authorization list, and then waiting for entering the next time slot;
so far, from step 9 to step 12, the conversion from the report message to the port authorization result is realized, the maintenance process of the authorization list is completed, and then the authorization process is entered;
an authorization process comprising the steps of:
step 13, when the next time slot starts, the switching controller configures the current optical switch array as the port authorization result calculated before the 2D +2 nd time slot; meanwhile, when the next time slot starts, whether the authorization list has a complete port authorization result is judged, and whether the authorization message is sent to the edge node is determined according to whether the complete port authorization result exists, which specifically comprises the following steps:
if the complete port authorization result exists, the exchange controller takes out the complete port authorization result with the most advanced time position from the authorization list to form an authorization message, then sends the authorization message to the edge node, and jumps to the step 14;
if not, if no complete port authorization result exists, no authorization message is sent, and the step 13 is skipped;
step 14, the edge multiplexer executes a waiting time after receiving the authorization message;
step 15, the edge node enters a second time slot of the authorization process, the edge node of the time slot gathers the data traffic into optical packets, and then waits for entering a third time slot of the authorization process;
step 16, after the third time slot of the authorization process, the edge node continues to execute a waiting time;
step 17, sending the data packet and the report message to the optical switching node at the same time;
so far, the authorization process of the device is completed from step 13 to step 17, and data packet exchange between non-equidistant edge nodes is realized;
to this end, from step 1 to step 17, the apparatus completes optical packet switching between non-equidistant edge nodes through a timing positioning process, a reporting process, a maintenance process of an authorization list, and an authorization process.
2. An optical packet switching method according to claim 1, wherein: in step 1.1, the handshake message includes the protocols supported by the node and the MAC address physical attributes.
3. An optical packet switching method according to claim 2, wherein: in step 2, the clock calibration and positioning measurement of an edge node includes the following substeps:
step 2.1, the edge node firstly sends a timing message to the optical switching node, wherein the message contains the current time;
step 2.2, after receiving the timing message, the optical switching node writes the current time into the message and updates the timing message into a two-time timing message;
step 2.3, starting timing from the current time, and sending a two-time timing message after a fixed time, wherein the fixed time is known;
wherein the fixed time ranges from 2 to 20 ns;
step 2.3, after receiving the timing messages of the two moments, the edge node records the current receiving moment and calculates to obtain the optical fiber transmission delay;
step 2.4, calibrating the edge node clock to the clock of the optical switching node based on the optical fiber transmission delay calculated in step 2.3;
thus, through steps 2.1 to 2.4, clock calibration and positioning measurement of one edge node are completed.
4. An optical packet switching method according to claim 3, wherein: in step 2.3, the optical fiber transmission delay is equal to the current receiving time minus the current time of step 2.1, and then the time obtained by subtracting the fixed time is divided by 2.
5. An optical packet switching method according to claim 4, wherein: in step 7, the size of the waiting time is the time slot minus the phase delay of the edge node.
6. An optical packet switching method according to claim 5, wherein: in step 8, N is the slot delay of the edge node.
7. An optical packet switching device depending on an optical packet switching method according to claim 6, characterized in that: the optical switching node comprises an optical switching node and an edge node;
the optical switching node comprises a switching controller, an optical switch array, a switching demultiplexer at each input end and a switching multiplexer at each output end;
the switching controller is provided with a logic unit with a high-speed signal processing function, and comprises a port conflict processing module which can process a control packet and control the optical switch array; the optical switch array comprises N optical switch modules which are switched rapidly;
each fast switching optical switch module is connected with one input end and N output ends, the input end and any one output end can be switched, and the switching time delay range of the optical switch is 3-5 nanoseconds;
the edge node comprises a packet header processor, a VOQ queue, a controller, a control cache, two pairs of optical transceivers, an edge multiplexer of a transmitting end and an edge demultiplexer of a receiving end; the edge node has an OPS signal processing function;
wherein, the packet header processor comprises an electric switch, and the two pairs of optical transceivers comprise transmitters Tx1 and Tx2 and receivers Rx1 and Rx 2; the packet header processor processes all data traffic reaching the edge node, specifically including uplink traffic from a traffic source node and downlink traffic of the optical switch; the electric switch forwards the flow switched inside the edge node, and the flow needing to pass through the optical switch is sent to the VOQ queue; in the VOQ queues, a corresponding queue is arranged aiming at the output port of each optical switching node; the controller processes the authorization message, selects the VOQ queue, controls the cache and is responsible for the timing transmission of the two transmitters, and the control cache is used for storing the report message of the edge node; the edge demultiplexer separates the authorization message from the data packet from the optical switching node, the authorization message is sent to the controller, and the data packet is sent to the packet header processor; the controller comprises a port conflict processing module for processing the authorization information and controlling the timing transmission of the transmitter; the control packets and the data packets are respectively sent to an input port of an edge multiplexer, and the edge multiplexer converges the control packets and the data packets to an output port of the edge multiplexer;
the connection relationship of the optical packet switching device is as follows: the switching controller in the optical switching node is respectively connected with the switching multiplexer and the switching demultiplexer, the switching demultiplexer is connected with the optical switch array, and the optical switch array is connected with the switching multiplexer; the packet header processor in the edge node is connected with the VOQ queue, and the VOQ queue is connected with the Tx1 and the controller; the multiplexer is respectively connected with Tx1 and Tx2, and Tx2 is connected with the control buffer; the controller is connected with Rx 2; rx1 and Rx2 are respectively connected with the demultiplexer; rx1 is connected to the packet header processor and demultiplexer;
the working process of the optical packet switching device is as follows: after the control message and the data packet are respectively sent to the switching demultiplexer through the control channel and the data channel, the control message is sent to the switching controller through the switching demultiplexer, and the data packet is sent to the optical switch array through the switching demultiplexer; after the port conflict processing module in the exchange controller resolves the port conflict, the authorization result is put into an authorization list; the data packet and the authorization message are sent to an edge multiplexer of the edge node through two independent channels after passing through the switch multiplexer; the authorization message is sent to the edge receiver Rx2 after passing through the edge demultiplexer, and the data packet is sent to the edge receiver Rx1 after passing through the edge demultiplexer;
wherein the control channel occupies one independent wavelength in the optical fiber communication, and the data channel occupies another independent wavelength in the optical fiber communication.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110847054.1A CN113794956B (en) | 2021-07-27 | 2021-07-27 | Method and device for optical packet switching |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110847054.1A CN113794956B (en) | 2021-07-27 | 2021-07-27 | Method and device for optical packet switching |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113794956A CN113794956A (en) | 2021-12-14 |
| CN113794956B true CN113794956B (en) | 2022-07-22 |
Family
ID=78876931
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110847054.1A Active CN113794956B (en) | 2021-07-27 | 2021-07-27 | Method and device for optical packet switching |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113794956B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1716824A (en) * | 2004-06-14 | 2006-01-04 | 电子科技大学 | Method for resoving light burst resource conflict |
| CN101193049A (en) * | 2006-11-24 | 2008-06-04 | 中兴通讯股份有限公司 | A core node switching structure for integrated edge node |
| CN106464412A (en) * | 2014-06-03 | 2017-02-22 | 华为技术有限公司 | Optical packet sending method and device, processing method and optical switching device |
| CN109451377A (en) * | 2018-11-28 | 2019-03-08 | 杭州电子科技大学 | Optical switch based on integrated form hybrid switching technique |
| CN112260788A (en) * | 2020-10-19 | 2021-01-22 | 北京邮电大学 | Dynamic high-precision time-frequency synchronous network |
| CN112995804A (en) * | 2020-12-29 | 2021-06-18 | 北京理工大学 | Optical switching method, device and system |
-
2021
- 2021-07-27 CN CN202110847054.1A patent/CN113794956B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1716824A (en) * | 2004-06-14 | 2006-01-04 | 电子科技大学 | Method for resoving light burst resource conflict |
| CN101193049A (en) * | 2006-11-24 | 2008-06-04 | 中兴通讯股份有限公司 | A core node switching structure for integrated edge node |
| CN106464412A (en) * | 2014-06-03 | 2017-02-22 | 华为技术有限公司 | Optical packet sending method and device, processing method and optical switching device |
| CN109451377A (en) * | 2018-11-28 | 2019-03-08 | 杭州电子科技大学 | Optical switch based on integrated form hybrid switching technique |
| CN112260788A (en) * | 2020-10-19 | 2021-01-22 | 北京邮电大学 | Dynamic high-precision time-frequency synchronous network |
| CN112995804A (en) * | 2020-12-29 | 2021-06-18 | 北京理工大学 | Optical switching method, device and system |
Non-Patent Citations (2)
| Title |
|---|
| Edge-buffered Contention-resolution Hybrid Optical/Electrical Switch for Data Center Networks;Bao Huang;《2018 Asia Communications and Photonics Conference (ACP)》;20181231;1-3 * |
| 光分组交换网络中光纤延迟线缓存技术;刘焕淋;《光电技术应用》;20090228;1-5 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113794956A (en) | 2021-12-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9391768B2 (en) | Transparent clock for precision timing distribution | |
| US7457542B2 (en) | Optical access network system | |
| US8571068B2 (en) | Network timing distribution and synchronization using virtual network delays | |
| KR102138731B1 (en) | Reduced synchronization errors in Common Public Radio Interface (CPRI) passive optical networks (PONS) | |
| JP2018520534A (en) | Time synchronization method and system, and network device | |
| Nakayama et al. | Wavelength and bandwidth allocation for mobile fronthaul in TWDM-PON | |
| WO2020057187A1 (en) | Passive optical network (pon) channel bonding protocol | |
| US10492160B2 (en) | Network node | |
| US20170238335A1 (en) | Method for radio source scheduling | |
| CN117157915A (en) | Time synchronization in passive optical networks | |
| US20120099854A1 (en) | Layer One Path Delay Compensation | |
| JP4913876B2 (en) | Bandwidth allocation apparatus and bandwidth allocation method | |
| CN113794956B (en) | Method and device for optical packet switching | |
| Pfeiffer | Considerations on transport latency in passive optical networks | |
| JP5504210B2 (en) | Optical communication system and optical communication method | |
| White et al. | Design of a control-channel-based media-access-control protocol for HORNET | |
| US11924592B2 (en) | Fast flow-controlled and clock-distributed optical switching system for optical data center network | |
| US7154879B1 (en) | Point to multipoint network | |
| JP2011166328A (en) | Optical transmission system, optical line terminal and upward transmission control method | |
| Sasu et al. | Packet Delay Variation Correction for Time Sensitive Networking with Frame Preemption | |
| WO2024045815A1 (en) | Registration method, related device and optical communication system | |
| WO2024038540A1 (en) | Switching system, optical packet transmission apparatus, and method of generating optical packet | |
| KR101531689B1 (en) | Apparatus and method for bonding channel in a data transfer system | |
| JP2016213770A (en) | Optical signal relay device, optical communication system, and optical signal relay method | |
| US9294370B2 (en) | System and method for reduced latency reporting in an ethernet passive optical network (EPON) |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CB03 | Change of inventor or designer information |
Inventor after: Wang Fu Inventor after: Xin Xiangjun Inventor after: Chang Tianhai Inventor after: Zhang Qi Inventor after: Gao Ran Inventor after: Yao Haipeng Inventor after: Wang Guangquan Inventor before: Wang Fu Inventor before: Xin Xiangjun Inventor before: Zhang Qi Inventor before: Gao Ran Inventor before: Yao Haipeng Inventor before: Wang Guangquan |
|
| CB03 | Change of inventor or designer information |