Time-sensitive network-based full-time global power communication network time delay guarantee method
By establishing a business model and priority division in the full-time, all-domain power communication network, combining genetic algorithms and taboo search algorithms, a time-sensitive traffic scheduling algorithm is proposed, which solves the difficulty of deterministic delay of time-sensitive services caused by multi-type terminal service access. The issue of guarantee achieves transmission guarantees with low latency and low jitter.
Patent Information
- Application Number
- CN202411035710.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-31
- Publication Date
- 2024-12-03
AI Technical Summary
The simultaneous access of multiple types of terminal services and massive data transmission in the full-time, all-domain power communication network leads to the problem that time-sensitive services are difficult to guarantee the certainty delay.
Time-delay guarantee method for full-time all-domain power communication network based on time-sensitive networks, through establishing business models and prioritization, using a joint optimization algorithm based on genetic algorithm and taboo search algorithm, a time-sensitive traffic scheduling algorithm is proposed to ensure the gated scheduling table Computes a traffic scheduling mechanism that satisfies the deterministic low-latency transmission of time-sensitive streams during the transmission cycle, and sets constraints for the transmission start time on each link through the time-sensitive network-aware shaping mechanism.
It improves scheduling performance, ensures the deterministic delay of time-sensitive services, reduces delay jitter, and realizes bounded low-latency transmission of massive data transmission.
Smart Images

Figure CN119071237A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of multi-service deterministic delay guarantee in all-time and all-domain electric power communication networks, and relates to a multi-service deterministic delay guarantee method in all-time and all-domain electric power communication networks, and in particular to a delay guarantee method in all-time and all-domain electric power communication networks based on time-sensitive networks. Background Art
[0002] With the vigorous construction of new power systems, the increased penetration rate of renewable energy and the large-scale application of power electronic equipment have made the system more complex. The trend of digital upgrading of power systems is obvious. The power communication network is gradually transforming into a more robust, flexible, multi-service full-time and full-domain communication network, supporting the integrated development of source, grid, load and storage, and laying a solid foundation for the safe and stable operation of new power systems.
[0003] Considering that intensive access of terminal devices brings massive business data, and emerging businesses such as distributed new energy and smart charging of electric vehicles continue to emerge, the demand for low-latency and high-reliability transmission of business data in the power communication network is becoming increasingly obvious. For example, the deterministic delay requirement of control command services usually needs to be less than 150ms, and the deterministic delay of telemetry transmission usually needs to be less than 200ms. The jitter requirements of the above scenarios are all in microseconds. In comparison, the deterministic transmission requirements of collection services are slightly looser.
[0004] Therefore, it is necessary to propose a delay guarantee method for the full-time and full-domain power communication network based on time-sensitive network, which can be used for massive new services in the full-time and full-domain power communication network, realize deterministic data transmission in the same network, that is, periodic key control communication data and non-periodic data are integrated and transmitted in the same network, and provide deterministic bounded low-latency transmission guarantee for time-sensitive business traffic, thereby optimizing network performance.
[0005] After searching, no patent documents of the prior art that are identical or similar to the present invention were found. Summary of the invention
[0006] The purpose of the present invention is to overcome the shortcomings of the prior art and propose a delay guarantee method for a full-time and global power communication network based on a time-sensitive network, which can solve the problem that it is difficult to guarantee the deterministic delay of time-sensitive services caused by the simultaneous access of multiple types of terminal services in the full-time and global power communication network and the transmission of massive data.
[0007] The present invention solves the practical problem by adopting the following technical solutions:
[0008] A method for ensuring time delay of a full-time and full-domain electric power communication network based on a time-sensitive network comprises the following steps:
[0009] Step 1: Establish a full-time and global electric power communication network service model and prioritize it;
[0010] Step 2: Based on the result of the multi-type service priority division in step 1, a traffic scheduling mechanism based on time-sensitive network-aware shaping is established, and based on the traffic scheduling mechanism, in order to ensure that the gating schedule calculation satisfies the deterministic low-latency transmission of the time-sensitive flow within the transmission cycle, a constraint condition is established on the transmission start time of each link;
[0011] Step 3: Taking the minimum end-to-end delay of time-sensitive traffic as the optimization goal, a heuristic optimization algorithm based on the genetic algorithm and the taboo search algorithm is proposed to propose a time-sensitive traffic scheduling algorithm based on the joint optimization of genetic taboo search, calculate and output the gating list of time-sensitive traffic, and output the scheduling plan with the minimum end-to-end delay.
[0012] Moreover, the specific steps of step 1 include:
[0013] (1) Introduce TSN technology into the full-time and full-domain power communication network to establish business models for new smart grid services, including smart grid distribution network automation services, smart grid power consumption information collection services, smart interactive marketing and smart power consumption management services, microgrids and distributed energy access;
[0014] (2) By analyzing the communication network requirements of various services, the multi-type services in the full-time and global power communication network are prioritized according to the service latency requirements and the importance of the services to the grid operation control;
[0015] (3) Using the priority field of the IEEE 802.1Q protocol, different types of data are divided into eight priorities:
[0016] Periodic real-time data is marked as the highest priority because of its high requirements for real-time and periodicity of communication;
[0017] Non-periodic real-time data requires lower communication latency and is therefore marked as the second highest priority;
[0018] The remaining priorities are marked according to traffic type and data usage;
[0019] Data of different priorities corresponds to different priority buffer queues to ensure orderly transmission of data.
[0020] Moreover, the specific steps of establishing the traffic scheduling mechanism based on time-sensitive network-aware shaping in step 2 are:
[0021] (1) The topological structure of the full-time and global power communication network is modeled as G = (E, V), where E represents the set of all links in the communication network, V represents the set of all nodes in the communication network, and the service flow f i ∈F from the starting node v 1∈v sends data through the switching node to the final node v according to the period T 2 ;
[0022] (2) Using the TAS gating list, the communication time is divided into cycles, and each cycle is further divided into two time slices. In the first time slice, the gate state of the highest priority queue is set to open. In the second time slice, the gate state of the queues with priorities 0 to 5 is set to open, allowing non-real-time data to be transmitted. Non-periodic real-time data is not restricted by the gate state and is transmitted within the sending interval of periodic real-time data. It can also preempt low-priority non-real-time data.
[0023] (3) A guard band is set at the end of each cycle. During the guard band, no data is allowed to be sent to avoid the data transmission of the next cycle being affected by the unfinished data transmission of the previous cycle. After the highest priority periodic real-time data enters the buffer queue, it is transmitted in the first time slice according to the gate state and the reserved transmission bandwidth, and its transmission time is recorded in the flow reservation information table. When non-periodic data is transmitted, the flow reservation information table is first queried and the data is transmitted during the idle period of periodic data transmission. If the sending port is transmitting low-priority non-real-time data, the non-periodic real-time data can interrupt its transmission and be sent first.
[0024] Moreover, in step 2, based on the traffic scheduling mechanism, in order to ensure that the gated scheduling table calculation satisfies the deterministic low-latency transmission of the time-sensitive flow within the transmission period, a constraint condition is established for the transmission start time on each link, and the constraint condition includes:
[0025] (1) Frame constraint: This constraint ensures that the transmission of a frame on the first link of the transmission path cannot start before the send time and must be completed on the last link before the deadline. The constraint expression is:
[0026]
[0027] Among them, v j and v k Represents the second node and the second to last node; For flow i In (v 1 ,v 2 ) transmission offset, The flow f i The sending time and deadline, The time taken to transmit any data frame.
[0028] (2) End-to-end delay constraint: The TT flow is transmitted from the sender to the receiver along the calculated optimal path. Its end-to-end delay includes propagation delay, processing delay, and transmission delay. The expression of the delay D between one switching node and the next switching node is as follows.
[0029]
[0030] The end-to-end delay of flow fi from the sending end to the receiving end through j switches should be less than or equal to Di. The constraint expression is:
[0031]
[0032] (3) Link constraint: When two different frames on a link overlap in transmission time, the propagation delay of the two frames will increase, resulting in the indeterminacy of the transmission of the two frames. Therefore, the transmission of two different frames on a link cannot overlap in time. That is, for each pair of different frames on the same link, the transmission of one frame must be completed before the transmission of the other frame begins, and vice versa; this constraint must take into account all frames in the cycle; due to the zero jitter requirement, the link (v i ,v j The transmission offset of the kth frame of stream fi on )∈E is equal to The constraint expression is:
[0033]
[0034] Among them, Ti represents the period of TT flow, and the cycle period of the gating list is equal to the least common multiple L of all flows.
[0035] (4) Queue constraint: This constraint ensures that each frame is correctly assigned a transmission queue on each egress port. The constraint expression is as follows:
[0036]
[0037] Among them, λ represents the queue information of frame storage, and Q represents the number of queues available on the link egress port.
[0038] Moreover, the specific steps of step 3 include:
[0039] (1) Setting the parameters of the genetic algorithm involves population size, crossover probability, mutation probability, and maximum number of iterations;
[0040] With the help of the centralized user configuration CUC defined in the IEEE 802.1Qcc standard and the centralized network configuration CNC supporting the NETCONF or RESTCONF protocol, a complete centralized management mechanism is provided. CUC is responsible for providing registration services for time-sensitive flows for key applications, collecting information such as the flow period, length, sending and receiving nodes, and maximum delay, and registering new devices and exchanging information with CNC through the user network interface UNI; CNC is responsible for calculating and configuring GCL to meet the strict real-time requirements for dynamic scheduling in large-scale network models; in the dynamic scheduling process, the core scheduling solution is required to have a low time complexity to meet the challenges brought by changes in network topology and traffic;
[0041] Whenever the traffic or network topology changes, the centralized management mechanism can promptly update algorithm-related parameters and adjust individual genes to ensure the normal iteration of the algorithm.
[0042] (2) Population initialization: The first step is to randomly initialize all feasible routes between the sender and receiver of each flow. Then, the Cartesian product of the routes selected by each flow is calculated, that is, the combination of the routes selected by each flow. The number of populations required for initialization by the algorithm is selected from them, rather than completely traversing the Cartesian product. Next, it is verified whether the generated individuals meet the constraints. If they do, the corresponding individual genes are generated; otherwise, they are discarded.
[0043] In the initialization phase of route optimization, the load data volume in each link is accumulated. Then, the link load value with the maximum load is selected and weighted summed with the total length of the route. Finally, the smallest individual is selected as the initial member of the population.
[0044] (3) Use the OX crossover operator to perform crossover operations on the population, and use the uniform mutation operator to perform mutation operations. Select individuals in the population according to the principle of survival of the fittest based on end-to-end delay. Use the taboo search algorithm to optimize the generated new individuals to prevent loops from occurring during the search process, which may lead to falling into local optimality. After a certain number of iterations, output the scheduling solution with the minimum end-to-end delay.
[0045] Advantages and beneficial effects of the present invention:
[0046] Aiming at the problem that it is difficult to guarantee the deterministic delay of time-sensitive services caused by the simultaneous access of multiple types of terminal services and the transmission of massive data in the full-time and global power communication network, the present invention proposes a multi-service deterministic delay guarantee technology for the full-time and global power communication network based on a time-sensitive network. The algorithm is based on the business model and priority division of the full-time and global power communication network, proposes a traffic scheduling mechanism based on time-sensitive network-aware shaping and establishes constraints; for the goal of minimizing end-to-end delay, a heuristic optimization algorithm based on a combination of a genetic algorithm and a taboo search algorithm is proposed to propose a time-sensitive traffic scheduling algorithm, calculate and output the gating list of time-sensitive traffic, improve scheduling performance, and guarantee the deterministic delay of time-sensitive services. BRIEF DESCRIPTION OF THE DRAWINGS
[0047] Figure 1 Schematic diagram of the end-to-end average delay of TT flows in different scheduling methods of the present invention;
[0048] Figure 2 It is a schematic diagram of the delay jitter of each priority service flow of the present invention. DETAILED DESCRIPTION
[0049] The embodiments of the present invention are further described in detail below with reference to the accompanying drawings:
[0050] The present invention proposes a method for ensuring the delay of a full-time and global power communication network based on a time-sensitive network. The method first analyzes the characteristics, importance, periodicity and delay requirements of various types of services in the full-time and global power communication network, establishes a service model and prioritizes them. On this basis, a traffic scheduling mechanism based on time-sensitive network-aware shaping is proposed. In order to ensure that the gated scheduling table calculation meets the deterministic low-latency transmission of time-sensitive flows within the transmission cycle, constraints are established on the transmission start time of each link; for the goal of minimizing end-to-end delay, a time-sensitive traffic scheduling algorithm is proposed based on a heuristic optimization algorithm combined with a genetic algorithm and a taboo search algorithm, and the gated list of the output time-sensitive traffic is calculated to improve the scheduling performance and ensure the deterministic delay of time-sensitive services.
[0051] A method for ensuring time delay of a full-time and full-domain electric power communication network based on a time-sensitive network comprises the following steps:
[0052] Step 1: Establish a full-time and global electric power communication network service model and prioritize it;
[0053] TSN (Time-Sensitive Networking) is a set of protocol standards developed by the TSN working group in the IEEE 802.1 working group. It is a protocol family that achieves deterministic minimum time delay. It defines the time-sensitive mechanism for Ethernet data transmission to ensure real-time, deterministic and reliable data transmission. There are multiple data streams in TSN, which are usually divided into TT streams and non-TT streams. TT streams are common in periodic real-time data streams. The transmission has periodic characteristics. It is usually necessary to ensure that these TT streams have deterministic low latency and jitter to improve transmission performance. Non-TT streams refer to AVB streams, best-effort streams, or non-periodic sporadic streams.
[0054] The specific steps of step 1 include:
[0055] (1) Introduce TSN technology into the full-time and full-domain power communication network to establish business models for new smart grid services, including smart grid distribution network automation services, smart grid power consumption information collection services, smart interactive marketing and smart power consumption management services, microgrids and distributed energy access;
[0056] (2) By analyzing the communication network requirements of various services, the multi-type services in the full-time and global power communication network are prioritized according to the service latency requirements and the importance of the services to the grid operation control;
[0057] (3) Using the priority field of the IEEE 802.1Q protocol, different types of data are divided into eight priorities:
[0058] Periodic real-time data is marked as the highest priority because of its high requirements for real-time and periodicity of communication;
[0059] Non-periodic real-time data requires lower communication latency and is therefore marked as the second highest priority;
[0060] The remaining priorities are marked according to traffic type and data usage;
[0061] Data of different priorities corresponds to different priority buffer queues to ensure orderly transmission of data.
[0062] In this embodiment, the main service priorities are divided as shown in Table 1:
[0063] Table 1 Priority division of all-time and all-area power communication network services
[0064]
[0065] Step 2: Based on the result of the multi-type service priority division in step 1, a traffic scheduling mechanism based on time-sensitive network-aware shaping is established, and based on the traffic scheduling mechanism, in order to ensure that the gating schedule calculation satisfies the deterministic low-latency transmission of the time-sensitive flow within the transmission cycle, a constraint condition is established on the transmission start time of each link;
[0066] In this embodiment, time-aware shaping (TAS) is introduced in the IEEE 802.1Qbv standard to optimize the transmission performance of frames of different priorities. TAS divides the communication time into equal periods and uses a gating list to determine which priority data queues can be transmitted. In a specific time slice, TAS may suspend the transmission of low-priority traffic and allow only high-priority, time-sensitive data to pass, thereby significantly reducing transmission delay and delay jitter and ensuring the timeliness of information delivery.
[0067] The gate control list contains the switch status of the gate, represented by an 8-bit binary number, each bit corresponds to a priority queue. 0 means that the buffer queue is closed, and 1 means it is open. The gate status will be updated periodically as the synchronous clock changes.
[0068] The specific steps of establishing the traffic scheduling mechanism based on time-sensitive network-aware shaping in step 2 are as follows:
[0069] (1) The topological structure of the full-time and global power communication network is modeled as G = (E, V), where E represents the set of all links in the communication network, V represents the set of all nodes in the communication network, and the service flow f i ∈F from the starting node v 1 ∈v sends data through the switching node to the final node v according to the period T 2 ;
[0070] (2) Using the TAS gating list, the communication time is divided into cycles, and each cycle is further divided into two time slices. In the first time slice, the gate state of the highest priority queue is set to open. In the second time slice, the gate state of the queues with priorities 0 to 5 is set to open, allowing non-real-time data to be transmitted. Non-periodic real-time data is not restricted by the gate state and is transmitted within the sending interval of periodic real-time data. It can also preempt low-priority non-real-time data.
[0071] (3) A guard band is set at the end of each cycle. During the guard band, no data is allowed to be sent to avoid the data transmission of the next cycle being affected by the unfinished data transmission of the previous cycle. After the highest priority periodic real-time data enters the buffer queue, it is transmitted in the first time slice according to the gate state and the reserved transmission bandwidth, and its transmission time is recorded in the flow reservation information table. When non-periodic data is transmitted, the flow reservation information table is first queried and the data is transmitted during the idle period of periodic data transmission. If the sending port is transmitting low-priority non-real-time data, the non-periodic real-time data can interrupt its transmission and be sent first.
[0072] In step 2, based on the traffic scheduling mechanism, in order to ensure that the gated scheduling table calculation satisfies the deterministic low-latency transmission of the time-sensitive flow within the transmission period, a constraint condition is established for the transmission start time on each link. The constraint condition includes:
[0073] Based on the above traffic scheduling mechanism, in order to ensure that the gating schedule calculation satisfies the deterministic low-latency transmission of time-sensitive flows within the transmission cycle, it is necessary to impose certain constraints on the transmission start time on each link:
[0074] (1) Frame constraint: This constraint ensures that the transmission of a frame on the first link of the transmission path cannot start before the send time and must be completed on the last link before the deadline. The constraint expression is:
[0075]
[0076] Among them, v j and v k Represents the second and second to last node. For flow i In (v 1 ,v 2 ) transmission offset, The flow f i The sending time and deadline, The time taken to transmit any data frame.
[0077] (2) End-to-end delay constraint: The TT flow is transmitted from the sender to the receiver along the calculated optimal path. Its end-to-end delay includes propagation delay, processing delay, and transmission delay. The expression of the delay D between one switching node and the next switching node is as follows.
[0078]
[0079] The end-to-end delay of flow fi from the sending end to the receiving end through j switches should be less than or equal to D i , the constraint expression is:
[0080]
[0081] (3) Link constraint: When two different frames on a link overlap in transmission time, the propagation delay of the two frames will increase, resulting in the indeterminacy of the transmission of the two frames. Therefore, the transmission of two different frames on a link cannot overlap in time. That is, for each pair of different frames on the same link, the transmission of one frame must be completed before the transmission of the other frame begins, and vice versa; this constraint must take into account all frames in the cycle; due to the zero jitter requirement, the link (v i ,v j The transmission offset of the kth frame of stream fi on )∈E is equal to The constraint expression is:
[0082]
[0083] Among them, T i Represents the period of the TT flow, and the cycle period of the gating list is equal to the least common multiple L of all flows.
[0084] (4) Queue constraint: This constraint ensures that each frame is correctly assigned a transmission queue on each egress port. The constraint expression is as follows:
[0085]
[0086] Among them, λ represents the queue information of frame storage, and Q represents the number of queues available on the link egress port.
[0087] Step 3: Taking the minimum end-to-end delay of time-sensitive traffic as the optimization goal, a heuristic optimization algorithm based on the genetic algorithm and the taboo search algorithm is proposed to propose a time-sensitive traffic scheduling algorithm based on the joint optimization of genetic taboo search, calculate and output the gating list of time-sensitive traffic, and output the scheduling plan with the minimum end-to-end delay;
[0088] Based on the above time-aware shaping traffic scheduling mechanism, several constraints are considered and the minimum end-to-end delay of time-sensitive traffic is adopted as the optimization goal. Considering that TSN traffic scheduling is an NP-hard problem, heuristic algorithms and other methods are needed to calculate the gating list of output time-sensitive traffic.
[0089] Among heuristic algorithms, the genetic algorithm (GA) has parallel search and calculation, strong adaptability and self-learning ability, is suitable for solving scheduling calculation problems, and has excellent global search ability. However, its local search ability is weak. On the contrary, the taboo search algorithm (TS) shows strong local search ability by avoiding repeated paths in the short term or selectively avoiding specific paths, but its global search ability is weak. In view of this, the present invention proposes a time-sensitive traffic scheduling algorithm based on the joint optimization of genetic algorithm and taboo search, combining the advantages of both to improve the overall scheduling performance.
[0090] The specific steps of step 3 include:
[0091] (2) Setting the parameters of the genetic algorithm involves population size, crossover probability, mutation probability, and maximum number of iterations.
[0092] Among them, the population size determines the number of individuals in each generation, the crossover probability determines the degree of application of the crossover operation in the population, the mutation probability affects the degree of mutation of individual genes, and the maximum number of iterations specifies the maximum number of times the algorithm is executed. The settings of these parameters directly affect the performance and convergence speed of the genetic algorithm;
[0093] With the help of the Centralized User Configuration (CUC) defined in the IEEE 802.1Qcc standard and the Centralized Network Configuration (CNC) supporting the NETCONF or RESTCONF protocol, a complete centralized management mechanism is provided. CUC is responsible for providing registration services for time-sensitive flows for key applications, collecting information such as the flow period, length, sending and receiving nodes, and maximum delay, and registering new devices and exchanging information with CNC through the User Network Interface (UNI). CNC is responsible for calculating and configuring GCL to meet the strict real-time requirements for dynamic scheduling in large-scale network models. In the dynamic scheduling process, the core scheduling solution is required to have a low time complexity to cope with the challenges brought by changes in network topology and traffic.
[0094] Whenever the traffic or network topology changes, the centralized management mechanism can promptly update algorithm-related parameters and adjust individual genes to ensure the normal iteration of the algorithm.
[0095] (2) Population initialization: The first step is to randomly initialize all feasible routes between the sender and receiver of each flow. Then, the Cartesian product of the routes selected by each flow is calculated, that is, the combination of the routes selected by each flow. The number of populations required for initialization by the algorithm is selected from them, rather than completely traversing the Cartesian product. Next, it is verified whether the generated individuals meet the constraints. If they do, the corresponding individual genes are generated; otherwise, they are discarded.
[0096] In the initialization phase of route optimization, the load data volume in each link is accumulated. Then, the link load value with the maximum load is selected and weighted summed with the total length of the route. Finally, the smallest individual is selected as the initial member of the population.
[0097] (3) Use the OX crossover operator to perform crossover operations on the population, and use the uniform mutation operator to perform mutation operations. Select individuals in the population according to the principle of survival of the fittest based on end-to-end delay. Use the taboo search algorithm to optimize the generated new individuals to prevent loops from occurring during the search process, which may lead to falling into local optimality. After a certain number of iterations, output the scheduling solution with the minimum end-to-end delay.
[0098] Figure 1 The results of the operation under the average end-to-end delay of TT flows under the influence of different global flow numbers are shown. It can be seen that the average end-to-end delay of time-sensitive flows is the lowest when the joint heuristic scheduling method is used. The genetic algorithm is 15.79% higher than the results obtained by the joint genetic taboo search algorithm on average, and the taboo search algorithm is 10.48% higher than the results of the joint genetic taboo search algorithm on average. It can be seen that the joint genetic taboo search algorithm has a significant improvement on the average end-to-end delay of traffic scheduling in TSN.
[0099] Figure 2 It shows that when the routing constraints and scheduling constraints are jointly considered in the traffic scheduling constraints, the average delay jitter of the time-sensitive flow is very small, and the delay jitter of the highest and second highest priority business flows is only about 2μs. This is because the gated scheduling gives priority to the delay jitter of the time-sensitive flow, ensuring the end-to-end deterministic transmission of important time-sensitive flows in the full-time and full-domain power communication network. However, the transmission delay jitter of the non-time-sensitive flow is larger after the gated scheduling is turned on, and the delay jitter of the lowest priority business flow reaches about 50μs. This is because the time slots of the gated scheduling are allocated to the time-sensitive flow first. After ensuring the transmission requirements of the time-sensitive flow, the remaining time slots are allocated to the non-time-sensitive flow, resulting in the transmission delay jitter of the non-time-sensitive flow cannot be guaranteed, so the delay jitter is larger.
[0100] In summary, the time-aware shaping gating scheduling calculation method based on genetic taboo search combined with the proposed invention can effectively ensure the deterministic delay of important time-sensitive services in the full-time and full-area communication network and reduce its delay jitter.
[0101] The innovation of the present invention lies in:
[0102] The present invention proposes a multi-service deterministic delay guarantee technology for a full-time and full-domain electric power communication network based on a time-sensitive network. After analyzing the characteristics, importance, periodicity and delay requirements of various types of services in the full-time and full-domain electric power communication network, a service model is established and prioritized. On this basis, a traffic scheduling mechanism based on time-sensitive network-aware shaping is proposed. In order to ensure that the calculation of the gating schedule meets the deterministic low-latency transmission of time-sensitive flows within the transmission cycle, constraints are established on the transmission start time of each link. Then, with the goal of minimizing end-to-end delay, a gating list method for calculating and outputting time-sensitive traffic is proposed based on a heuristic optimization algorithm combined with a genetic algorithm and a taboo search algorithm. The method improves scheduling performance, ensures deterministic delays for time-sensitive services, and reduces delay jitter. It is an intelligent and efficient delay guarantee technology. The key points and pre-protection points of this patent are as follows:
[0103] Step 1: Analyze the characteristics, importance, periodicity and latency requirements of various types of services in the full-time and global power communication network, establish a business model and prioritize them.
[0104] Step 2: Analyze the basic principles of time-sensitive network-aware shaping, and establish a traffic scheduling mechanism based on the business model and business priority in step 1. In order to ensure that the gating schedule calculation meets the deterministic low-latency transmission of time-sensitive flows within the transmission cycle, constraints are established on the transmission start time on each link.
[0105] Step 3: Aiming at the goal of minimizing end-to-end delay, a heuristic optimization algorithm based on the genetic algorithm and the taboo search algorithm is proposed to calculate the gated list method for outputting time-sensitive traffic, improve scheduling performance, ensure deterministic delay of time-sensitive services, and reduce delay jitter.
[0106] It should be emphasized that the embodiments of the present invention are illustrative rather than restrictive. Therefore, the present invention includes but is not limited to the embodiments described in the specific implementation modes. Any other implementation modes derived by those skilled in the art based on the technical solutions of the present invention also fall within the scope of protection of the present invention.
Claims
1. A method for ensuring time delay in a full-time and full-domain electric power communication network based on a time-sensitive network, characterized in that: The following steps are involved: Step 1: Establish a full-time and global electric power communication network service model and prioritize it; Step 2: Based on the result of the multi-type service priority division in step 1, a traffic scheduling mechanism based on time-sensitive network-aware shaping is established, and based on the traffic scheduling mechanism, in order to ensure that the gating schedule calculation satisfies the deterministic low-latency transmission of the time-sensitive flow within the transmission cycle, a constraint condition is established on the transmission start time of each link; Step 3: Taking the minimum end-to-end delay of time-sensitive traffic as the optimization goal, a heuristic optimization algorithm based on the joint optimization of genetic algorithm and taboo search algorithm is proposed. A time-sensitive traffic scheduling algorithm based on genetic taboo search joint optimization is proposed, the gating list of time-sensitive traffic is calculated and output, and the scheduling plan with the minimum end-to-end delay is output.
2. According to claim 1, a method for ensuring time delay in a full-time and full-domain electric power communication network based on a time-sensitive network is characterized in that: The specific steps of step 1 include: (1) Introduce TSN technology into the full-time and full-domain power communication network to establish business models for new smart grid services, including smart grid distribution network automation services, smart grid power consumption information collection services, smart interactive marketing and smart power consumption management services, microgrids and distributed energy access; (2) By analyzing the communication network requirements of various services, the multi-type services in the full-time and global power communication network are prioritized according to the service latency requirements and the importance of the services to the grid operation control; (3) Using the priority field of the IEEE 802.1Q protocol, different types of data are divided into eight priorities: Periodic real-time data is marked as the highest priority because of its high requirements for real-time and periodicity of communication; Non-periodic real-time data requires lower communication latency and is therefore marked as the second highest priority; The remaining priorities are marked according to traffic type and data usage; Data of different priorities corresponds to different priority buffer queues to ensure orderly transmission of data.
3. The method for ensuring time delay of a full-time and full-domain electric power communication network based on a time-sensitive network according to claim 1 is characterized by: The specific steps of establishing the traffic scheduling mechanism based on time-sensitive network-aware shaping in step 2 are as follows: (1) The topological structure of the full-time and global power communication network is modeled as G = (E, V), where E represents the set of all links in the communication network, V represents the set of all nodes in the communication network, and the service flow f i ∈F sends data from the starting node v1∈v through the switching node to the end node v2 according to the period T; (2) Using the TAS gating list, the communication time is divided into cycles, and each cycle is further divided into two time slices. In the first time slice, the gate state of the highest priority queue is set to open. In the second time slice, the gate state of the queues with priorities 0 to 5 is set to open, allowing non-real-time data to be transmitted. Non-periodic real-time data is not restricted by the gate state and is transmitted within the transmission interval of periodic real-time data. It can also preempt low-priority non-real-time data. (3) A guard band is set at the end of each cycle. During the guard band, no data is allowed to be sent to prevent the data transmission of the next cycle from being affected by the unfinished data transmission of the previous cycle. After the highest priority periodic real-time data enters the buffer queue, it is transmitted in the first time slice according to the gate status and the reserved transmission bandwidth. Its transmission time is recorded in the stream reservation information table. When non-periodic data is transmitted, the stream reservation information table is first queried and the data is transmitted during the idle period of periodic data transmission. If the sending port is transmitting low-priority non-real-time data, the non-periodic real-time data can interrupt its transmission and be sent first.
4. The method for ensuring time delay of a full-time and full-domain electric power communication network based on a time-sensitive network according to claim 1 is characterized in that: In step 2, based on the traffic scheduling mechanism, in order to ensure that the gated scheduling table calculation satisfies the deterministic low-latency transmission of the time-sensitive flow within the transmission period, a constraint condition is established for the transmission start time on each link. The constraint condition includes: (1) Frame constraint: This constraint ensures that the transmission of a frame on the first link of the transmission path cannot start before the sending time and must be completed on the last link before the deadline. The constraint expression is: Among them, v j and v k Represents the second node and the second to last node; For flow i The transfer offset at (v1,v2), The flow f i The sending time and deadline, The time taken to transmit any data frame; (2) End-to-end delay constraint: The TT flow is transmitted from the sender to the receiver along the calculated optimal path. Its end-to-end delay includes propagation delay, processing delay, and transmission delay. The expression of the delay D between one switching node and the next switching node is as follows: The end-to-end delay of flow fi from the sending end to the receiving end through j switches should be less than or equal to Di. The constraint expression is: (3) Link constraint: When two different frames on a link overlap in transmission time, the propagation delay of the two frames will increase, resulting in the non-deterministic transmission of the two frames. Therefore, the transmission of two different frames on a link cannot overlap in time. That is, for each pair of different frames on the same link, the transmission of one frame must be completed before the transmission of the other frame begins, and vice versa. This constraint must take into account all frames in the cycle. Due to the zero jitter requirement, the link (v i ,v j The transmission offset of the kth frame of stream fi on )∈E is equal to The constraint expression is: Among them, Ti represents the period of TT flow, and the cycle period of the gating list is equal to the least common multiple L of all flows; (4) Queue constraint: This constraint ensures that each frame is correctly assigned a transmission queue on each egress port. The constraint expression is as follows: Among them, λ represents the queue information of frame storage, and Q represents the number of queues available on the link egress port.
5. The method for ensuring time delay of a full-time and full-domain electric power communication network based on a time-sensitive network according to claim 1 is characterized in that: The specific steps of step 3 include: (1) Setting the parameters of the genetic algorithm involves population size, crossover probability, mutation probability, and maximum number of iterations; With the help of the centralized user configuration CUC defined in the IEEE 802.1Qcc standard and the centralized network configuration CNC supporting the NETCONF or RESTCONF protocol, a complete centralized management mechanism is provided; CUC is responsible for providing registration services for time-sensitive flows for key applications, collecting information such as the flow period, length, sending and receiving nodes, and maximum delay, and registering new devices and exchanging information with CNC through the user network interface UNI; CNC is responsible for calculating and configuring GCL to cope with the strict real-time requirements for dynamic scheduling in large-scale network models; in the dynamic scheduling process, the core scheduling solution is required to have a low time complexity to cope with the challenges brought by changes in network topology and traffic; Whenever traffic or network topology changes, the centralized management mechanism can promptly update algorithm-related parameters and adjust individual genes to ensure the normal iteration of the algorithm; (2) Population initialization: The first step is to randomly initialize all feasible routes between the sender and receiver of each flow. Then, the Cartesian product of the routes selected by each flow is calculated, that is, the combination of the routes selected by each flow. The number of populations required for the algorithm to be initialized is selected from them, rather than completely traversing the Cartesian product. Next, it is verified whether the generated individuals meet the constraints. If they do, the corresponding individual genes are generated. Otherwise, they are discarded. In the initialization phase of route optimization, the load data volume in each link is accumulated; then, the load value of the link with the maximum load is selected and weighted summed with the total length of the route; finally, the smallest individual is selected as the initial member of the population; (3) Use the OX crossover operator to perform crossover operations on the population, and use the uniform mutation operator to perform mutation operations; select individuals in the population according to the principle of survival of the fittest based on end-to-end delay; use the taboo search algorithm to optimize the generated new individuals to prevent loops from occurring during the search process, which may lead to falling into local optimality; after a certain number of iterations, output the scheduling plan with the minimum end-to-end delay.
Citation Information
Cited By
Online reconfiguration method for time-sensitive network
CN119865423A
Wireless deterministic traffic access control method and device, equipment and storage medium
CN119967602A
Hybrid intelligent driven multi-service flow deterministic scheduling method and related equipment
CN121462514A
Dynamic data packet scheduling method and system
CN122002590A
Dynamic priority preemption and time-sensitive network fused low-delay networking method and device
CN122137803A
