A time-sensitive network traffic routing and scheduling method based on centralized control

Abstract

The application belongs to the technical field of network transmission, and particularly relates to a time-sensitive network traffic routing and scheduling method based on centralized control; comprising: a CNC constructs a network topology and generates global topology information; a CUC collects demand information of each to-be-transmitted traffic from each source node and divides traffic priority according to the demand information; the CNC generates an optimal transmission path for each traffic according to the global topology information; the CNC allocates a transmission time slot for each traffic and dynamically adjusts the transmission time slot; the optimal transmission path and the transmission time slot are synchronously allocated and the transmission time slot is optimized and the transmission time slot jitter is controlled; a scheduling table is generated according to the final allocation result of the path and the time slot, and the CNC distributes the scheduling table to each network device; when the network state changes, the scheduling table is dynamically adjusted to realize real-time routing and scheduling of the traffic; the application realizes dynamic management of traffic priority, enhances the utilization rate of network resources, and guarantees the service quality and high availability of transmission.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of network transmission, and particularly relates to a time-sensitive network traffic routing and scheduling method based on centralized control. BACKGROUND

[0002] With the rapid development of intelligent manufacturing and industrial Internet of Things, modern industrial systems have higher requirements for the real-time performance, reliability and resource utilization of networks. Especially in critical tasks that require precise timing control, Time Sensitive Network (TSN) has gradually become one of the important technologies to meet the needs of industrial control. TSN realizes low-delay and high-reliability transmission of critical traffic by introducing high-precision time scheduling and deterministic routing, thereby supporting the strict requirements of industrial applications for deterministic communication. Traditional TSN scheduling methods are mostly based on distributed control. In complex network scenarios and high-density critical traffic environments, distributed methods have limitations in resource utilization and scheduling flexibility. In the face of uncertain traffic changes and multi-priority service requirements, distributed scheduling is difficult to quickly adjust resource allocation, which can easily cause uneven utilization of network resources, thereby increasing delay and wasting bandwidth. Therefore, more and more researches have begun to focus on centralized control models in order to achieve efficient utilization of resources in complex network topologies. Therefore, there is an urgent need for a centralized routing and scheduling method that can ensure deterministic transmission of critical traffic and effectively improve the utilization efficiency of remaining resources in the network. SUMMARY

[0003] In view of the deficiencies of the prior art, the application provides a time-sensitive network traffic routing and scheduling method based on centralized control, which comprises the following steps:

[0004] S1: The CNC constructs a network topology according to the current network state and generates global topology information;

[0005] S2: The CUC collects the demand information of each to-be-transmitted traffic from each source node, and divides the traffic priority according to the demand information, including high-priority traffic, medium-priority traffic and low-priority traffic. The priority division of traffic is based on bandwidth demand and delay demand. The high-priority traffic has the most stringent delay demand, the delay requirement of the medium-priority traffic is relatively relaxed, and the low-priority traffic tolerates higher delay and jitter.

[0006] S3: The CNC generates the optimal transmission path for each traffic according to the global topology information. The high-priority traffic is given priority consideration in the path selection process to ensure the minimum transmission delay. The CNC will select the best path for each priority traffic according to factors such as link bandwidth, delay, load state, etc. In the path selection, the path of the high-priority traffic will avoid links with high load and high packet loss rate.

[0007] S4: The CNC allocates transmission slots for each flow and dynamically adjusts the transmission slots. High-priority flows will have priority in obtaining slots, ensuring that their low-latency requirements are met; low-priority flows' slots will be allocated according to remaining resources, which may tolerate higher latency.

[0008] S5: Synchronize the allocation of optimal transmission paths and transmission slots, and optimize the transmission latency and jitter, complete the final allocation of paths and slots; bandwidth utilization and packet loss rate will be considered, to ensure efficient transmission of flows, avoid delay increase due to insufficient bandwidth or high packet loss rate.

[0009] S6: Generate a scheduling table according to the final allocation of paths and slots, and distribute the scheduling table to each network device; during the generation of the scheduling table, the CNC will refer to the bandwidth utilization and packet loss rate to optimize the scheduling to improve the efficiency of network resource use.

[0010] S7: Monitor the network state, when the network state changes, the CNC will dynamically adjust the scheduling table according to the bandwidth utilization, packet loss rate and port state of the link, and preferentially adjust the transmission path and slot of high-priority flow, to ensure the stable transmission of critical flow.

[0011] Preferably, the process of generating global topology information includes:

[0012] S11: The CNC scans all nodes and links in the network to generate a topology structure diagram;

[0013] S12: The CNC records the basic properties of each link, including bandwidth, delay, load state, packet loss rate and port state;

[0014] S13: The CNC sets a fixed time interval to refresh the link state to obtain the latest link information.

[0015] Preferably, the process of dividing flow priority according to demand information includes:

[0016] Demand information includes bandwidth demand BW req and latency requirement D req ;

[0017] According to the bandwidth demand BW req , the latency requirement D req , the priority of the flow is divided: high-priority flow satisfies D req ≤D th and BW req ≥BW th , medium-priority flow is D th <D req ≤D max, low priority traffic meets D req >D max ; BW th represents bandwidth requirement threshold, D th represents delay requirement threshold, D max represents maximum delay requirement.

[0018] Preferably, the process of generating optimal transmission path for each traffic includes:

[0019] S31: generating candidate path set for each traffic according to topology graph, each candidate path meets bandwidth requirement BW req and delay requirement D req of the traffic;

[0020] S32: eliminating candidate paths in the candidate path set that do not meet link load and packet loss rate;

[0021] S33: calculating fitness and finding out optimal transmission path from the candidate path set according to fitness by using PSO algorithm.

[0022] Further, the formula of calculating fitness is:

[0023] F path = a · L total (P) + b · D total (P)

[0024] wherein, F path represents fitness, L total (P) represents total load of candidate path set P,

[0025] D total (P) represents total delay of candidate path set P; a, b represent load weight and delay weight.

[0026] Preferably, the process of CNC allocating transmission gap for each traffic and dynamically adjusting transmission time slot includes:

[0027] S41: CNC allocates initial transmission time slot for each priority traffic; high priority traffic has strictest delay requirement, low priority traffic can tolerate higher delay;

[0028] S42: optimizing time slot offset of each traffic from high to low according to priority order by using SAC algorithm, obtaining optimal time slot offset; adjusting transmission time slot according to optimal time slot offset, obtaining transmission gap allocated for each traffic;

[0029] S43: Calculate the score of each flow according to the delay requirement, bandwidth requirement and jitter tolerance; sort the flows according to the scores; and allocate the time slots to the flows in order of priority to ensure that high-priority flows are given low-delay transmission resources first.

[0030] Preferably, the process of synchronously allocating the optimal transmission path and transmission time slot and optimizing the transmission delay and jitter includes:

[0031] S51: Synchronously allocate the optimal transmission path and transmission time slot of each flow in order of priority from high to low to ensure that the transmission time slot of each flow is arranged within a suitable time window.

[0032] S52: Optimize the transmission time slot according to the delay requirement of the flow, and reduce the transmission delay of the critical flow by dynamically adjusting the position and length of the time slot.

[0033] S53: Control the transmission time slot jitter T jitter to meet the transmission requirement target Minimize T jitter = |T slot -T req |, T slot represents the transmission time slot, and T req represents the required time slot.

[0034] Preferably, the time-sensitive network flow routing and scheduling method further includes: when a link or node fails, the CNC recalculates a backup path; and the formula for calculating the backup path is:

[0035]

[0036] wherein P new represents the backup path, L link (i) represents the load of path i, D limit (i) represents the delay of path i, and F usa represents the non-link failure set.

[0037] The beneficial effects of the present application are: the present application realizes the global optimization configuration of traffic in the centralized control mode by introducing the particle swarm optimization (PSO) and soft actor critic (SAC) algorithms. The centralized user configuration manager (CUC) is responsible for collecting and managing the demand information of each service traffic, including bandwidth, delay and jitter tolerance, and classifying the traffic according to priority. The central network controller (CNC) generates a candidate path set for each type of traffic based on the real-time updated network topology information and link load, and selects the optimal path in the candidate path using the PSO algorithm. The SAC algorithm further optimizes the time slot allocation of the key traffic, ensuring that the high-priority traffic is transmitted in the low-delay time slot. Combined with the time-aware scheduling (QBV) and time-sensitive scheduling (TAS) mechanisms, the delay and jitter are reduced in the synchronous optimization of path and time slot, and the transmission reliability of the key traffic is ensured. The present application realizes the dynamic management of traffic priority, enhances the utilization of network resources, and guarantees the service quality and high availability of transmission. BRIEF DESCRIPTION OF DRAWINGS

[0038] Figure 1 Flow chart of the centralized control-based time-sensitive network traffic routing and scheduling method of the present application;

[0039] Figure 2 Schematic diagram of the network topology structure of the present application;

[0040] Figure 3 Experience playback and feedback mechanism diagram of the TSN network routing and scheduling of the present application. DETAILED DESCRIPTION

[0041] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0042] The present application proposes a centralized control-based time-sensitive network traffic routing and scheduling method, aiming to solve the contradiction between the deterministic transmission demand of key service flow and the efficient utilization of network resources. The method combines the particle swarm optimization (PSO) algorithm and the soft actor critic (SAC) algorithm to realize global path selection and time slot scheduling optimization, ensuring low-delay and high-reliability transmission of high-priority traffic while effectively improving the utilization of remaining network resources; as shown in Figure 1 The method comprises:

[0043] S1: The CNC constructs the network topology according to the current network state and generates global topology information.

[0044] Topology of CNC construction network, CNC and CUC collaborative working mode as shown in Figure 2 The CNC (Central Network Controller) systematically scans all nodes and links in the entire network, and constructs a global topology graph G=(V, E) by obtaining the connection state and link characteristics of the device, where V is the node set in the network, including switches, routers and other network devices, and E is the link set.

[0045] The CNC collects and records the basic attributes of each link, including bandwidth B link (e ij ) the maximum transmission capacity of the link, delay D link (e ij ) the data transmission delay from one end of the link to the other end, load state L usage (e ij ) the current usage of the link, the packet loss rate, i.e. the ratio of the number of lost data packets to the transmitted data group, and the port state S port (e ij ) records the connectivity and running state of the port.

[0046] The CNC sets a fixed time interval to refresh the link state to obtain the latest link information, and real-time feedback of the state change of the link, and the updated information will be fed back to the central network controller (CNC) and related network devices, so as to make dynamic routing adjustment and time slot allocation according to the current network state.

[0047] S2: CUC collects the demand information of each to-be-transmitted traffic from each source node, and divides the traffic priority according to the demand information, including high-priority traffic, medium-priority traffic and low-priority traffic.

[0048] The priority division of the traffic is based on bandwidth demand and delay demand. The high-priority traffic has the most stringent delay demand, and the bandwidth demand is relatively high, so it needs to be allocated to links and time slots with sufficient bandwidth in priority; the delay requirement of the medium-priority traffic is relatively loose, and the bandwidth demand is moderate; the low-priority traffic tolerates higher delay, but its bandwidth demand can be appropriately reduced.

[0049] CUC collects the demand information of each to-be-transmitted traffic from each source node, including bandwidth demand BW req and delay demand D req . According to these demands, CUC divides the traffic into high-priority, medium-priority and low-priority; specifically: high-priority traffic satisfies D req ≤D th and BW req ≥BW th , medium-priority traffic is D th <D req ≤D maxLow-priority traffic satisfies D. req >D max BW th D represents the bandwidth requirement threshold. th D represents the latency requirement threshold. max This indicates the maximum latency requirement.

[0050] S3: CNC generates the optimal transmission path for each traffic based on global topology information.

[0051] Consider the link's bandwidth utilization, bandwidth requirements, and packet loss rate. CNC prioritizes links with sufficient bandwidth and low packet loss rates to ensure that traffic with high bandwidth requirements can obtain the necessary resources and optimize traffic transmission quality.

[0052] S31: Generate a set of candidate paths P for each flow based on the topology graph G = (V, E).

[0053] {p1,p2,...,p n Each candidate path satisfies the bandwidth requirement BW of the traffic. req and delayed demand D req .

[0054] S32: Eliminate candidate paths from the candidate path set that do not meet the link load and packet loss rate requirements.

[0055] S33: Calculate the fitness and use the PSO algorithm to find the optimal transmission path from the candidate path set based on the fitness.

[0056] The CNC employs PSO and SAC algorithms to optimize paths and time slots respectively, achieving joint optimization. Within the candidate path set, the CNC uses the Particle Swarm Optimization (PSO) algorithm for path selection. The PSO algorithm treats each candidate path as a "particle," and its fitness function is:

[0057] F path =α·L total (P)+β·D total (P)

[0058] Among them, L total (P)=∑ i∈p L link (i) represents the total load of the path, D total (P)=∑ i∈p D link (i) represents the total path delay, L link (i) represents the load of link i, D link (i) represents the delay of link i; α and β represent the load weight and delay weight, respectively.

[0059] Each particle is based on its individual optimal value P.best and global optimal value g best The speed and position are updated as follows:

[0060]

[0061] where, is the speed of the particle at iteration t, is the specific solution or path of the current particle in the path selection space, ω is the inertia weight, c1 and c2 are learning factors, and r1 and r2 are random numbers. After multiple iterations, the algorithm finally converges to the optimal path.

[0062] According to the fitness function results of the PSO optimization, the candidate path set P cand is sorted, and the path P opt with the optimal fitness is selected as the optimal transmission path. p∈pcand F path .

[0063] S4: The CNC allocates transmission slots for each traffic according to the traffic priority and dynamically adjusts the transmission slots.

[0064] S41: The CNC allocates an initial transmission slot T slot for each priority traffic.

[0065] The delay requirement D req , jitter tolerance J tol and bandwidth requirement BW req of the traffic are considered. High-priority traffic will be given low-latency slots to ensure that its transmission delay requirement is met, and appropriate slot length will be allocated according to the bandwidth requirement. Traffic with low jitter tolerance will also be given priority in slot allocation to reduce the impact of slot jitter. The slot allocation for low-priority traffic will be based on the remaining resources and may tolerate higher latency and jitter.

[0066] S42: The SAC algorithm is used to optimize the slot offset for each traffic in priority order from high to low, and the optimal slot offset is obtained; the transmission slots are adjusted according to the optimal slot offset to obtain the transmission gap allocated for each traffic.

[0067] The SAC algorithm optimizes the transmission slots through reinforcement learning to minimize the delay. In SAC, the actor network calculates the action a t , i.e. the slot offset, according to the current state s t ; the critic network estimates the value of the action and gives the corresponding reward. The reward design takes into account factors such as delay, bandwidth utilization and jitter,

[0068] After optimization, the state, action, reward, and other information are stored in the experience replay pool. The loss function is updated The allocation of each time slot is optimized so that the overall performance of the network is improved.

[0069] S43: Calculate the score of each traffic, which is evaluated by factors such as latency requirement, bandwidth requirement, and jitter tolerance. According to the score, the traffic is sorted, and the sorted traffic is allocated time slots in order of priority, ensuring that high-priority traffic is given low-latency transmission resources first, and low-priority traffic can be delayed appropriately in high-load situations.

[0070] S5: Synchronize the allocation of optimal transmission paths and transmission time slots, and optimize the transmission time slots and control the transmission time slot jitter to complete the final allocation of paths and time slots.

[0071] As shown in Figure 3 , the system realizes dynamic adjustment of paths and time slots through a multi-layer feedback mechanism combined with QBV and TAS strategies.

[0072] S51: Combine the time-aware scheduling (QBV) mechanism to synchronize the allocation of optimal transmission paths and transmission time slots for each traffic in order of priority from high to low, to ensure that each traffic is arranged in the appropriate time window, and then to ensure that the transmission time slot matches the demand time slot, which is the target value of the ideal transmission time slot of the traffic, to ensure that the traffic completes data transmission on time when the network load allows.

[0073] S52: Optimize the transmission time slot according to the latency requirement of the traffic, and reduce the transmission delay of critical traffic by dynamically adjusting the position and length of the time slot.

[0074] For traffic with low latency requirement, allocate earlier time slots and minimize overlap with other traffic; for traffic with high latency tolerance, allocate later time slots. Use SAC to optimize time slot offset to further reduce latency and improve traffic transmission efficiency.

[0075] S53: Control the transmission time slot jitter T jitter to meet the transmission requirement target Minimize T jitter = |T slot -T req |, T slot represents the transmission time slot, and T req represents the demand time slot.

[0076] S6: Generate a scheduling table according to the final allocation of paths and time slots, and distribute the scheduling table to each network device by CNC.

[0077] In the process of generating the schedule table, the CNC will refer to the bandwidth utilization and packet loss rate to optimize the scheduling to improve the efficiency of network resource utilization.

[0078] After completing the path and time slot optimization, the system generates a schedule table S sched , including the best path P opt , priority P k and corresponding time slot T slot for each flow; each item in the schedule table is S i ={P opt , T slot, P k}, indicating the transmission of flow F i on path P opt and time slot T slot .

[0079] S7: Monitor the network state, when the network state changes, the CNC will dynamically adjust the schedule table according to the bandwidth utilization, packet loss rate and port state of the link, etc. Real-time routing and scheduling of traffic is achieved. The transmission path and time slot of high-priority traffic are adjusted first to ensure the stable transmission of critical traffic.

[0080] The CNC and CUC monitor the link load L usage (e ij ), topology G and flow changes F of the network in real time. When the network state changes significantly (such as link failure, new flow joining, etc.), the CNC and CUC will repeat steps S1-S6 according to the latest network state to recalculate the optimal transmission path and time slot allocation scheme. The CUC can extract historical experience from the experience replay pool to optimize the time slot offset to reduce delay and jitter and quickly respond to the time slot optimization needs of traffic. At the same time, the bandwidth demand BW adjusted =

[0081] f(L ratio , Q alloc ) will be recalculated, and the CNC will dynamically adjust the path selection and priority allocation based on the bandwidth demand, link load and time slot requirement to ensure that high-priority traffic gets enough network resources and efficiently uses the remaining resources.

[0082] The time-sensitive network traffic routing and scheduling method further comprises: when a link or node fails, the CNC recalculates a backup path; when calculating the backup path, ensure that the backup path can meet the bandwidth demand and its performance is not lower than the current path. The formula for calculating the backup path is:

[0083]

[0084] where P new represents the backup path, L link(i) represents the load of path i, D limit F represents the delay of path i, F set F represents the set of link failures set E represents the set of links, F usa E-F set F represents the set of non-link failures usa F represents the set of non-link failures

[0085] In summary, by introducing particle swarm optimization (PSO) and soft actor critic (SAC) algorithms, the global optimization configuration of traffic is realized in the centralized control mode. The centralized user configuration manager (CUC) is responsible for collecting and managing the demand information of each service traffic, including bandwidth, delay and jitter tolerance, and classifying the traffic according to priority. The central network controller (CNC) generates a candidate path set for each type of traffic based on the real-time updated network topology information and link load, and uses the PSO algorithm to select the appropriate optimal path in the candidate path. The SAC algorithm further optimizes the time slot allocation of critical traffic to ensure that high-priority traffic is transmitted in low-delay time slots. Combined with the time-aware scheduling (QBV) and time-sensitive scheduling (TAS) mechanisms, the delay and jitter are reduced in the synchronous optimization of path and time slot, ensuring the transmission reliability of critical traffic. The invention realizes the dynamic management of traffic priority, enhances the utilization of network resources, and guarantees the service quality and high availability of transmission.

[0086] The above examples further illustrate the purpose, technical solutions and advantages of the present application. It should be understood that the above examples are only preferred embodiments of the present application and do not limit the present application. Any modification, equivalent replacement, improvement, etc. made to the present application within the spirit and principles of the present application should be included in the protection scope of the present application.

Claims

1. A centralized control based time sensitive network traffic routing and scheduling method, characterized in that, Comprise: S1: CNC constructs network topology according to current network state and generates global topology information; S2: CUC collects demand information of each to-be-transmitted traffic from each source node, and divides traffic priority according to demand information, including high-priority traffic, medium-priority traffic and low-priority traffic; The process of dividing traffic priority according to demand information comprises: The demand information includes a bandwidth demand BW req and a delay demand D req ; According to bandwidth requirement BW req , delay requirement D req , traffic is prioritized: high priority traffic meets D req ≤D th and BW req ≥BW th , medium priority traffic is D th <D req ≤D max , and low priority traffic meets D req >D max ; BW th denotes bandwidth requirement threshold, D th denotes delay requirement threshold, and D max denotes maximum delay requirement; S3: CNC generates optimal transmission path for each traffic according to global topology information; the process of generating optimal transmission path for each traffic comprises: S31: generating a candidate path set for each flow according to the topology graph, each candidate path satisfying the bandwidth requirement BW and the delay requirement D of the flow req and the delay requirement D req ; S32: eliminating candidate paths in candidate path set that do not meet link load and packet loss rate; S33: calculating fitness and finding out optimal transmission path from candidate path set according to fitness by using PSO algorithm; the formula of calculating fitness is: F path = a · L total (P) + b · D total (P) where F path represents fitness, L total (P) represents total load of paths in candidate path set P, D total (P) represents total delay of paths in candidate path set P; and α, β represent load weight and delay weight. S4: CNC allocates transmission time slot for each traffic according to traffic priority and dynamically adjusts transmission time slot; the process of CNC allocating transmission time slot for each traffic and dynamically adjusting transmission time slot comprises: S41: CNC allocates initial transmission time slot for each priority traffic; high-priority traffic has the most stringent delay requirement, and low-priority traffic can tolerate higher delay; S42: optimizing time slot offset of each traffic from high to low according to priority order by using SAC algorithm, to obtain optimal time slot offset; adjusting transmission time slot according to optimal time slot offset, to obtain transmission gap allocated for each traffic; S43: calculating score of each traffic according to delay requirement, bandwidth requirement and jitter tolerance; sorting traffic according to score; the sorted traffic will be allocated time slot according to priority order in turn, to ensure that high-priority traffic obtains low-delay transmission resource in priority; S5: synchronously allocating optimal transmission path and transmission time slot, and optimizing transmission time slot and controlling transmission time slot jitter, to complete final allocation of path and time slot; the process of synchronously allocating optimal transmission path and transmission time slot, and optimizing transmission time delay and jitter comprises: S51: synchronously allocating optimal transmission path and transmission time slot of each traffic from high to low according to priority order, to ensure that transmission time slot of each traffic is arranged in appropriate time window; S52: optimizing transmission time slot according to delay requirement of traffic, to reduce transmission delay of critical traffic by dynamically adjusting position and length of time slot; S53: Control the transmission time slot jitter T jitter such that it meets the transmission demand target Minimize T jitter = |T slot -T req |, T slot denotes the transmission time slot, T req denotes the demand time slot; S6: generating scheduling table according to final allocation result of path and time slot, and CNC distributes scheduling table to each network device; S7: monitoring network state, and CNC dynamically adjusts scheduling table when network state changes, to realize real-time routing and scheduling of traffic; The time-sensitive network traffic routing and scheduling method further comprises: when link or node fails, CNC recalculates backup path; the formula of calculating backup path is: where P new represents a backup path, L link (i) represents the load of path i, D limit (i) represents the delay of path i, F usa represents the set of non-link failures.

2. The centralized control based time sensitive network traffic routing and scheduling method of claim 1, wherein, The process of generating global topology information comprises: S11: CNC scans all nodes and links in network, to generate topology structure diagram; S12: CNC records basic attributes of each link, including bandwidth, delay, load state, packet loss rate and port state; S13: CNC sets fixed time interval to refresh link state, to obtain latest link information.

Images