TSN-5G high-speed rail vehicle-mounted network traffic scheduling method based on enhanced immune algorithm

Abstract

The invention provides a TSN-5G high-speed rail vehicle-mounted network traffic scheduling method based on an enhanced immune algorithm. The method comprises the following steps: acquiring all possible routes of a traffic set to be scheduled in a TSN-5G high-speed rail vehicle-mounted network and the time delay of each route; various operators of the enhanced immune algorithm are set, a DQN network is used as a clone mutation operator of the enhanced immune algorithm, and time delay and load are comprehensively considered to serve as an optimization target of the enhanced immune algorithm; the enhanced immune algorithm utilizes various operators to iteratively obtain the sending sequence and routing of each flow and the sending time of each flow in a TSN switch; and obtaining a gating scheduling list of the TSN switch according to the sending time of each flow, and performing end-to-end deterministic transmission on each flow by the TSN switch according to the gating scheduling list. According to the method, the immune algorithm and reinforcement learning are combined, the defects of an existing scheduling method are overcome, the scheduling efficiency and flexibility are improved, and deterministic scheduling of the key traffic in the TSN-5G train on-board network is achieved.

Description

Technical Field

[0001] The present invention relates to the field of vehicle-mounted network technology, and in particular to a TSN-5G high-speed rail vehicle-mounted network traffic scheduling method based on an enhanced immune algorithm. Background Art

[0002] Currently, due to the complex deployment environment of on-board terminals, there are many influencing factors such as electromagnetic interference, dust and oil pollution, vehicle vibration and signal blocking, which makes the on-board 5G wireless network have strong uncertainty problems such as node failure and link failure. In the TSN (Time-Sensitive Networking)-5G train on-board network, a traffic scheduling algorithm that can respond quickly and adjust dynamically is urgently needed to ensure the deterministic transmission of key business flows.

[0003] Many scholars at home and abroad have conducted research on TSN-5G network traffic scheduling. One solution formulated the specification of the joint message segmentation and scheduling problem through the optimization module theory, and used the existing solver to find the optimal solution. The study also proposed a heuristic algorithm to construct a low-latency schedule through the worst-case delay analysis, which significantly improved the scheduling performance.

[0004] In order to solve the routing and scheduling problems of dynamic traffic, a solution has been proposed to establish a joint optimization model for routing and scheduling of TSN-5G networks, and an online heuristic algorithm has been proposed. This solution considers both transmission delay and network load factors to determine the route, and uses two trimming operations to quickly determine the scheduling time, thus ensuring the transmission performance of real-time traffic in a dynamically changing network environment.

[0005] Aiming at the no-wait packet scheduling problem of TSN-5G network, a scheme is proposed to map it to the no-wait job shop scheduling problem (NW-JSP), and a heuristic optimization method based on taboo search is proposed. In addition, the proposed scheduling compression technology effectively reduces the number of guard bands and improves bandwidth utilization.

[0006] There is a scheme to introduce the ant colony optimization algorithm into the TSN-5G network, and propose an improved ant colony optimization algorithm as a scheduling algorithm for time-triggered flows. Simulation results show that the algorithm meets the strict end-to-end delay requirements of industrial applications, and is superior to the traditional ant colony algorithm in terms of convergence speed and optimization capability.

[0007] In the traffic scheduling algorithm of TSN-5G network, a solution found that when TSN flow in 5G-TSN integrated network is transmitted in 5G bridge, it will compete with 5G eMBB flow for time slot and channel resources, which will affect network throughput and end-to-end delay. Therefore, a transmission solution is proposed to calculate the priority of TSN flow and eMBB flow according to parameters such as delay and packet size, and allocate resources according to priority.

[0008] There is a scheme to use Network Function Virtualization (NFV) technology to realize QoS-aware mapping and scheduling in 5G-TSN networks. First, an incremental greedy algorithm is designed to map NFV to 5G and TSN resources, and a 5G resource scheduling scheme based on dynamic priority preemption is proposed to provide wait-free transmission for higher priority applications.

[0009] The disadvantages of the optimization theory algorithm of the TSN-5G network in the above-mentioned prior art include: although this method can provide a theoretically optimal solution, it faces multiple challenges in the TSN-5G vehicle network scenario, such as difficulty in formulating constraints, long solution time, and is not suitable for dynamic environments.

[0010] The disadvantages of the heuristic algorithm-based traffic scheduling algorithm of the TSN-5G network in the above-mentioned prior art include: this method has some inherent defects, such as difficulty in ensuring global optimality, strong parameter dependence, and slow convergence speed. Summary of the invention

[0011] An embodiment of the present invention provides a TSN-5G high-speed rail on-board network traffic scheduling method based on an enhanced immune algorithm to achieve efficient transmission of traffic in the TSN-5G high-speed rail on-board network.

[0012] In order to achieve the above object, the present invention adopts the following technical scheme.

[0013] A TSN-5G high-speed rail vehicle network traffic scheduling method based on enhanced immune algorithm, comprising:

[0014] Obtain all possible routes for the set of traffic to be scheduled in the TSN-5G high-speed rail vehicle network, as well as the latency of each route;

[0015] Set various operators of the enhanced immune algorithm, use the DQN network as the clone mutation operator of the enhanced immune algorithm, and take the minimum weighted sum of end-to-end delay, load balancing index and routing hops as the optimization goal of the enhanced immune algorithm;

[0016] All possible routes of the set of traffic to be scheduled, the delay of each route and the network information of the TSN-5G high-speed rail vehicle network are input into the enhanced immune algorithm, and the enhanced immune algorithm uses various operators to iteratively obtain the sending order and route of each traffic, as well as the sending time of each traffic on the time-sensitive network TSN switch;

[0017] A gating scheduling list of the TSN switch is obtained according to the sending time of each flow, and the TSN switch performs end-to-end deterministic transmission for each flow according to the gating scheduling list.

[0018] Preferably, the acquisition of all possible routes of the set of traffic to be scheduled in the TSN-5G high-speed rail vehicle network and the delay of each route includes:

[0019] Assume the flow rate f i There are n possible routes in the TSN-5G high-speed rail vehicle network, and the route set is recorded as R, r ij (j∈[1,2,...,n]) is a route in n, from the sender Passing through k intermediate nodes Arrival at the receiving end Then the route r ij Expressed as

[0020] Set the end-to-end delay of each route, including TSN network delay and 5G network delay, and the traffic f i The delay of selecting the jth route in the TSN network It consists of 4 parts: sending delay Transmission delay Processing delay and queuing delay Latency in TSN Networks The calculation formula is:

[0021]

[0022] Flow rate i Latency in 5G Networks It consists of 4 parts: sending delay Propagation Delay Processing delay and queuing delay Latency in 5G Networks The calculation formula is:

[0023]

[0024] Flow rate i The end-to-end delay d of selecting the jth route ij for:

[0025]

[0026] Flow rate i The total queuing delay for selecting the jth route is:

[0027]

[0028] Queuing delay in 5G networks TTI is the transmission time interval;

[0029] Queuing delay in TSN network k is the flow rate f i Select the number of TSN switches that the jth route passes through.

[0030] Preferably, the various operators of the enhanced immune algorithm are set, the DQN network is used as the cloning mutation operator of the enhanced immune algorithm, and the weighted sum of the end-to-end delay, the load balancing index and the number of routing hops is minimized as the optimization goal of the enhanced immune algorithm, including:

[0031] The operators set for the enhanced immune algorithm include: affinity calculation operator, antibody concentration calculation operator, incentive calculation operator, immune selection operator, DQN-based clonal mutation operator, and population refresh operator;

[0032] (8) Population initialization

[0033] The antibodies in population P are coded discretely, and each antibody p i Record a possible sending order s of the set of flows to be scheduled F i and routing number r i ;

[0034] (9) Affinity evaluation operator

[0035] Calculate the end-to-end delay D of the xth antibody according to the sending order and routing number of the antibody records x , load balancing index V x And the number of routing hops H x , the affinity aff(x) is the optimization target. The algorithm goal is to continuously iterate and find an antibody with the smallest affinity. The affinity calculation formula is:

[0036] aff(x)=ln(D x )+ln(V x )+ln(H x )

[0037] (10) Antibody concentration evaluation operator

[0038] Define the order of sending traffic within two antibodies only when s i With routing number r i If they are the same, the two antibodies are considered to be the same, and the antibody concentration den(x) of the x-th antibody is defined as:

[0039]

[0040] (11) Excitation degree calculation operator

[0041] The excitation degree sim(x) of the x-th antibody is:

[0042] sim(x)=w·aff(x)+(1-w)·den(x)

[0043] (12) Immune selection operator

[0044] Define the immune selection operator to select the top three antibodies with the highest degree of motivation;

[0045] (13) Population Refresh Operator

[0046] The population refresh operator randomly constructs new antibodies to complete the task.

[0047] (14) DQN-based cloning mutation operator

[0048] Use the DQN algorithm in reinforcement learning as the cloning and mutation operator in the immune algorithm;

[0049] State space s: the current population’s incentive distribution and the vector representation of antibodies;

[0050] Action space a: defines the optional number of clones, mutation strategies, and mutation probabilities;

[0051] The number of clones is 1 to Z, Z = S / N, S is the number of antibodies to be cloned, and N is the population size;

[0052] There are three mutation strategies:

[0053] (4) No mutation is done.

[0054] (5) Change the traffic scheduling order according to the mutation probability.

[0055] (6) Change traffic routing according to mutation probability.

[0056] Reward function r: If the overall affinity of the population is improved after cloning and mutation, a positive reward is given, otherwise a negative reward is given.

[0057] Preferably, the optimization goal of the enhanced immune algorithm is to minimize the weighted sum of the end-to-end delay, the load balancing index and the number of routing hops, including:

[0058] Assume the flow rate f i The load of the switch when the jth route is selected is x ij , the routing hop count is |r ij |, define the end-to-end delay of the entire set of traffic to be scheduled F as D = min(d ij ), the load balancing index is Routing hop count The optimization goal of the enhanced immune algorithm is set to min(ln(D)+ln(V)+ln(H)), that is, the weighted sum of the end-to-end delay, load balancing index and routing hops is minimized.

[0059] Preferably, all possible routes of the set of traffic to be scheduled, the delay of each route and the network information of the TSN-5G high-speed rail vehicle network are input into the enhanced immune algorithm, and the enhanced immune algorithm uses various operators to iteratively obtain the sending order and route of each traffic, as well as the sending time of each traffic on the TSN switch, including:

[0060] The current TSN-5G vehicle network switch bandwidth, network topology and flow information to be scheduled are input into the enhanced immune algorithm to initialize the population and randomly generate antibodies p i represents a possible sending order s of the set of flows to be scheduled F i and routing number r i , several antibodies form a population, and the end-to-end delay D of the flow is calculated using the sending order and routing number of the flow saved in each antibody x , load balancing index V x And the number of routing hops H x , use the affinity operator to calculate the affinity of each antibody in the population, use the antibody concentration evaluation operator to calculate the similarity between each antibody, use the incentive calculation operator to jointly calculate the antibody concentration and antibody affinity, and obtain the final evaluation index of each antibody;

[0061] Use the immune selection operator to select excellent antibodies according to the evaluation indicators, and use the DQN-based cloning mutation operator to adaptively clone and mutate the excellent antibodies to generate new antibodies. These new antibodies are combined with excellent antibodies to initially form a new population. Use the population refresh operator to randomly construct new antibodies to fill the gap or delete excess antibodies to obtain the final new population. Repeat the above process. When the population affinity meets the termination condition, stop the iterative update of the population, and output the sending order and routing number of the flow to be scheduled stored in the current population antibody.

[0062] Preferably, the gating scheduling list of the TSN switch is obtained according to the sending time of each flow, and the TSN switch performs end-to-end deterministic transmission of each flow according to the gating scheduling list, including:

[0063] According to the sending order of the flow and the routing path pointed to by the routing number, the sending time of each TSN switch through which the traffic reaches the destination address is calculated, and the sending time of each flow in each TSN switch is integrated to obtain the gating scheduling table of each port of each TSN switch. When the traffic is sent by the sender, each TSN switch it passes through controls the sending time of the traffic according to the gating scheduling table to achieve end-to-end deterministic transmission of the traffic.

[0064] It can be seen from the technical solutions provided by the above-mentioned embodiments of the present invention that the present invention proposes a traffic scheduling method based on an enhanced immune algorithm. By combining the immune algorithm with reinforcement learning, it overcomes the shortcomings of existing scheduling methods, improves scheduling efficiency and flexibility, and realizes deterministic scheduling of key traffic in the TSN-5G train on-board network.

[0065] Additional aspects and advantages of the present invention will be given in part in the following description, which will become obvious from the following description, or may be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0067] Figure 1 An architecture diagram of a 5G-TSN train on-board network provided in an embodiment of the present invention.

[0068] Figure 2 A processing flow chart of a TSN-5G high-speed rail vehicle network traffic scheduling method based on an enhanced immune algorithm proposed in an embodiment of the present invention;

[0069] Figure 3 A processing flow chart of an enhanced immune algorithm provided by the present method embodiment;

[0070] Figure 4 A DQN algorithm flow chart provided for an embodiment of the method;

[0071] Figure 5 A population schematic diagram provided for an embodiment of the present invention. DETAILED DESCRIPTION

[0072] The embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be interpreted as limiting the present invention.

[0073] It will be understood by those skilled in the art that, unless expressly stated, the singular forms "one", "said", and "the" used herein may also include plural forms. It should be further understood that the term "comprising" used in the specification of the present invention refers to the presence of the features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof. It should be understood that when we refer to an element as being "connected" or "coupled" to another element, it may be directly connected or coupled to the other element, or there may be intermediate elements. In addition, the "connection" or "coupling" used herein may include wireless connection or coupling. The term "and / or" used herein includes any unit and all combinations of one or more associated listed items.

[0074] It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as those generally understood by those skilled in the art in the art to which the present invention belongs. It should also be understood that terms such as those defined in common dictionaries should be understood to have meanings consistent with the meanings in the context of the prior art, and will not be interpreted with idealized or overly formal meanings unless defined as herein.

[0075] To facilitate understanding of the embodiments of the present invention, several specific embodiments will be further explained below with reference to the accompanying drawings, and each embodiment does not constitute a limitation on the embodiments of the present invention.

[0076] The immune algorithm has good adaptability and diversity retention capabilities by simulating the process of antibody recognition, adaptation and elimination of antigens in the biological immune system. The present invention improves the immune algorithm by introducing a reinforcement learning DQN (Deep Q-Network) network to obtain a reinforced immune algorithm. In the iterative process of the reinforced immune algorithm, a reinforcement learning DQN network is introduced to assist in decision-making. The DQN network adaptively optimizes the decision-making strategy through continuous interaction with the immune algorithm population, so that the search ability and convergence speed of the immune algorithm are significantly improved. The improvements include:

[0077] 1: Enhanced global search capability: Reinforcement learning dynamically adjusts the search strategy by continuously interacting with the immune algorithm population, thereby overcoming the deficiency that the immune algorithm may fall into local optimality in complex problems, thereby enhancing its global search capability.

[0078] 2: Improve convergence speed: Reinforcement learning can make decisions based on historical experience and continuously optimize the solution space exploration route of the immune algorithm, which greatly accelerates the convergence speed of the algorithm and shortens the solution time.

[0079] 3: Reduce parameter dependence: Using the DQN network as the cloning mutation operator of the immune algorithm greatly reduces the number of parameters that need to be adjusted in the algorithm and reduces the sensitivity of the algorithm to parameter adjustment. This not only simplifies the algorithm design, but also improves the stability and adaptability of the algorithm.

[0080] The embodiment of the present invention provides a 5G-TSN train on-board network architecture as follows Figure 1 Assume that the set of traffic to be scheduled is F, and a five-tuple is used to represent a traffic in F. in and Indicates the sending and receiving ends of the data stream, prd i Indicates the sending cycle of the data stream, size i Indicates the frame size of the data stream, ddl i Indicates the maximum allowed transmission delay of a data stream.

[0081] Based on the above Figure 1 The 5G-TSN train on-board network shown in FIG. 1 is a processing flow of a TSN-5G high-speed rail on-board network traffic scheduling method based on an enhanced immune algorithm proposed in an embodiment of the present invention. Figure 2 As shown, the processing steps include the following:

[0082] Step S10: Obtain all possible routes of the traffic set F to be scheduled in the TSN-5G high-speed rail on-board network, as well as the delay of each route.

[0083] Step S20: Set various operators of the enhanced immune algorithm, use the DQN network as the clone mutation operator of the enhanced immune algorithm, and minimize the weighted sum of the end-to-end delay, load balancing index and routing hops.

[0084] min(ln(D)+ln(V)+ln(H)) is used as the optimization target of the enhanced immune algorithm;

[0085] Step S30: input all possible routes of the set of traffic to be scheduled, the delay of each route, and the network information of the TSN-5G high-speed rail vehicle network (topology information, switch bandwidth, data transmission rate, etc.) into the enhanced immune algorithm, and the enhanced immune algorithm uses various operators to iterate to obtain the sending order and route of each traffic, as well as the sending time of each traffic on the TSN switch;

[0086] Step S40: Obtain a GCL (Gate Control List) of the TSN switch according to the sending time of each flow, and the TSN switch performs end-to-end deterministic transmission for each flow according to the gate control scheduling list.

[0087] The above step S10 includes, assuming that the flow rate fi There are n possible routes in the TSN-5G high-speed rail vehicle network, and the route set is recorded as R, r ij (j∈[1,2,...,n]) is a route in n, from the sender Passing through k intermediate nodes Arrival at the receiving end Then the route r ij It can be expressed as

[0088] The end-to-end latency of each route includes TSN network latency and 5G network latency.

[0089] Flow rate i The delay of selecting the jth route in the TSN network It consists of 4 parts: sending delay Transmission delay Processing delay and queuing delay Latency in TSN Networks The calculation formula is:

[0090]

[0091] In the TSN-5G network, 5G only acts as a logical bridge, so the traffic f i Latency in 5G Networks Only the queuing delay is affected by the selected route j, which consists of four parts: sending delay Propagation Delay Processing delay and queuing delay Latency in 5G Networks The calculation formula is:

[0092]

[0093] So the flow rate f i The end-to-end delay d of selecting the jth route ij for:

[0094]

[0095] The sending delay, transmission delay and processing delay are mainly determined by the link length or hardware performance. Therefore, these delays are relatively fixed and are not easy to be further reduced by algorithm optimization. Therefore, the optimization focus of the present invention is on the queuing delay. Queuing delay refers to the time that a data packet waits for processing in the buffer queue of a network node. It is affected by network traffic, task allocation and scheduling strategy, and has a large optimization space. The traffic f i The total queuing delay for selecting the jth route is:

[0096]

[0097] 5G data frames can only obtain resources to transmit data at an integer multiple of TTI (Transmission Time Interval).

[0098] The TSN queuing delay is the sum of the queuing delays of each switch, so k is the flow rate f i Select the number of TSN switches that the jth route passes through.

[0099] In order to improve network performance and resource utilization efficiency, avoid single point failure caused by overload of a switch or link, reduce latency and congestion, improve throughput and response speed, ensure QoS (Quality of Service) of key services, and improve network stability and scalability, the present invention also considers the load and routing hops of TSN switches. i The load of the switch when the jth route is selected is x ij , the routing hop count is |r ij |.

[0100] The above step S20 includes: the processing flow of a strengthened immune algorithm provided by the present method embodiment is as follows: Figure 3 As shown in Figure 2, the evolutionary optimization process of the immune algorithm is realized through operators.

[0101] The operators of the enhanced immune algorithm designed by the present invention include: an affinity calculation operator, an antibody concentration calculation operator, an incentive calculation operator, an immune selection operator, a DQN-based cloning mutation operator, and a population refresh operator.

[0102] (1) Population initialization

[0103] The antibodies in population P are coded discretely, and each antibody p i Record a possible sending order s of the set of flows to be scheduled F i and routing number r i .

[0104] (2) Affinity evaluation operator

[0105] Calculate the end-to-end delay D of the xth antibody according to the sending order and routing number of the antibody records x , load balancing index V x And the number of routing hops H x , the affinity aff(x) is the optimization target, so the goal of the algorithm is to continuously iterate and find an antibody with the smallest affinity. The affinity calculation formula is:

[0106] aff(x)=ln(D x )+ln(V x )+ln(H x )

[0107] (3) Antibody concentration evaluation operator

[0108] Define the order of sending traffic within two antibodies only when s i With routing number r i If they are the same, the two antibodies are considered to be the same, so the antibody concentration den(x) of the x-th antibody is defined as:

[0109]

[0110] (4) Excitation degree calculation operator

[0111] The antibody incentive is the final evaluation result of the antibody quality. It needs to comprehensively consider the antibody affinity and antibody concentration. Generally, antibodies with high affinity and low concentration will get a higher incentive. The antibody incentive can usually be obtained by the weighted sum of antibody affinity and antibody concentration. Therefore, the incentive sim(x) of the x-th antibody is:

[0112] sim(x)=w·aff(x)+(1-w)·den(x)

[0113] (5) Immune selection operator

[0114] Define the immune selection operator to select the top three antibodies in terms of excitation.

[0115] (6) Population Refresh Operator

[0116] After adding the cloned and mutated antibodies to the new population, the population size may be insufficient, and the population refresh operator is needed to randomly construct new antibodies to fill the gap, which is conducive to maintaining the diversity of antibodies, achieving global search, and exploring new feasible solution space areas.

[0117] (7) DQN-based cloning mutation operator

[0118] The cloning operator copies the antibody individuals selected by the immune selection operator. The cloning operator is used in the immune algorithm to enhance the diversity and efficiency of the search process. By copying individuals with higher fitness and mutating them, the cloning operator can explore different areas of the solution space, which is the key to accelerating convergence and avoiding local optimal solutions. The traditional cloning operator that statically determines the number of clones limits the algorithm's ability to explore the solution space and is prone to falling into local optimal solutions. In addition, the traditional immune algorithm considers the cloning operator and the mutation operator separately, lacks collaborative optimization, and is difficult to find a suitable operator.

[0119] Figure 4 A DQN algorithm flow chart provided for the embodiment of the method. Therefore, the present invention proposes to use the DQN algorithm in reinforcement learning as the cloning and mutation operator in the immune algorithm. DQN evaluates the value of each antibody through a neural network to determine the number of clones and mutation strategies of the antibody, which can better explore the combination of different clones and mutations and help the immune algorithm converge to the optimal solution faster.

[0120] State space s: the current population’s incentive distribution and vector representation of antibodies.

[0121] Action space a: defines the optional number of clones, mutation strategies, and mutation probabilities.

[0122] The number of clones is 1 to Z, Z = S / N, S is the number of antibodies to be cloned, and N is the population size.

[0123] There are three mutation strategies:

[0124] (7) No mutation is made.

[0125] (8) Change the traffic scheduling order according to the mutation probability.

[0126] (9) Change traffic routing according to mutation probability.

[0127] Reward function r: If the overall affinity of the population is improved after cloning and mutation, a positive reward is given, otherwise a negative reward is given.

[0128] Define the end-to-end delay of the entire set of traffic to be scheduled F as D = min(d ij ), the load balancing index is Routing hop count Since the numerical differences between these three parameters are too large, the natural logarithm is taken to narrow the numerical differences, and the optimization target is min(ln(D)+ln(V)+ln(H)).

[0129] The above step S30 includes: firstly obtaining the current TSN-5G vehicle network switch bandwidth, network topology and other information, and obtaining the information of the flow to be scheduled. Next, the population is initialized and the antibody p is randomly generated. i represents a possible sending order s of the set of flows to be scheduled F i and routing number r i , several antibodies constitute a population, and a schematic diagram of a population provided by an embodiment of the present invention is as follows Figure 5 Then, the end-to-end delay D of the flow is calculated using the sending order and routing number of the flow saved in each antibody. x , load balancing index V x And the number of routing hops H x, then use the affinity operator to calculate the affinity of each antibody in the population. Then use the antibody concentration evaluation operator to calculate the similarity between each antibody, and finally use the incentive calculation operator to jointly calculate the antibody concentration and antibody affinity to obtain the final evaluation index for each antibody.

[0130] The immune selection operator is used to select excellent antibodies according to the evaluation indicators. The DQN-based cloning mutation operator adaptively clones and mutates the excellent antibodies to generate new antibodies. These new antibodies are combined with excellent antibodies to initially form a new population. At this time, the number of antibodies in the new population may be insufficient or excessive, and the population refresh operator is required to randomly construct new antibodies to make up for the shortage or delete the excess antibodies to obtain the final new population. The above process is repeated. When the population affinity meets the termination condition, the population iterative update is stopped, and the sending order and routing number of the scheduled flow stored in the current population antibody are output.

[0131] The above step S40 includes: the sending time of each TSN switch through which the traffic reaches the destination address can be calculated according to the sending order of the flow and the routing path pointed to by the routing number, and the sending time of each flow in each TSN switch can be integrated to obtain the GCL gating scheduling table of each port of each TSN switch. When the traffic is sent by the sender, each TSN switch passed by will control the sending time of the traffic according to the GCL gating scheduling table to ensure the deterministic transmission of the traffic.

[0132] In summary, the embodiment of the present invention constructs a delay model of the TSN-5G vehicle-mounted network, clarifies that the key factors affecting the end-to-end delay are the queuing delay of the TSN switch and the 5G base station, and comprehensively considers important factors such as the switch load and the number of routing hops, and proposes a synchronous scheduling method based on the enhanced immune algorithm. This method effectively reduces the delay and congestion in the network, ensures the stability and reliability of the network under high-load scenarios, and is especially capable of coping with the complexity and dynamic changes of the TSN-5G vehicle-mounted network, and ultimately realizes the deterministic scheduling of the critical traffic of the train, meeting the needs of high real-time communication.

[0133] The present invention integrates the cloning operator and the mutation operator into the DQN network, and uses the adaptive ability of deep learning to flexibly adjust the decision-making strategy according to the changes in the real-time network environment. Through the integrated DQN cloning mutation operator, the global search capability of the immune algorithm is enhanced, and the population diversity is significantly improved, so that the algorithm can more efficiently explore different scheduling schemes when facing complex and changeable network conditions.

[0134] Those skilled in the art can understand that the accompanying drawings are only schematic diagrams of an embodiment, and the modules or processes in the accompanying drawings are not necessarily required to implement the present invention.

[0135] It can be known from the description of the above implementation methods that those skilled in the art can clearly understand that the present invention can be implemented by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which can be stored in a storage medium such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in the various embodiments of the present invention or certain parts of the embodiments.

[0136] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device or system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment. The device and system embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0137] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A TSN-5G high-speed rail vehicle network traffic scheduling method based on enhanced immune algorithm, characterized in that: include: Obtain all possible routes for the set of traffic to be scheduled in the TSN-5G high-speed rail vehicle network, as well as the latency of each route; Set various operators of the enhanced immune algorithm, use the DQN network as the clone mutation operator of the enhanced immune algorithm, and take the minimum weighted sum of end-to-end delay, load balancing index and routing hops as the optimization goal of the enhanced immune algorithm; All possible routes of the set of traffic to be scheduled, the delay of each route and the network information of the TSN-5G high-speed rail vehicle network are input into the enhanced immune algorithm, and the enhanced immune algorithm uses various operators to iteratively obtain the sending order and route of each traffic, as well as the sending time of each traffic on the time-sensitive network TSN switch; A gating scheduling list of the TSN switch is obtained according to the sending time of each flow, and the TSN switch performs end-to-end deterministic transmission for each flow according to the gating scheduling list.

2. The method according to claim 1, characterized in that The method of obtaining all possible routes of the set of traffic to be scheduled in the TSN-5G high-speed rail vehicle network and the delay of each route includes: Assume the flow rate f i There are n possible routes in the TSN-5G high-speed rail vehicle network, and the route set is recorded as R, r ij (j∈[1,2,...,n]) is a route in n, from the sender Passing through k intermediate nodes Arrival at the receiving end Then the route r ij Expressed as Set the end-to-end delay of each route, including TSN network delay and 5G network delay, and the traffic f i The delay of selecting the jth route in the TSN network It consists of 4 parts: sending delay Transmission delay Processing delay and queuing delay Latency in TSN Networks The calculation formula is: Flow rate i Latency in 5G Networks It consists of 4 parts: sending delay Propagation Delay Processing delay and queuing delay Latency in 5G Networks The calculation formula is: Flow rate i The end-to-end delay d of selecting the jth route ij for: Flow rate i The total queuing delay for selecting the jth route is: Queuing delay in 5G networks TTI is the transmission time interval; Queuing delay in TSN network k is the flow rate f i Select the number of TSN switches that the jth route passes through.

3. The method according to claim 2, characterized in that The various operators of the enhanced immune algorithm are set, the DQN network is used as the clone mutation operator of the enhanced immune algorithm, and the weighted sum of the end-to-end delay, the load balancing index and the number of routing hops is minimized as the optimization goal of the enhanced immune algorithm, including: The operators set for the enhanced immune algorithm include: affinity calculation operator, antibody concentration calculation operator, incentive calculation operator, immune selection operator, DQN-based clonal mutation operator, and population refresh operator; (1) Population initialization The antibodies in population P are coded discretely, and each antibody p i Record a possible sending order s of the set of flows to be scheduled F i and routing number r i ; (2) Affinity evaluation operator Calculate the end-to-end delay D of the xth antibody according to the sending order and routing number of the antibody records x , load balancing index V x And the number of routing hops H x , the affinity aff(x) is the optimization target. The algorithm goal is to continuously iterate to find an antibody with the smallest affinity. The affinity calculation formula is: aff(x)=ln(D x )+ln(V x )+ln(H x ) (3) Antibody concentration evaluation operator Define the order of sending traffic within two antibodies only when s i With routing number r i If they are the same, the two antibodies are considered to be the same, and the antibody concentration den(x) of the x-th antibody is defined as: (4) Excitation degree calculation operator The excitation degree sim(x) of the x-th antibody is: sim(x)=w·aff(x)+(1-w)·den(x) (5) Immune selection operator Define the immune selection operator to select the top three antibodies with the highest degree of motivation; (6) Population Refresh Operator The population refresh operator randomly constructs new antibodies to complete the task. (7) DQN-based cloning mutation operator Use the DQN algorithm in reinforcement learning as the cloning and mutation operator in the immune algorithm; State space s: the current population’s incentive distribution and the vector representation of antibodies; Action space a: defines the optional number of clones, mutation strategies, and mutation probabilities; The number of clones is 1 to Z, Z = S / N, S is the number of antibodies to be cloned, and N is the population size; There are three mutation strategies: (1) No mutation is done. (2) Change the traffic scheduling order according to the mutation probability. (3) Change traffic routing according to the mutation probability. Reward function r: If the overall affinity of the population is improved after cloning and mutation, a positive reward is given, otherwise a negative reward is given.

4. The method according to claim 3, characterized in that The optimization goal of the enhanced immune algorithm is to minimize the weighted sum of the end-to-end delay, the load balancing index and the number of routing hops, including: Assume the flow rate f i The load of the switch when the jth route is selected is x ij , the routing hop count is |r ij |, define the end-to-end delay of the entire set of traffic to be scheduled F as D = min(d ij ), the load balancing index is Routing hop count The optimization goal of the enhanced immune algorithm is set to min(ln(D)+ln(V)+ln(H)), that is, the weighted sum of the end-to-end delay, load balancing index and routing hops is minimized.

5. The method according to claim 4, characterized in that The method of inputting all possible routes of the set of traffic to be scheduled, the delay of each route and the network information of the TSN-5G high-speed rail vehicle network into the enhanced immune algorithm, and the enhanced immune algorithm uses various operators to iteratively obtain the sending order and route of each traffic, as well as the sending time of each traffic on the TSN switch, including: The current TSN-5G vehicle network switch bandwidth, network topology and flow information to be scheduled are input into the enhanced immune algorithm to initialize the population and randomly generate antibodies p i represents a possible sending order s of the set of flows to be scheduled F i and routing number r i , several antibodies form a population, and the end-to-end delay D of the flow is calculated using the sending order and routing number of the flow saved in each antibody x , load balancing index V x And the number of routing hops H x , use the affinity operator to calculate the affinity of each antibody in the population, use the antibody concentration evaluation operator to calculate the similarity between each antibody, use the incentive calculation operator to jointly calculate the antibody concentration and antibody affinity, and obtain the final evaluation index of each antibody; Use the immune selection operator to select excellent antibodies according to the evaluation indicators, and use the DQN-based cloning mutation operator to adaptively clone and mutate the excellent antibodies to generate new antibodies. These new antibodies are combined with excellent antibodies to initially form a new population. Use the population refresh operator to randomly construct new antibodies to fill the gap or delete excess antibodies to obtain the final new population. Repeat the above process. When the population affinity meets the termination condition, stop the iterative update of the population, and output the sending order and routing number of the flow to be scheduled stored in the current population antibody.

6. The method according to claim 5, characterized in that The gating scheduling list of the TSN switch is obtained according to the sending time of each flow, and the TSN switch performs end-to-end deterministic transmission of each flow according to the gating scheduling list, including: According to the sending order of the flow and the routing path pointed to by the routing number, the sending time of each TSN switch through which the traffic reaches the destination address is calculated, and the sending time of each flow in each TSN switch is integrated to obtain the gating scheduling table of each port of each TSN switch. When the traffic is sent by the sender, each TSN switch it passes through controls the sending time of the traffic according to the gating scheduling table to achieve end-to-end deterministic transmission of the traffic.

Images