Probability analysis method and optimization scheduling method for ensuring transmission reliability of 5g-tsn network

Abstract

The application discloses a kind of guarantee 5G-TSN network transmission reliability probability analysis method and optimization scheduling method, 5G-TSN network is modeled probabilistically, and the probability characteristics of 5G data injection TSN are highlighted, so that the time slot occupation of TSN network is described in the form of expectation, a novel 5G-TSN modeling method is provided. The reliability of 5G-TSN network is analyzed in detail by SINR, packet loss rate and other indicators, and the end-to-end transmission reliability of 5G-TSN network is improved by using the K times repeated retransmission without feedback. In 5G uplink access, the number of retransmission expression is calculated by using signal-to-noise ratio, packet loss rate, modulation and coding technology (MCS) and reliability threshold to more accurately analyze the transmission of 5G-TSN network and reserve resources.

Description

TECHNICAL FIELD

[0001] The present application mainly relates to the field of communication technology, and in particular to a probability analysis method and an optimization scheduling method for guaranteeing the transmission reliability of a 5G-TSN network. BACKGROUND

[0002] With the rise of latency-sensitive services, traditional industrial networks are evolving towards deterministic networks. Time Sensitive Networking (TSN) has become an important support scheme for realizing deterministic transmission in industrial sites due to its series of fine-grained traffic scheduling protocols. To further break away from wired constraints and enable ubiquitous sensing and flexible manufacturing, a 5G and TSN converged network architecture has been proposed, which is expected to usher in a new era of industrial communication networks. However, the harsh production environment of factories has seriously affected the reliability of 5G air interface transmission, exacerbating the convergence bottleneck of 5G-TSN industrial networks in terms of deterministic performance matching. Therefore, how to improve the transmission performance of 5G networks and build seamless and deeply integrated 5G-TSN networks has become a key to enabling industrial scenarios.

[0003] As one of the key technologies for improving the reliability of 5G networks, retransmission technology can alleviate packet loss and improve the determinism of air interface transmission by sending multiple copies. There are two main retransmission technologies: feedback-based retransmission and feedback-free retransmission. Among them, the feedback-based retransmission scheme, such as Hybrid Automatic Repeat reQuest (HARQ), must wait for the reception of feedback signals before retransmission, which increases the transmission delay at the cost of improving the reliability of 5G air interface. On the other hand, the feedback-free retransmission scheme, such as K times repetition, is introduced in 5G URLLC scenarios, which does not require waiting for feedback and actively sends multiple consecutive copies. Compared with feedback-based retransmission, the transmission delay of feedback-free retransmission scheme will be significantly reduced and is more and more suitable for latency-sensitive scenarios. However, due to the absence of feedback signals, the time when the base station successfully decodes the data packet is not known, which brings uncertainty to the time of injecting 5G data into the TSN network. Therefore, how to characterize the uncertainty of the time of injecting 5G data into the TSN network under the feedback-free retransmission of 5G networks and establish a reliable 5G-TSN end-to-end transmission model is crucial for realizing deterministic transmission of latency-sensitive services.

[0004] As an important mathematical tool, probability is widely used to measure and deal with uncertainty. Based on the probability framework, we can make reasonable inferences and decisions in the face of incomplete information. Based on this, we assume that 5G data injection TSN has probability characteristics, considers the uncertainty caused by 5G retransmission through probability form, so as to infer the data transmission and load distribution of TSN network, and then give the important basis of 5G-TSN network resource allocation. In the worst case, 5G data needs to wait for all the copies to be transmitted before it can be injected into the TSN network, which will produce a higher queuing delay. Through probabilistic modeling, 5G data can realize no-wait forwarding from the base station to the TSN source node, greatly reducing the end-to-end transmission delay, and more meeting the demand of low delay and high reliable transmission in industrial scene.

[0005] The most similar implementation scheme found through the existing literature search is Chinese Patent Application No. 202110718076.8, entitled "5G and TSN joint scheduling method based on wireless channel information". The specific scheme is to design a retransmission factor based on wireless channel information, and to establish a transmission delay budget for the 5G system based on the retransmission factor to eliminate the transmission time jitter of data packets in the 5G system. Although this method also uses a retransmission scheme to deal with the impact of wireless channel instability on end-to-end transmission determinism, the 5G transmission delay budget it builds is an analysis of the worst case, which is significantly different from the research focus of the present invention. Patent Application No. 202210649601.X, entitled "5G-TSN industrial heterogeneous virtual network architecture and virtual resource fine-grained scheduling method", the specific scheme is to use SDN and NFV technology to model 5G-TSN, introduce information age (AoI) to characterize the real-time performance of 5G access, and dynamically adjust the priority of TSN injection based on this to ensure end-to-end deterministic delivery. This scheme focuses on end-to-end latency and designs an ideal environment that does not consider packet loss that may occur in 5G access. Patent Application No. 202311251389.2, entitled "5G and TSN collaborative flow scheduling system and method for industrial internet", the specific scheme is to design a flow client in the CG resource of GF-NOMA technology, and jointly optimize the 5G transmission CG time-frequency resource and the time slot offset of the injected TSN network through deep learning, to realize seamless transmission of 5G network and TSN network. This scheme adopts a deterministic forwarding strategy when setting the base station to TSN gateway process, that is, data will arrive at the TSN network deterministically, which is significantly different from the purpose of the present invention to solve the uncertainty of injection caused by the unknown time of successful decoding at the base station. Patent Application No. 202311748248.1, entitled "Probabilistic repetition coding access method for uplink sporadic load", the specific scheme is that users with sporadic load use a competitive access strategy based on probabilistic continuous repetition coding when accessing the 5G network, and maximize transmission reliability by jointly optimizing the activation probability and the number of retransmissions. However, this scheme only addresses 5G network access issues and does not consider 5G-TSN fusion network application scenarios. In summary, the current research on 5G-TSN sets an ideal environment and there is almost no analysis of transmission reliability in network modeling, and there is no related patent on probabilistic modeling of 5G-TSN networks. Therefore, it is a very innovative research idea to analyze the transmission reliability and model of 5G-TSN networks in a probabilistic form. SUMMARY

[0006] In view of the above-mentioned defects of the prior art, the technical problems to be solved by the present invention include:

[0007] How to design a probabilistic analysis method to ensure the reliability of 5G-TSN network transmission, and how to design an optimized scheduling method based on the above probabilistic analysis method.

[0008] To achieve the above object, the present invention provides a probability analysis method for ensuring 5G-TSN network transmission reliability, which is characterized by comprising the steps of:

[0009] Step 1: Calculate the number of retransmissions k based on the signal-to-noise ratio, MCS, packet loss rate, and reliability threshold.

[0010] Step 2: Allocate k retransmissions in the 5G network and establish a 5G load balancing indicator;

[0011] Step 3: Calculate the transmission success rate of 5G data in the allocated time slot;

[0012] Step 4: Calculate the forwarding time slot and forwarding probability after 5G data is injected into the TSN network;

[0013] Step 5: Describe the TSN network load balancing indicators in the desired form;

[0014] Step 6: Establish a 5G-TSN network load balancing optimization model.

[0015] Furthermore, step 1 is specifically as follows:

[0016] Calculating users Signal-to-noise ratio , select an MCS Derive the packet loss rate under this channel environment , and calculate the 5G time-frequency domain resources required to transmit one copy based on the selected MCS ;

[0017] Obtain reliability threshold based on user QoS requirements , according to the packet loss rate and signal-to-noise ratio Establish retransmission times The calculation expression of .

[0018] Furthermore, step 2 is specifically as follows:

[0019] Convert the packet deadline to the maximum time slot offset that the 5G network can allocate , in the time slot The calculated value in step 1 is The retransmission is allocated, where the user In the The number of retransmissions per time slot is expressed as ;

[0020] Meanwhile, the 5G system needs to be configured with a corresponding 5G transmission resource-time-frequency domain resource block (RB) for the retransmission, wherein the RB is represented as whether the RB is allocated to a user ;

[0021] Therefore, the load balancing index in the 5G system is defined as the bandwidth occupancy rate of each time slot , wherein the total number of RBs that can be allocated on the time slot, i.e.

[0022] .

[0023] Further, step 3 is specifically:

[0024] constructing the 5G data transmission success rate of the user on the time slot .

[0025] .

[0026] Further, in step 4,

[0027] the calculation method of the forwarding probability of the 5G data injected into the TSN network is

[0028] By selecting the 5G parameter set and configuring the mini-slot, the time slot length of 5G and TSN is set to be consistent, both being .

[0029] In the 5G network, after the data packet completes a time slot transmission, it will be immediately injected into the TSN network, and the user from the th 5G time slot , the probability of injecting the th TSN time slot is , wherein .

[0030] In the TSN network, the data will carry this probability for transmission;

[0031] the calculation method of the forwarding time slot of the 5G data injected into the TSN is

[0032] In the TSN network, by using the forwarding characteristics of the CQF model, the forwarding time slot of the flow on the link can be calculated by the number of transmission hops .

[0033] constructing the flow ​​whether in time slot through the link Boolean variable is in the form of representing the flow period:

[0034] .

[0035] Further, step 5 is specifically:

[0036] Through the packet size , forwarding probability and existence variable The load expression is constructed in the form of time slot and link The time slot occupancy of is:

[0037] .

[0038] Further, step 6 is specifically:

[0039] Define the TSN load balancing index as the time slot occupancy of the link, and the TSN gateway usually has the same queue capacity According to the time slot occupancy , the time slot occupancy of the TSN network in time slot and link is:

[0040]

[0041] A multi-objective optimization model for 5G-TSN network load balancing is constructed, which aims to improve the schedulability of the network by minimizing and at the same time.

[0042] .

[0043] An optimization scheduling method based on the foregoing probability analysis method for ensuring the transmission reliability of 5G-TSN network, comprising:

[0044] Phase 1, initialization scheme;

[0045] Phase 2, group result merging;

[0046] Phase 3, dynamic adjustment within the group;

[0047] Phase 4, termination judgment.

[0048] ​Further, the stage 1 is specifically: traffic with the same period and transmission path is divided into a group, and according to the period attribute, packet size attribute and retransmission number attribute of the traffic set in each group, 5G RBs transmission resources and TSN queue resources are allocated; at the same time, the priority of each traffic in the group is set , and then the initial scheduling scheme of each group is obtained by greedy, incremental scheduling according to the priority;

[0049] The stage 2 is specifically: the initial scheduling scheme of each group generated in the stage 1 is combined to form a global scheduling scheme; in this process, a tabu search method is used to find the optimal scheme when the scheduling schemes of each group are combined.

[0050] Further, the stage 3 is specifically: based on the combination result of the stage 2, the group with the TSN load peak value is found, and the current TSN load balancing index value of the group is recorded ; this stage takes the index as a new constraint condition, and re-schedules in the scheduling framework of the stage 1, that is, when the traffic in the group is greedily and incrementally scheduled according to the priority, it is additionally checked whether the load balancing index of each time slot of the TSN network in the group is less than ;

[0051] The stage 4 is specifically: the termination condition is that the change amount of the 5G-TSN load balancing index no longer presents a downward trend, that is , wherein and are weight factors for balancing the magnitude of the 5G and TSN load balancing indexes; if the termination condition is met, the scheduling is ended, and the final scheduling scheme is formed; if not, the stage 2 is returned, and the stage 2 and the stage 3 are iteratively executed.

[0052] Compared with the prior art, the technical effect of the present application is that:

[0053] The present application probabilistically models the 5G-TSN network, focuses on describing the probability characteristics of 5G data injection into TSN, thus describes the time slot occupation of the TSN network in the form of expectation, provides a novel 5G-TSN modeling method, well describes the coupling relationship between 5G and TSN under the condition of no feedback retransmission, and more reasonably analyzes and infers the load distribution of the 5G-TSN network.

[0054] The present application analyzes the reliability of the 5G-TSN network in detail through SINR, packet loss rate and other indexes, and improves the end-to-end transmission reliability of the 5G-TSN network by adopting the K-time repeated feedback-free retransmission mode. Transmission is carried out under the condition of reliability guarantee, which greatly improves the network performance and more robustly copes with the interference generated by the industrial complex environment.

[0055] The application adopts signal-to-noise ratio, packet loss rate, modulation and coding technology (MCS) and reliability threshold to jointly establish a retransmission number calculation expression in 5G uplink access, so as to more accurately analyze the transmission of 5G-TSN network and make resource reservation. The more scientific setting of the retransmission number makes the delay calculation and load analysis of 5G-TSN network more accurate, which is more in line with the resource reservation characteristics of TSN fine granularity, while ensuring the deterministic delivery of data, and maximally reducing the waste of transmission delay. BRIEF DESCRIPTION OF DRAWINGS

[0056] Figure 1 is a 5G-TSN industrial heterogeneous network diagram of the application;

[0057] Figure 2 is a flowchart of the probability analysis method of the application;

[0058] Figure 3 is a flowchart of the optimization scheduling method of the application. DETAILED DESCRIPTION

[0059] The following reference description of the drawings introduces the preferred embodiments of the application, so that the technical content of the application is more clear and easy to understand. The application can be embodied in many different forms of embodiments, and the protection scope of the application is not limited to the embodiments mentioned in the text.

[0060] The application aims to solve the problem of uncertainty of 5G data injection into TSN when applying the no-feedback retransmission technology in 5G-TSN network to guarantee the end-to-end transmission reliability. For this purpose, the application proposes a probability analysis method to effectively depict the uncertainty caused by 5G retransmission, and further proposes an optimization scheduling method based on the probability analysis result to improve the schedulability of 5G-TSN network. The method will provide a complete scheduling scheme for the 5G-TSN industrial heterogeneous network as shown in Figure 1 The flowchart of the probability analysis method is as shown in Figure 2 , and the flowchart of the scheduling scheme based on the probability analysis is as shown in Figure 3 .

[0061] The embodiment adopts the 5G-TSN industrial heterogeneous network architecture as shown in Figure 1 , the bottom layer device data is uploaded through the 5G access network, and the TSN network is responsible for data forwarding as the core network. The 5G system adopts K times repetition scheme as the reliability guarantee mechanism, that is, the user equipment performs fixed K times of transmission attempt for each data packet. At the same time, the TSN adopts the Cyclic Quening Forwarding (CQF) model proposed in IEEE 802.1 Qch protocol for configuration.

[0062] AsFigure 2 As shown, a probability analysis method for ensuring 5G-TSN network transmission reliability is characterized by comprising the steps of:

[0063] Step 1: Calculate the number of retransmissions k based on the signal-to-noise ratio, MCS, packet loss rate, and reliability threshold.

[0064] Step 2: Allocate k retransmissions in the 5G network and establish a 5G load balancing indicator;

[0065] Step 3: Calculate the transmission success rate of 5G data in the allocated time slot;

[0066] Step 4: Calculate the forwarding time slot and forwarding probability after 5G data is injected into the TSN network;

[0067] Step 5: Describe the TSN network load balancing indicators in the desired form;

[0068] Step 6: Establish a 5G-TSN network load balancing optimization model.

[0069] Step 1 is as follows:

[0070] Calculating users Signal-to-noise ratio , select an MCS Derive the packet loss rate under this channel environment , and calculate the 5G time-frequency domain resources required to transmit one copy based on the selected MCS ;

[0071] Obtain reliability threshold based on user QoS requirements , according to the packet loss rate and signal-to-noise ratio Establish retransmission times The calculation expression of .

[0072] Step 2 is as follows:

[0073] Convert the packet deadline to the maximum time slot offset that the 5G network can allocate , in the time slot The calculated value in step 1 is The retransmission is allocated, where the user In the The number of retransmissions per time slot is expressed as ;

[0074] At the same time, it is necessary to The retransmission is allocated with corresponding 5G transmission resources - time-frequency domain resource blocks (RBs), Used to indicate RB Is it assigned to the user? ;

[0075] Therefore, the load balancing indicator in the 5G system is defined as the bandwidth occupancy of each time slot ,in Indicates the total number of RBs that can be allocated on the time slot, that is:

[0076] .

[0077] Step 3 is as follows:

[0078] if If at least one of the copies is correctly decoded, the data transmission in the 5G network is considered successful. Assume that the 5G base station has a filtering function, that is, if multiple data packets are successfully decoded in the same time slot, the base station only allows one data packet to be injected into TSN. Based on the above two points, the user In the time slot 5G data transmission success rate on for:

[0079] .

[0080] Furthermore, in step 4,

[0081] The forwarding probability of the 5G data when injected into the TSN network is calculated as follows:

[0082] By selecting the 5G parameter set and configuring the mini-time slot, the time slot lengths of 5G and TSN are set to be consistent. ;

[0083] In the 5G network, after a data packet completes the transmission of a time slot, it will be immediately injected into the TSN network. From 5G time slots , inject The probability of a TSN time slot is ,in ;

[0084] In the TSN network, data will be transmitted with this probability;

[0085] The calculation method of the forwarding time slot after the 5G data is injected into the TSN network is as follows:

[0086] In TSN network, the forwarding characteristics of CQF model can be used to Computational Flow In the link forwarding slots on ;

[0087] Building a representation flow whether in time slot through the link Boolean variable is in the form of representing the flow period:

[0088] .

[0089] Step 5 is specifically:

[0090] Due to the absence of feedback signals, the time slot of successful decoding of 5G data packets is unknown, which will cause randomness of TSN load distribution. For this problem, we use Chebyshev's law to infer and describe the TSN network load, and construct the load expression in the form of expectation through the data packet size , forwarding probability and existence variable , that is, the time slot occupancy of time slot and link is:

[0091] .

[0092] Step 6 is specifically:

[0093] Define the TSN load balancing index as the time slot occupancy of the link, and the TSN gateway usually has the same queue capacity , according to the time slot occupancy , the time slot occupancy rate of TSN network on time slot and link is:

[0094]

[0095] Since the field data is uploaded through the 5G network and forwarded by the TSN network, and the time and probability of 5G data injection into the TSN network have been constructed in steps 3 and 4. Therefore, we can plan the load distribution of 5G-TSN network through the resource allocation and retransmission allocation of 5G network, so as to simplify the complexity of heterogeneous network cooperative scheduling. According to the 5G, TSN load balancing indexes shown in steps 2 and 5 respectively, we construct a multi-objective optimization model for 5G-TSN network load balancing, aiming to improve the schedulability of the network by minimizing and at the same time.

[0096] .

[0097] ​​To solve the above optimization problem, based on the aforementioned probability analysis method, an optimization scheduling method is proposed, including:

[0098] Stage 1, initialization scheme;

[0099] Stage 2, group result merging;

[0100] Stage 3, intra-group dynamic adjustment;

[0101] Stage 4, termination judgment.

[0102] The stage 1 is specifically: the flows with the same period and transmission path are divided into a group, and according to the period attribute, packet size attribute and retransmission number attribute of the flow set in each group, 5G RBs transmission resource and TSN queue resource are allocated; at the same time, the priority of each flow in the group is set , then the greedy and incremental scheduling is performed according to the priority, so as to obtain the initial scheduling scheme of each group;

[0103] The stage 2 is specifically: the initial scheduling scheme of each group generated in stage 1 is combined to form a global scheduling scheme; in this process, the tabu search method is used to find the optimal scheme when the scheduling schemes of each group are combined.

[0104] The stage 3 is specifically: based on the merging result of stage 2, the group with TSN load peak value is found, and the current TSN load balancing index value of the group is recorded ; this stage takes the index as a new constraint condition and adds it to the scheduling framework of stage 1 for rescheduling, that is, when the flows in the group are greedily and incrementally scheduled according to the priority, it is additionally checked whether the load balancing index of each time slot of the TSN network in the group is less than ;

[0105] The stage 4 is specifically: the termination condition is defined as the change amount of 5G-TSN load balancing index no longer presents a downward trend, that is , wherein and are weight factors for balancing the magnitude of 5G and TSN load balancing index; if the termination condition is met, the scheduling ends and the final scheduling scheme is formed; if not, return to stage 2 and continue to iterate stage 2 and stage 3.

[0106] The application is significantly different from the previous researches which only focus on the optimization of 5G-TSN network end-to-end delay. It deeply considers the interference and influence of the actual factory environment on the 5G channel, and takes reliability as the research focus of 5G-TSN network. Through probability analysis, it predicts and reduces the packet loss rate in harsh environment, thereby providing more accurate and reliable solutions in practical applications, improving the adaptability and robustness of the network. Applying probability theory to the transmission analysis of 5G-TSN network is an innovative method that can more accurately predict and cope with uncertainties in the network, thereby better enabling 5G-TSN network in industrial sites. In addition, the application can dynamically adjust data transmission strategies according to changes in the current network environment, including adjusting the transmission timing of data packets, the allocation of retransmission times, etc., thereby effectively reducing packet loss while maintaining low latency requirements, meeting the strict real-time requirements of industrial scenarios. Through detailed probability analysis and scheduling, the application can optimize the allocation of 5G-TSN network resources such as bandwidth, time slots, etc., avoiding unnecessary resource waste. While ensuring transmission reliability, it improves the overall efficiency and throughput of the network, providing stronger network support for industrial Internet of Things applications.

[0107] Under complex industrial environments, the application of the application is expected to significantly reduce the packet loss rate in 5G-TSN networks, achieving 99.9-99.999% transmission reliability and ensuring the complete transmission of critical data. While ensuring high reliability, it can still effectively control end-to-end latency to meet the real-time needs of industrial applications, with an average end-to-end transmission latency range controlled within milliseconds. Through probability modeling to quantify 5G-TSN network load distribution, reasonable scheduling methods are used to reduce resource idling and waste, improving overall network performance.

[0108] With the rapid development of Industry 4.0 and smart manufacturing, the transmission reliability of 5G-TSN network as a key infrastructure for future networks is directly related to the production efficiency and product quality of factories. The probability analysis and scheduling method of the application solves the bottleneck problem of transmission reliability in the actual environment, providing strong technical support for the wide application of 5G-TSN network in the industrial field. It is expected that this method will be widely applied in smart manufacturing, automated production lines, remote monitoring and maintenance, etc., promoting the digital transformation of industry and improving industrial competitiveness.

[0109] Other embodiments of the application will be apparent to those skilled in the art from consideration of the specification and practice of the application disclosed herein. It is intended that the application cover any and all variations of the application that come within the scope of the

[0110] It is understood that the application is not limited to the precise construction herein disclosed and shown in the drawings, and that various modifications and changes can be made by those skilled in the art without departing from the scope of the application. The scope of the application is limited only by the claims that follow.

Claims

1. A probabilistic analysis method for ensuring 5G-TSN network transmission reliability, characterized by: Including steps; Step 1: Calculate the number of retransmissions k based on the signal-to-noise ratio, MCS, packet loss rate, and reliability threshold. Step 2: Allocate k retransmissions in the 5G network and establish a 5G load balancing indicator; Step 3: Calculate the transmission success rate of 5G data in the allocated time slot; Step 4: Calculate the forwarding time slot and forwarding probability after 5G data is injected into TSN; Step 5: Describe the TSN network load balancing indicators in the desired form; Step 6: Establish a 5G-TSN network load balancing optimization model; The step 1 is specifically as follows: Calculating users Signal-to-noise ratio , select an MCS Derive the packet loss rate under the channel environment , and calculate the 5G time-frequency domain resources required to transmit one copy based on the selected MCS ; Obtain reliability threshold based on user QoS requirements , according to the packet loss rate and signal-to-noise ratio Establish retransmission times The calculation expression of ; The step 2 is specifically as follows: Convert the packet deadline to the maximum time slot offset that the 5G network can allocate , in the time slot The calculated value in step 1 is The retransmission is allocated, where the user In the The number of retransmissions per time slot is expressed as ; At the same time, it is necessary to The retransmission is allocated with corresponding 5G transmission resources - time-frequency domain resource blocks (RBs), Used to indicate RB Is it assigned to the user? ; Therefore, the load balancing indicator in the 5G system is defined as the bandwidth occupancy of each time slot ,in Indicates the total number of RBs that can be allocated on the time slot, that is: ； The step 3 is specifically as follows: Build User In the time slot 5G data transmission success rate on for: ； In the step 4, The forwarding probability of the 5G data when injected into the TSN network is calculated as follows: By selecting the 5G parameter set and configuring the mini-time slot, the time slot lengths of 5G and TSN are set to be consistent. ; In the 5G network, after a data packet completes the transmission of a time slot, it will be immediately injected into the TSN network. From 5G time slots , inject The probability of a TSN time slot is ,in ; In the TSN network, data will be transmitted with this probability; The calculation method of the forwarding time slot after the 5G data is injected into the TSN network is as follows: In TSN network, the forwarding characteristics of CQF model can be used to Computational Flow In the link forwarding slots on ; Building a representation flow Is it in the time slot Through the link Boolean variable In the following form, Representation Flow Cycle: ； The step 5 is specifically as follows: By packet size , forwarding probability and existential variables Construct the following expected load expression, that is, in the time slot and links Time slot occupancy for: ； The step 6 is specifically as follows: The TSN load balancing indicator is defined as the time slot occupancy of the link. TSN gateways usually have the same queue capacity. , according to the time slot occupancy , calculate the TSN network in the time slot and links The time slot occupancy rate for: A multi-objective optimization model for 5G-TSN network load balancing is constructed to minimize and To improve the schedulability of the network, 。 2. An optimization scheduling method based on the probability analysis method for ensuring 5G-TSN network transmission reliability according to claim 1, characterized in that: include: Phase 1, initialization scheme; Stage 2, the results between groups were combined; Stage 3, dynamic adjustment within the group; Stage 4, judgment termination; Phase 1 specifically involves: dividing traffic with the same period and transmission path into a group, and allocating 5G RBs transmission resources and TSN queue resources based on the period attribute, packet size attribute, and retransmission count attribute of the traffic set in each group; At the same time, set the priority for each flow in the group ,Then greedy ,incremental scheduling is performed according to the priority to obtain the ,initial scheduling scheme for each group; The second stage specifically comprises: combining the initial scheduling solutions generated in the first stage to form a global scheduling solution; in this process, using a tabu search method to find the optimal solution when combining the scheduling solutions of the groups; The specific steps of stage 3 are: based on the merged results of stage 2, find the group with the TSN load peak and record the current TSN load balancing index value of the group ; In this stage, this indicator is used as a new constraint and added to the scheduling framework of stage 1 for rescheduling. That is, when the traffic in the group is greedily and incrementally scheduled according to priority, it is necessary to additionally check whether the load balancing indicator of each time slot of the TSN network in the group is less than ; The stage 4 is specifically as follows: the termination condition is defined as the change in the 5G-TSN load balancing indicator no longer shows a downward trend, that is, ,in and is a weight factor used to balance the magnitude of 5G and TSN load balancing indicators. If the termination condition is met, the scheduling ends and the final scheduling plan is formed. If not, the process returns to stage 2 and continues to iterate stages 2 and 3.

Images