A Random Backoff Method for User Access

Abstract

The present invention discloses a random backoff method for user access, which allocates high priority to important users and very important services; queues users who apply for services according to their priorities, with users with higher priorities being placed in the front of the queue and users with lower priorities being placed in the back of the queue; wherein, for users with the same priority, competition-based spectrum allocation and user access are adopted according to the time sequence of service application and competition strategy, and those who win the competition can be placed in the front of the same priority queue, otherwise they are placed in the back; and / or competition-free spectrum allocation and user access strategy is adopted, so that users can obtain resources and access the system through completely random access. When the network is congested and cannot meet the application requirements of all users, the tolerance of different services to delay, rate, etc. is grouped and processed, and the requirements of users of different service types are met with different degrees of differentiation.

Description

Technical Field

[0001] The present invention belongs to the field of wireless communication, and particularly relates to a random backoff method for user access. Background Art

[0002] In traditional terrestrial communication, congestion control generally starts from the network layer and the transport layer. On the one hand, it controls the routing of network flows from the network layer to avoid congested paths. On the other hand, it controls the traffic injected into the network from the transport layer.

[0003] In a satellite communication system, limited resources are shared by all users in the system. When a large number of users perform random access simultaneously, congestion will inevitably occur, resulting in packet loss. In severe cases, the system service will become unavailable. Therefore, it is necessary to study congestion control methods suitable for satellite large-capacity communication. The following are several commonly used communication congestion control methods.

[0004] First, transport layer congestion control: (1) Access type delay mechanism (ACB, Access Class Barring); The access class (AC) is proposed by 3GPP to classify different types of users by priority. When the channel condition is poor, users with higher priority will be allowed to access first, while users with lower priority must wait. ACB defines 16 access types, among which emergency services, security services, and public facility services are given higher priorities. When the system load is heavy, the base station will broadcast the ACB parameters to all M2M users in the system. This parameter contains the access probability corresponding to each priority user and the backoff waiting time. The M2M user that receives the ACB parameters will first determine its access probability according to its own priority, then randomly generate a number between 0 and 1 before initiating a random access request, and compare the generated number with the access probability. If the random number is less than the access probability, the user is allowed to access. Otherwise, the user access is prohibited and the backoff process starts. It can only access again after the backoff time ends, and the above process is repeated. Since this scheme uses the access probability, accurate control can be achieved when the number of users in the system is known, keeping the channel load at a low level, thus maximizing the system throughput and making full use of system resources. (2) Backoff time adjustment mechanism; The backoff time adjustment scheme delays the user's access request. When the system load is large, the satellite transponder will assign different backoff parameters W to the users in the system. Each user that receives the backoff parameter will randomly select a number between 0 and W as its backoff time, and is only allowed to initiate access again after the backoff time ends, so as to achieve the purpose of dispersing intensive access requests within a short period of time to different time periods. It can achieve good results when the number of users is not too large. However, when considering that there are more than one hundred thousand users within the coverage of a satellite beam, this mechanism takes a long time to evenly disperse the user's access requests to different time periods, increasing the system transmission delay. (3) Resource allocation scheme; As mentioned above, the resources in the system are limited. When a large number of users request random access simultaneously, it will cause system congestion. And how to maximize the utilization rate of system resources is a problem worthy of research. The resource allocation scheme is the solution proposed for this problem. This scheme mainly considers allocating access resources such as preambles and physical random access channels (PRACH) in the system to different communication types. In the preamble allocation scheme, there is an independent resource allocation method, that is, different preambles are respectively allocated to different types of communication, allowing them to use different resources for communication, so as to reduce the mutual influence between different types of communication.In some other shared resource allocation schemes, some communication types can use all system resources, while some can only use a part of them, which can also reduce communication interference for certain communication types. In the PRACH resource allocation scheme, the main idea is to increase the communication resources available in the system to solve the shortage of access resources. PRACH mainly has time resources and frequency resources. When congestion occurs, time slots are dynamically increased in the time domain or resource blocks are increased in the frequency domain to alleviate congestion. When the traffic volume decreases, these extra added resources are released again, hoping to completely solve the problem of system resource shortage in this way. (4) Slotting access mechanism; this scheme provides dedicated time slots for each user in the system for random access. Each user has a unique identifier, which calculates the time slot when it should initiate an access request according to the random access period broadcast by the base station. It is not difficult to imagine that when there are many users, allocating a dedicated time slot for each user for random access will result in a very long access period for the system, which some delay-sensitive users cannot accept. Through the above analysis, considering aspects such as operation difficulty, efficiency, and feasibility, the most effective method is to implement congestion control based on the access probability of ACB. Compared with the backoff scheme, the satellite transponder can achieve more accurate and effective control by broadcasting the access probability, keeping the number of users initiating access requests simultaneously at a relatively low level. The prerequisite for implementing this scheme is to obtain the information on the number of users requesting random access in the system. Although the satellite transponder cannot directly obtain this value, it can estimate it using the received data, and then calculate the optimal access probability using the load estimate value to achieve the best congestion control.

[0005] Secondly, congestion handling at the network layer access end. To alleviate congestion at the access end and appropriately reduce the load at the access end, congestion handling at the access end is an important part of access control, and access control plays a very important role in dealing with network congestion. To meet the user's connection anytime, anywhere and the subsequent service requirements, the system must select the most suitable interface for the user to access. With the access of new users or the departure of old users, the network load condition also changes dynamically. If access users are not reasonably allocated, it may block the arriving or upcoming service requests. When designing the access control algorithm, multiple factors such as network conditions, user preferences, and service types need to be considered. Since there are certain contradictions among these factors, that is, the optimal network under certain factors is often relatively poor for other factors. It is impossible to have a system that is optimal in all network standards. Therefore, the access control algorithm can only try to select a relatively optimal and most suitable network to provide services for users. The current network access control algorithms can be mainly divided into the following categories: algorithms based on received signal strength, algorithms based on multi-attribute decision-making, algorithms based on fuzzy logic, algorithms based on utility functions, algorithms based on game theory, and algorithms based on Markov decision processes. The above access congestion control methods have been verified and well applied in terrestrial networks, but due to the very long transmission delay of satellite communication, it is often difficult to obtain the user, network status, and management and control message transmission in a timely manner, and it is difficult to perform access control and resource scheduling according to the real-time status. Summary of the Invention

[0006] The main object of the present invention is to propose a random backoff method for user access, aiming to solve the above existing technical problems.

[0007] To achieve the above object, a random backoff method for user access is proposed, and the method includes:

[0008] For important users and very important services, high priorities are assigned;

[0009] Users applying for services are queued according to their priorities. The higher the priority, the more forward the queuing position, and the lower the priority, the more backward the queuing position;

[0010] Among them, users with the same priority adopt competition-based spectrum allocation and user access according to the time sequence of applying for services and competition strategies. Those who win the competition can be ranked in front of the same priority queue, otherwise ranked behind; and / or adopt a non-competition-based spectrum allocation and user access strategy, and users obtain resources and access the system through completely random access.

[0011] Preferably, the method further includes:

[0012] When the system load is heavy, the satellite broadcasts the ACB parameters to all users in the system. These parameters include the access probabilities corresponding to users of each priority level and the backoff waiting time required.

[0013] Upon receiving the ACB parameters, a user first determines its own access probability based on its priority. Then, before initiating a random access request, the user randomly generates a number between 0 and 1 and compares the generated number with the access probability. If the random number is less than the access probability, the user is allowed to access;

[0014] otherwise, the user's access is prohibited and the backoff process begins. Only after the backoff time ends can the user attempt another access, and the above process is repeated.

[0015] Preferably, the method further includes:

[0016] The user detects the confirmation information for the access request. If the satellite gives a feedback message, there is no congestion. If the satellite still does not give a feedback message after multiple requests, it is very likely that the signaling information for the access request has collided with the information of other users and been lost;

[0017] On the satellite side, the energy and bit error rate of the signaling signals sent by the user are detected. If the user's signaling signal is submerged in a large number of energy signals and cannot be detected, it is confirmed that a collision and congestion have occurred. Similarly, if the detected signal has a very high bit error rate, resulting in the information not being correctly received, it is confirmed that a collision and congestion have occurred;

[0018] After performing signal energy detection, signal bit error rate detection, and collision count detection, learning, reasoning, and decision-making are carried out based on the detection results and learning algorithms to determine the occurrence of congestion.

[0019] Preferably, the method further includes a congestion control step, which combines the estimation method based on a learning automaton and the access adjustment strategy based on ACB, and includes:

[0020] Step (1), mapping the optimization problem of the number of successfully accessed users to an LA model;

[0021] Step (2), at the end of the random access process for each time slot, the satellite acquires the preamble states of all users attempting to access, and then obtains the corresponding collision probability and idle probability;

[0022] Step (3), using the LA-based estimation method to solve the optimization problem of the collision probability, and calculating the collision probability and reward probability;

[0023] Step (4), according to the reward probability, using the DLRI algorithm to update the action probability vector P;

[0024] Step (5), determine whether the optimal state is reached. In the optimal case, calculate the number of users N attempting to access the system through the ACB scheme in each time slot using the optimal λ value; otherwise, execute step (2).

[0025] Step (6), obtain an approximately optimal ACB factor p by estimating Ni.

[0026] Step (7), control the access of each user according to the optimal ACB factor p.

[0027] In addition, a random backoff device for user access is also proposed. The device includes:

[0028] An allocation module that assigns high priority to important users and very important services.

[0029] A control module that queues the users applying for services according to their priorities. The higher the priority, the more forward the queuing position, and the lower the priority, the more backward the queuing position.

[0030] Among them, for users with the same priority, based on the time order of applying for services and the competition strategy, adopt competition-based spectrum allocation and user access. Those who win the competition can be ranked in front of the same priority queue, otherwise behind; and / or adopt a spectrum allocation and user access strategy without competition, and users obtain resources and access the system through completely random access.

[0031] Preferably, the control module further includes:

[0032] When the system load is heavy, the satellite broadcasts ACB parameters to all users in the system. This parameter contains the access probabilities corresponding to users of each priority and the time for backoff waiting.

[0033] Upon receiving the ACB parameters, users will first determine their access probabilities according to their priorities, then randomly generate a number between 0 and 1 before initiating a random access request, and compare the generated number with the access probability. If the random number is less than the access probability, access is allowed;

[0034] otherwise, the user's access is prohibited and the backoff process starts. Only after the backoff time ends can the user perform the next access, and the above process is repeated.

[0035] Preferably, the control module further includes:

[0036] The user detects the confirmation information of the access request. If the satellite gives feedback information, there is no congestion; if the satellite still does not give feedback information after multiple requests, the signaling information of the access request is likely to collide and be lost with the information of other users.

[0037] On the satellite side, the energy and bit error rate of the signaling signal sent by the user are detected. If the user signaling signal is submerged in a large number of energy signals and cannot be detected, it is confirmed that a collision and congestion have occurred; similarly, if the detected signal has a high bit error rate, resulting in the information not being correctly received, it is confirmed that a collision and congestion have occurred.

[0038] After performing signal energy detection, signal bit error rate detection, and collision count detection, learning, reasoning, and decision-making are carried out according to the detection results and learning algorithms to determine the occurrence of congestion.

[0039] Preferably, the device further includes a congestion control module, which combines the estimation method based on the learning automaton and the access adjustment strategy based on ACB, and includes the following operations:

[0040] Step (1), map the problem of optimizing the number of successfully accessed users to the LA model;

[0041] Step (2), at the end of the random access process in each time slot, the satellite will obtain the preamble status of all users attempting to access, and then obtain the corresponding collision probability and idle probability;

[0042] Step (3), use the LA-based estimation method to solve the optimization problem of the collision probability, and calculate the collision probability and reward probability;

[0043] Step (4), according to the reward probability, use the DLRI algorithm to update the action probability vector P;

[0044] Step (5), determine whether the optimum is reached, and in the optimum case, calculate the number of users N attempting to access the system through the ACB scheme in each time slot through the optimum λ value; otherwise, execute step (2);

[0045] Step (6), obtain an approximate optimum ACB factor p by estimating Ni;

[0046] Step (7), control the access of each user according to the optimum ACB factor p.

[0047] In addition, an electronic device is also proposed, including a processor and a memory, where the memory stores a computer program, and the computer program is executed by the processor to implement the random backoff method for user access as described above.

[0048] In addition, a computer storage medium is also proposed, where the computer storage medium stores a program; the program is loaded and executed by the processor to implement the random backoff method for user access as described above.

[0049] The random backoff method for user access provided by the present invention assigns high priorities to important users and very important services; queues the users applying for services according to the priorities, where the higher the priority, the more forward the queuing position, and the lower the priority, the more backward the queuing position; among them, users with the same priority adopt competition-based spectrum allocation and user access according to the time sequence of applying for services and the competition strategy, and those who win the competition can be ranked in front of the same priority queue, otherwise ranked behind; and / or adopt a competition-free spectrum allocation and user access strategy, and users obtain resources and access the system through completely random access. When the network is congested and cannot meet the application requirements of all users, the present invention performs grouped processing on the tolerances of different services for delay, rate, etc., and meets the needs of users of different service types with different discrimination degrees. By combining the learning automaton and the access type delay mechanism, congestion prediction is carried out on the basis of environmental learning and cognition, and the access delay strategy is determined by combining the network monitoring results and the prediction results, so as to reduce or eliminate the delay effect caused by the long delay of satellite communication by predicting in advance and taking measures in advance, and improve the efficiency and accuracy of access congestion control. Description of the Drawings

[0050] Figure 1 It is a flowchart of the random backoff method for user access according to Embodiment 1 of the present invention;

[0051] Figure 2 It is a topological diagram of the congestion perception architecture based on signals and collisions according to Embodiment 1 of the present invention;

[0052] Figure 3 It is a flowchart of the congestion control scheme based on LA and ACB according to Embodiment 1 of the present invention;

[0053] Figure 4 It is a schematic diagram of the performance simulation of the congestion control strategy based on the LA-ACB scheme according to Embodiment 1 of the present invention;

[0054] Figure 5 It is a schematic diagram of the structure of the random backoff device for user access according to Embodiment 1 of the present invention. Detailed Embodiments

[0055] The following describes the preferred embodiments of the present invention with reference to the accompanying drawings of the specification. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention. And without conflict, the embodiments in the present invention and the features in the embodiments can be combined with each other.

[0056] Embodiment 1

[0057] In the space-ground integrated network, there are various types of users with different characteristics, as well as diverse service types. The QoS requirements such as transmission rate, bandwidth, packet loss rate, and delay tolerance characteristics required by various service types are different. Therefore, users can be divided into multiple priorities and groups according to user type, service type, and user requirements. In the case of limited resources, various users queue to access the system according to priorities and service groups.

[0058] The hierarchical and grouped access scheme is to adopt different access schemes for users with different priorities and groups to achieve control over user access. The cost required is relatively high operation complexity. Different processing schemes need to be adopted for different user groups, and the design requirements for the receiving end are very strict.

[0059] In the space-ground integrated network, service division can be roughly classified into the following categories: (1) Voice call service: The voice call service is one of the most important services in communication. The required rate and output transmission reliability of the voice call service are not high, but the requirement for delay is relatively high, and it cannot be interrupted during the call; (2) Instant messaging service: Instant Messaging is a chat tool that sends / receives text, voice, or video in real time over the network. Instant messaging software usually transmits data through TCP or UDP protocols and belongs to P2P type applications. Common instant messaging software includes Tencent QQ, WeChat, etc.; (3) Data service: Data sent via satellite, including text, pictures, faxes, etc. A special data service, namely the short message service, is set up in the space-ground integrated network. The amount of data transmitted by the short message service is relatively small, and the requirement for real-time performance is not high, but the requirement for reliability is relatively high. It can be used to transmit some directive information; (4) Video service: Video refers to an online playback service based on streaming media technology or a large amount of video information. The data volume is relatively large. The online video service has a certain requirement for real-time performance, but video transmission does not have too high a requirement for real-time performance and has a certain tolerance for delay. The above service types have different requirements for rate, delay, bit error rate, etc. during access and transmission, so the resource requirements during access and transmission are also different. Therefore, when the network is congested and cannot meet the application requirements of all users, it is necessary to perform grouped processing on the tolerance of delay, rate, etc. for different services to meet the needs of users of different service types with different discrimination degrees.

[0060] Refer to Figure 1 , Figure 1 is a flowchart of a random backoff method for user access in this embodiment. In this embodiment, the method includes:

[0061] S1, for important users and very important services, assign high priorities;

[0062] S2. Queue the users applying for services according to their priorities. The higher the priority, the more forward the queuing position of the user; the lower the priority, the more backward the queuing position of the user. Among them, for users with the same priority, based on the time sequence of applying for services and the competition strategy, adopt competition-based spectrum allocation and user access. Those who win the competition can be ranked in front of the same priority queue, otherwise behind; and / or adopt a non-competitive spectrum allocation and user access strategy, and allow users to obtain resources and access the system through completely random access.

[0063] Specifically, in this embodiment, for important users and very important services, high priorities are assigned. For general users and services, different priorities can be assigned according to the system configuration requirements (the more priorities, the more refined the system's service and business support for users, but the higher the system complexity; on the contrary, the fewer priorities, the lower the refinement degree of the system's service and business support for users, but the system complexity is reduced and the overhead is decreased. Therefore, an appropriate number of priorities should be selected to achieve a balance between system performance and complexity). Then, queue the users applying for services according to their priorities. The higher the priority, the more forward the queuing position of the user; the lower the priority, the more backward the queuing position of the user. To avoid frequent occurrences of user queuing timeout situations, it is necessary to perform dynamic priority allocation and timeout setting according to the user service type and queuing time to ensure that users can obtain services within a certain period of time. Second, for users with the same priority in the same batch of applications, based on the time sequence of applying for services and the competition strategy, adopt competition-based spectrum allocation and user access. Those who win the competition can be ranked in front of the same priority queue, otherwise behind; or a non-competitive spectrum allocation and user access strategy can be adopted, and allow users to obtain resources and access the system through completely random access.

[0064] Preferably, the method further includes: when the system load is heavy, the satellite will broadcast ACB parameters to all users in the system. This parameter contains the access probabilities corresponding to users of each priority and the time for backoff waiting. And the users receiving the ACB parameters will first determine their access probabilities according to their own priorities, then randomly generate a number between 0 and 1 before initiating a random access request, and compare the generated number with the access probability. If the random number is less than the access probability, access is allowed; otherwise, the user is prohibited from accessing and the backoff process starts. It can only perform the next access until the backoff time ends, and repeat the above process.

[0065] Specifically, in this embodiment, 16 access types are defined in the 3GPP standard. Among them, emergency services, security services, public facility services, etc. are given higher priorities. When the system load is heavy, the satellite broadcasts the ACB parameter to all users in the system. This parameter contains the access probabilities corresponding to users of each priority and the backoff waiting time. The user who receives the ACB parameter will first determine its own access probability according to its priority, then randomly generate a number between 0 and 1 before initiating a random access request, and compare the generated number with the access probability. If the random number is less than the access probability, the user is allowed to access. Otherwise, the user access is prohibited and the backoff process starts. Only after the backoff time ends can the user perform the next access, and the above process is repeated.

[0066] In a wireless network communication environment, the occurrence of congestion is usually accompanied by the annihilation of signal energy, the increase in bit error rate, the increase in collision probability, etc. Therefore, the occurrence of network congestion can be perceived from the energy and bit error rate at the signal level, and the number of collisions at the user level.

[0067] In the process of access and transmission of control information, a congestion perception architecture based on signals and the number of collisions is designed. By detecting and inferring the occurrence and degree of congestion, it provides a decision basis for subsequent congestion control. The congestion perception architecture based on signals and collisions is as Figure 2 shown.

[0068] Preferably, the method further includes: the user detects the confirmation information of the access request. If the satellite gives feedback information, there is no congestion. If the satellite still does not give feedback information after multiple requests, the signaling information of the access request is likely to collide and be lost with the information of other users. The satellite side detects the energy and bit error rate of the signaling signal sent by the user. If the user signaling signal is annihilated in a large number of energy signals and cannot be detected, it is confirmed that a collision and congestion have occurred. Similarly, if the detected signal bit error rate is very high, resulting in the inability to correctly receive the information, it is confirmed that a collision and congestion have occurred. After performing signal energy detection, signal bit error rate detection, and number of collisions detection, learning, inference, and decision are performed according to the detection results and learning algorithms to determine the occurrence of congestion.

[0069] In this embodiment, the source - end congestion state awareness establishes a mechanism that combines explicit congestion notification and autonomous awareness based on link historical data and current state to sense and predict the congestion state and degree. It is intended to establish a user access behavior analysis model, classify user types, optimize the congestion window parameters and data transmission data according to user access behavior decisions, and achieve intelligent congestion control based on the space - based control protocol architecture. It takes into account both explicit congestion notification and autonomous congestion detection, and studies an adaptive congestion control strategy based on service priorities. Develop link perception and prediction technologies, form a dynamic prediction algorithm for link transmission performance, correct the congestion data packet loss decision - making process, enrich the differential intelligent congestion control mechanism for user, service, and link characteristics, and implement a fast data re - transmission and congestion recovery algorithm under overload conditions. In this embodiment, it is intended to develop relay - node congestion state awareness and congestion degree prediction technologies based on machine learning, and establish a relay hop - by - hop collaborative congestion control mechanism that adapts to link characteristics based on the space - based control protocol architecture. Construct a hop - by - hop collaborative congestion prediction model through the analysis of service rate, cache queue, and link load conditions to alleviate the ultra - long latency feedback effect of satellite network cognitive information. Based on network state awareness, intelligently calculate and mark the packet discard probability, optimize and implement the active cache queue management of relay nodes, and reduce the queuing delay of data packets at relay nodes. Under the condition of limited hardware resources, balance and improve link transmission efficiency and energy efficiency, develop a dynamic cache scheduling strategy based on service priorities, and ensure the quality of service of critical services.

[0070] In the above analysis, the advantages of the congestion control strategy for hierarchical and grouped control of users are elaborated. To implement this strategy, in this embodiment, based on the ACB scheme, a machine - learning method is introduced. Based on machine - learning perception and prediction, an estimation method based on Learning Automata (LA) is used to estimate future traffic and available information using the congestion state and link load state. In addition, an ACB factor adjustment strategy is used to dynamically adjust the ACB factor to control user access and prevent network congestion.

[0071] Preferably, the method further includes a congestion control step, which combines an estimation method based on a learning automaton and an access adjustment strategy based on ACB, and includes: Step (1), mapping the optimization problem of the number of successfully accessed users to an LA model; Step (2), at the end of the random access process in each time slot, the satellite will obtain the preamble states of all users attempting to access, and then obtain the corresponding collision probability and idle probability; Step (3), using the LA-based estimation method to solve the optimization problem of the collision probability, and calculating the collision probability and the reward probability; Step (4), according to the reward probability, using the DLRI algorithm to update the action probability vector P; Step (5), determining whether the optimum is reached, and in the optimum case, calculating the number of users N attempting to access the system through the ACB scheme in each time slot through the optimum λ value; otherwise, execute Step (2); Step (6), obtaining an approximate optimum ACB factor p by estimating Ni; Step (7), controlling the access of each user according to the optimum ACB factor p.

[0072] Specifically, in this embodiment, the congestion control scheme involved combines an estimation method based on a learning automaton and an access adjustment strategy based on ACB, and is collectively referred to as the LA-ACB-based congestion control scheme. Among them, a learning automaton adjusts itself by continuously interacting with the environment. That is to say, it obtains experience through continuous communication with the environment to improve its behavior, so as to select the optimum action in the available actions in this environment, and the optimum action is the action that can obtain the maximum environmental reward probability in the current environment. As Figure 3 shown, it is the flowchart of the congestion control scheme based on LA and ACB in this embodiment.

[0073] Specifically, in this embodiment, (1) the optimization problem is mapped to a learning automaton model. (a) Action: The set of action operations A in this problem is defined as the set of possible λ values (λ is the number of users successfully accessing the system through the ACB scheme). Theoretically, λ can be any value, but it does not exceed the total number of users. (b) Feedback: In the LA-ACB scheme, the feedback is defined as B = {0, 1}, where 0 represents a reward and 1 represents a penalty. (c) Environment: The environment is responsible for generating feedback on the selected action. In this model, the environment mapping simulates the random access process. In each iteration, in this embodiment, the generated collision probability and idle probability can be obtained from the simulated random access process through randomly selected values of λ and a random value γ. (d) Learning automaton: Considering the optimality and convenience of implementation, this embodiment will use the Discretized Linear Reward-Inaction automaton algorithm DLRI (Discretized Linear Reward-Inaction automaton) for learning. (2) At the end of the random access process in each time slot, the satellite will obtain the preamble states of all users attempting to access, and then obtain the corresponding collision probability and idle probability. (3) Use the LA-based estimation method to solve the optimization problem of the collision probability. In the LA-based learning process, at the j-th iteration, LA will randomly select an action aj according to the action probability vector P, and then calculate the collision probability and the reward probability. The smaller the gap between the estimated collision probability and the actual collision probability, the greater the likelihood of obtaining a reward. (4) Use the DLRI algorithm to update the action probability vector P. The probability of the action that obtains a reward being selected in the next iteration becomes larger. After multiple iterations, the optimal λ value is obtained. (5) Calculate the number of users Ni attempting to access the system through the ACB scheme in each time slot through the optimal λ value. (6) Obtain the approximate optimal ACB factor p through the estimation of Ni. (7) Control the access of each user according to the optimal ACB factor p.

[0074] To evaluate the performance of the LA-ACB scheme, simulations were carried out on it, as Figure 4 shown in the schematic diagram of the congestion control strategy performance simulation based on the LA-ACB scheme of this embodiment. Where R is the maximum number of users who need to access the system in each time slot, the curve is the simulation curve of the successful access probability of the ACB scheme when the satellite completely knows the number of users attempting to access the system, the curve is the simulation curve of the successful access probability of the ACB scheme when the satellite completely knows the number of users who successfully accessed the system in the previous time slot. The LA-ACB scheme is the simulation curve of the successful access probability based on the LA-ACB scheme when the satellite does not know the above prior information. It can be seen from the figure that the performance of the LA-ACB scheme is already close to that of the scheme and The solution is not sensitive to the total number of users and can still provide high-quality services to users when the network scale expands.

[0075] In addition, as Figure 5 shown, this embodiment also proposes a random backoff device for user access, and the device includes:

[0076] An allocation module 10 that assigns high priorities to important users and very important services;

[0077] A control module 20 that queues the users applying for services according to the priorities. The higher the priority, the more forward the queuing position of the user, and the lower the priority, the more backward the queuing position of the user;

[0078] Among them, for users with the same priority, based on the time sequence of applying for services and the competition strategy, a competition-based spectrum allocation and user access are adopted. Those who win the competition can be ranked in front of the same priority queue, otherwise ranked behind; and / or a non-competitive spectrum allocation and user access strategy is adopted, and users obtain resources and access the system through completely random access.

[0079] Preferably, the control module further includes:

[0080] When the system load is heavy, the satellite broadcasts the ACB parameters to all users in the system, and these parameters include the access probabilities corresponding to users of each priority and the time for backoff waiting;

[0081] The users receiving the ACB parameters will first determine their own access probabilities according to their priorities, and then before initiating a random access request, randomly generate a number between 0 and 1, and compare the generated number with the access probability. If the random number is less than the access probability, access is allowed;

[0082] Otherwise, the user's access is prohibited and the backoff process starts. It can only perform the next access after the backoff time ends, and the above process is repeated.

[0083] Preferably, the control module further includes:

[0084] The control module detects the confirmation information of the request for access. If the satellite gives feedback information, there is no congestion; if the satellite still does not give feedback information after multiple requests, the signaling information of the request for access is likely to collide and be lost with the information of other users;

[0085] On the satellite side, the energy and bit error rate of the signaling signal sent by the user are detected. If the user signaling signal is submerged in a large number of energy signals and cannot be detected, it is confirmed that a collision and congestion have occurred; similarly, if the detected signal has a high bit error rate, resulting in the information not being correctly received, it is confirmed that a collision and congestion have occurred.

[0086] After performing signal energy detection, signal bit error rate detection, and collision count detection, learning, reasoning, and decision-making are carried out according to the detection results and learning algorithms to determine the occurrence of congestion.

[0087] Preferably, the device further includes a congestion control module, which combines the estimation method based on the learning automaton and the access adjustment strategy based on ACB, including performing the following operations:

[0088] Step (1), map the problem of optimizing the number of successfully accessed users to the LA model;

[0089] Step (2), at the end of the random access process in each time slot, the satellite will obtain the preamble status of all users attempting to access, and then obtain the corresponding collision probability and idle probability;

[0090] Step (3), use the LA-based estimation method to solve the optimization problem of the collision probability, and calculate the collision probability and reward probability;

[0091] Step (4), according to the reward probability, use the DLRI algorithm to update the action probability vector P;

[0092] Step (5), determine whether the optimum is reached, and in the optimum case, calculate the number of users N attempting to access the system through the ACB scheme in each time slot through the optimum λ value; otherwise, execute step (2);

[0093] Step (6), obtain an approximately optimum ACB factor p by estimating Ni;

[0094] Step (7), control the access of each user according to the optimum ACB factor p.

[0095] In addition, this example also proposes an electronic device, including a processor and a memory, where the memory stores a computer program, and the computer program is executed by the processor to implement the random backoff method for user access as described above.

[0096] In addition, this example also proposes a computer storage medium, where the computer storage medium stores a program; the program is loaded and executed by the processor to implement the random backoff method for user access as described above.

[0097] In the solution of the above embodiments of the present invention, the provided random backoff method for user access assigns high priorities to important users and very important services; queues the users applying for services according to the priorities, where the users with higher priorities are queued closer to the front, and the users with lower priorities are queued closer to the back; among them, for users with the same priority, based on the time sequence of applying for services and the competition strategy, a competition-based spectrum allocation and user access is adopted, and those who win the competition can be ranked in front of the same priority queue, otherwise ranked behind; and / or a competition-free spectrum allocation and user access strategy is adopted, and users obtain resources and access the system through completely random access. When the network is congested and cannot meet the application requirements of all users, the present invention performs grouped processing on the tolerances of different services for delay, rate, etc., and meets the needs of users of different service types with different discrimination degrees. By combining the learning automaton and the access type delay mechanism, congestion prediction is performed on the basis of environmental learning and cognition, and the access delay strategy is determined by combining the network monitoring results and the prediction results, so as to reduce or eliminate the delay effect caused by the long delay of satellite communication by predicting in advance and taking measures in advance, and improve the efficiency and accuracy of access congestion control.

[0098] It should be noted that it should be understood that the division of each module of the above device is only a logical function division. In actual implementation, it can be fully or partially integrated into a physical entity, or physically separated. And these modules can all be implemented in the form of software called by a processing element. It can also all be implemented in the form of hardware. It can also be that some modules are implemented in the form of software called by a processing element, and some modules are implemented in the form of hardware.

[0099] In addition, an embodiment of the present invention further provides a readable storage medium, in which computer execution instructions are stored, and when a processor executes the computer execution instructions, the above information collection method for high-frequency experiment assessment is implemented.

[0100] The basic concepts have been described above. Obviously, for those skilled in the art, the above detailed disclosure is only an example and does not constitute a limitation to this specification. Although not explicitly stated here, those skilled in the art may make various modifications, improvements, and deviations to this specification. Such modifications, improvements, and deviations are proposed in this specification, so such modifications, improvements, and deviations still belong to the spirit and scope of the exemplary embodiments of this specification.

[0101] In addition, those skilled in the art can understand that various aspects of this specification can be illustrated and described by several patentable types or situations, including any new and useful process, machine, product, or composition of matter, or any new and useful improvement thereof. Accordingly, various aspects of this specification can be executed entirely by hardware, entirely by software (including firmware, resident software, microcode, etc.), or by a combination of hardware and software. The above hardware or software can all be referred to as "data block", "module", "engine", "unit", "component" or "system". In addition, various aspects of this specification may be embodied as a computer product located in one or more computer-readable media, which includes computer-readable program code.

[0102] It should be noted that if there are inconsistencies or conflicts between the descriptions, definitions, and / or uses of terms in the supplementary materials of this specification and the content described in this specification, the descriptions, definitions, and / or uses of terms in this specification shall prevail.

[0103] Finally, it should be understood that the embodiments described in this specification are only used to illustrate the principles of the embodiments of this specification. Other deformations may also fall within the scope of this specification. Therefore, by way of example and not limitation, alternative configurations of the embodiments of this specification can be regarded as consistent with the teachings of this specification. Accordingly, the embodiments of this specification are not limited to the embodiments explicitly introduced and described in this specification.

Claims

1. A random backoff method for user access, characterized in that, the method includes: For important users and very important services, assign high priorities; Queue the users applying for services according to the priorities. The higher the priority, the more forward the queuing position of the user, and the lower the priority, the more backward the queuing position of the user; Among them, for users with the same priority, based on the time sequence of applying for services and the competition strategy, adopt competition-based spectrum allocation and user access. Those who win the competition can be ranked in front of the same priority queue, otherwise ranked behind; and / or adopt a competition-free spectrum allocation and user access strategy, and let users obtain resources and access the system through completely random access; The method further includes a congestion control step, which combines an estimation method based on a learning automaton and an access adjustment strategy based on ACB, including: Step (1), map the optimization problem of the number of successfully accessed users to the LA model; Step (2), at the end of the random access process in each time slot, the satellite will obtain the preamble states of all users attempting to access, and then obtain the corresponding collision probability and idle probability; Step (3), use the LA-based estimation method to solve the optimization problem of the collision probability, and calculate the collision probability and the reward probability; Step (4), according to the reward probability, use the DLRI algorithm to update the action probability vector P; Step (5), determine whether the optimum is reached, and in the optimum case, calculate the number of users Ni attempting to access the system through the ACB scheme in each time slot through the optimum λ value; otherwise, execute step (2); Step (6), obtain an approximately optimum ACB factor p through the estimation of Ni; Step (7), control the access of each user according to the optimum ACB factor p.

2. The random backoff method for user access according to claim 1, characterized in that, the method further includes: When the system load is heavy, the satellite will broadcast the ACB parameters to all users in the system. This parameter includes the access probabilities corresponding to users of each priority and the time for backoff waiting; And the users receiving the ACB parameters will first determine their own access probabilities according to their priorities, and then before initiating a random access request, randomly generate a number between 0 and 1, and compare the generated number with the access probability. If the random number is less than the access probability, allow it to access; Otherwise, prohibit the user from accessing and start the backoff process. It can only access again after the backoff time ends, and repeat the above process.

3. The random backoff method for user access according to claim 1, characterized in that, the method further includes: The user detects the confirmation information of the access request. If the satellite gives feedback information, then there is no congestion; if the satellite still does not give feedback information after multiple requests, the signaling information of the access request is likely to collide with the information of other users and be lost; On the satellite side, the energy and bit error rate of the signaling signal sent by the user are detected. If the user signaling signal is submerged in a large number of energy signals and cannot be detected, it is confirmed that a collision and congestion have occurred. Similarly, if the detected signal has a high bit error rate, resulting in incorrect reception of information, it is confirmed that a collision and congestion have occurred. After performing signal energy detection, signal bit error rate detection, and collision count detection, learning, reasoning, and decision-making are carried out according to the detection results and learning algorithms to determine the occurrence of congestion.

4. A random backoff device for user access Characterized in that The device includes: An allocation module that assigns a high priority to important users and very important services. A control module that queues the users applying for services according to the priority. The higher the priority, the more forward the queuing position of the user, and the lower the priority, the more backward the queuing position of the user. Among them, users with the same priority adopt competition-based spectrum allocation and user access according to the time sequence of applying for services and competition strategies. Those who win the competition can be ranked in front of the same priority queue, otherwise ranked behind; and / or adopt a competition-free spectrum allocation and user access strategy, and users obtain resources and access the system through completely random access. The device further includes a congestion control module, which combines the estimation method based on the learning automaton and the access adjustment strategy based on ACB, and includes the following operations: Step (1), mapping the optimization problem of the number of successfully accessed users to the LA model. Step (2), at the end of the random access process in each time slot, the satellite will obtain the preamble status of all users attempting to access, and then obtain the corresponding collision probability and idle probability. Step (3), using the LA-based estimation method to solve the optimization problem of the collision probability, and calculating the collision probability and reward probability. Step (4), according to the reward probability, using the DLRI algorithm to update the action probability vector P. Step (5), determining whether the optimum is reached, and in the optimum case, calculating the number of users Ni attempting to access the system through the ACB scheme in each time slot through the optimum λ value; otherwise, execute step (2). Step (6), obtaining an approximately optimum ACB factor p through the estimation of Ni. Step (7), controlling the access of each user according to the optimum ACB factor p.

5. The random backoff device for user access according to claim 4 Characterized in that The control module further includes: When the system load is heavy, the satellite broadcasts the ACB parameters to all users in the system. This parameter includes the access probability corresponding to each priority user and the time to wait for backoff. And the users receiving the ACB parameters will first determine their access probability according to their own priority, then randomly generate a number between 0 and 1 before initiating a random access request, and compare the generated number with the access probability. If the random number is less than the access probability, access is allowed; Otherwise, the user is prohibited from accessing and the backoff process starts. It can only perform the next access after the backoff time ends, and repeat the above process.

6. The random backoff device for user access according to claim 4, wherein, the control module further includes: detect the confirmation information of the user's request for access. If the satellite gives feedback information, there is no congestion. If the satellite still does not give feedback information after multiple requests, the signaling information of the request for access is likely to collide and be lost with the information of other users; the satellite side detects the energy and bit error rate of the signaling signal sent by the user. If the user signaling signal is submerged in a large number of energy signals and cannot be detected, it is confirmed that a collision and congestion have occurred. Similarly, if the detected signal has a high bit error rate and the information cannot be correctly received, it is confirmed that a collision and congestion have occurred; after performing signal energy detection, signal bit error rate detection, and collision times detection, perform learning, reasoning, and decision-making according to the detection results and learning algorithms to determine the occurrence of congestion.

7. An electronic device, comprising a processor and a memory, the memory storing a computer program, the computer program being executed by the processor to implement the random backoff method for user access according to any one of claims 1-3 above.

8. A computer storage medium, the computer storage medium storing a program; the program is loaded and executed by a processor to implement the random backoff method for user access according to any one of claims 1-3 above.

Images