Distributed photovoltaic communication method and system based on multiple backoff strategies

Abstract

The invention discloses a distributed photovoltaic communication method and system based on a multi-backoff strategy, and the method comprises the steps: monitoring a channel of a distributed photovoltaic communication system, and judging whether a terminal needs to execute backoff or not according to whether a public channel is idle or not if there is data transmission in the channel; when it is judged that the terminal needs to execute backoff, whether backoff of other terminals in the channel is finished is judged, if not, channel monitoring is returned, otherwise, a Q matrix is constructed, and the current contention window value, the contention window threshold value and the conflict probability of the terminal are calculated according to the Q matrix; the terminal sends data and performs backoff in data transmission according to the current contention window value; and updating the contention window value in combination with the contention window threshold value and the conflict probability according to the size relationship between the current contention window value and the contention window threshold value and whether the data transmission of the terminal is successful or not. According to the invention, the window value is iteratively updated through the Q matrix, so that the situation that the network throughput is reduced due to overlong backoff time can be avoided.

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless network communication, and more specifically, relates to an improved BEB backoff algorithm for multi-backoff strategies of a distributed photovoltaic communication system. Background Art

[0002] Currently, power communication systems often involve the access of multiple distributed photovoltaic devices. In a half-duplex communication environment, both parties cannot send and receive data simultaneously and must alternate. This is like a one-way street where vehicles can only travel in one direction and must wait for oncoming vehicles to pass before they can reverse. Half-duplex communication is commonly used in devices such as walkie-talkies and early hubs. Although the hardware requirements for half-duplex communication are relatively low, it may lead to a reduction in data transmission efficiency because it requires switching between sending and receiving. In such a communication environment, when there is one terminal and multiple user terminals, data transmission and reception are very likely to collide. At this time, the node waiting to send data delays for a period of time according to the backoff algorithm and then competes for channel resources again.

[0003] The BEB algorithm is a traditional backoff algorithm that solves the problem of channel conflict to a certain extent. However, when the channel is congested, this algorithm has certain limitations. It stipulates that the backoff window (CW) of the in-network node that successfully sends data will directly jump to the minimum initial backoff window, which is beneficial for this node to compete for channel resources again in a short time, resulting in other nodes being unable to obtain the channel usage right for a long time and generating the "starvation effect". To alleviate this effect, improved algorithms based on BEB have emerged, such as multiplicative increase linear decrease backoff algorithm, exponential increase exponential decrease (EIED) backoff algorithm, slow contention window decrease (SD) backoff algorithm, etc. Although the BEB algorithm alleviates the "starvation effect" to a certain extent, however, in the case of a blocked channel, its network performance is still not good, and it has not properly handled the docking problem between the backoff algorithm and the military application environment. Because in the military application environment, communication between multiple combat systems such as unmanned aerial vehicles and ground-deployed vehicles is typically half-duplex communication, and obtaining an immediate response is particularly important, while the existing BEB algorithm cannot ensure the immediate transmission of high-priority data.

[0004] In the prior art, there is already a method for adjusting the contention window based on the Q-learning algorithm. However, in the prior art, the Q-value function is mainly used to iteratively calculate and weight thresholds to adjust the contention window. This method has the following technical defects: it is necessary to explicitly store each state-action pair, which is infeasible for large-scale problems and easily leads to a decrease in network throughput due to excessive backoff time; at the same time, it is difficult for the function calculation to handle delayed rewards. If the time delay between the reward and the action is long, it is difficult for the algorithm to associate the current action with the final reward, resulting in a decrease in learning efficiency. Summary of the Invention

[0005] To solve the deficiencies in the prior art, the present invention provides an improved BEB backoff algorithm for the multi-backoff strategy of a distributed photovoltaic communication system, which solves the problems of easy collision in multi-node half-duplex communication in the prior art, as well as the unfairness, large error, and frequent conflicts in the research of the aforementioned backoff algorithm.

[0006] The present invention adopts the following technical solutions.

[0007] A distributed photovoltaic communication method based on a multi-backoff strategy includes the following steps:

[0008] Monitor the channel of the distributed photovoltaic communication system to determine whether there is data in the channel. If there is data transmission, determine whether the terminal needs to perform backoff according to whether the common channel is idle;

[0009] When it is determined that the terminal needs to perform backoff, determine whether the backoff of other terminals in the channel has ended. If it has not ended, return to monitor the channel. Otherwise, construct a Q matrix and calculate the current contention window value, contention window threshold value, and collision probability of the terminal according to the Q matrix;

[0010] The terminal sends data and performs backoff during data transmission according to the current contention window value;

[0011] Update the contention window value according to the size relationship between the current contention window value and the contention window threshold value, and whether the data transmission of the terminal is successful, in combination with the contention window threshold value and the collision probability.

[0012] Preferably, monitoring the channel of the distributed photovoltaic communication system to determine whether there is data in the channel. If there is data transmission, determine whether the terminal needs to perform backoff according to whether the common channel is idle specifically includes:

[0013] If there is no data in the channel, end the process. If there is data transmission in the channel, further determine whether the common channel is idle;

[0014] If the common channel is not idle, it means that backoff needs to be performed;

[0015] If the common channel is idle, it means there is no data on the current channel. At this time, the terminal directly sends data and ends the process.

[0016] Preferably, the constructed Q matrix is in the following form:

[0017]

[0018] In the formula, n is the total number of terminals in the distributed photovoltaic communication system. Each element in the Q matrix represents a different state, and the element value of the next state is calculated based on the element values of the previous state and the current state:

[0019]

[0020] R(i,j) ∈ {Q(i - 1,j), Q(i,j - 1), Q(i - 1,j - 1)}

[0021]

[0022] In the formula, R(i,j) represents the element value of the previous state, and its value is randomly one of Q(i - 1,j), Q(i,j - 1), and Q(i - 1,j - 1);

[0023] is the element value of the current state;

[0024] Q(i + 1,j + 1) is the element value of the next state;

[0025] γ is the state parameter of learning;

[0026] Set in the initial state R(1,1) = 0;

[0027] According to the above calculation, the elements on the diagonal of the Q matrix can be obtained. For the Q values in the other directions except the diagonal elements, RL decision-making is used for value selection, specifically as follows:

[0028]

[0029] Preferably, the current contention window value of the terminal is calculated according to the Q matrix, specifically as follows:

[0030]

[0031] In the formula, CW current represents the current contention window value of the terminal. In the initial state, i = 1, j = 1, and the values of i and j increase by 1 after each iteration.

[0032] Preferably, the calculation of the contention window threshold specifically includes:

[0033] Based on whether the data transmission of other distributed photovoltaic terminals in the channel except this terminal is successful, this terminal combines the current contention window value and calculates the contention window threshold value using different methods respectively:

[0034] When the data transmission of other distributed photovoltaic terminals is successful, the backoff window CW changes to a positive value, and the contention window threshold value Th is calculated based on the Q matrix CW :

[0035]

[0036] In the formula, ε represents the learning direction parameter; CWcurrent represents the current state of the contention window; CWmin represents the minimum value of the contention window, satisfying:

[0037]

[0038] In the formula, in the initial state, i = 1, j = 1, and the values of i and j increase by 1 after each iteration

[0039] When the transmission fails, the contention window threshold value Th CW is calculated as follows:

[0040] Th CW = CW current - 1.

[0041] Preferably, the calculation of the collision probability specifically includes:

[0042] The collision probability refers to the probability that a node has to experience N collisions before successfully sending a piece of data, and the calculation formula is as follows:

[0043]

[0044] In the formula, P N is the collision probability; P C is the probability of a certain node colliding with other nodes; n is the total number of terminals;

[0045] Assume that the successful transmission probability of each active node of each terminal is not affected by the number of collisions. Then, when sending data, the probability of a certain node colliding with other nodes is P C as:

[0046]

[0047] In the formula, represents the probability that node X does not collide with other nodes when sending data; p0 represents the initial collision probability:

[0048]

[0049] Preferably, according to whether the data transmission of this terminal is successful, the contention window value is updated by combining the contention window threshold value and the collision probability, which specifically includes:

[0050] If the current contention window value CW of this terminal current satisfies: CW current ≤Th CW , then according to whether the data transmission of this terminal is successful, the contention window value of this terminal is updated as follows:

[0051]

[0052] If the current contention window value of this terminal and satisfies: CW current >Th CW , then according to whether the data transmission of this terminal is successful, the contention window value of this terminal is updated as follows:

[0053]

[0054] In the formula, T′ cw represents the updated contention window value of this terminal.

[0055] The present invention also proposes a distributed photovoltaic communication system based on a multi-backoff strategy for implementing the distributed photovoltaic communication method based on the multi-backoff strategy, including: a channel monitoring module, a matrix construction module, a calculation module, and a backoff update module;

[0056] The channel monitoring module is used to monitor the channel of the distributed photovoltaic communication system, including determining whether there is data in the channel, whether the common channel is idle, and whether the backoff of other terminals in the channel has ended;

[0057] The matrix construction module is used to construct a Q matrix when this terminal needs to perform backoff and the backoff of other terminals in the channel has ended;

[0058] The calculation module is used to calculate the current contention window value, contention window threshold value, and collision probability of this terminal according to the Q matrix;

[0059] The backoff update module is used to perform backoff during data transmission according to the current contention window value when this terminal sends data, and update the contention window value by combining the contention window threshold value and the collision probability according to the size relationship between the current contention window value and the contention window threshold value, and whether the data transmission of this terminal is successful.

[0060] The present invention also proposes a terminal, including a processor and a storage medium;

[0061] The storage medium is used to store instructions;

[0062] The processor is used to operate according to the instructions to execute the steps of the distributed photovoltaic communication method based on the multi-backoff strategy.

[0063] The present invention also proposes a computer-readable storage medium, on which a computer program is stored. When the program is executed by a processor, the steps of the distributed photovoltaic communication method based on the multi-backoff strategy are implemented.

[0064] The beneficial effects of the present invention are as follows. Compared with the prior art, the present invention provides a strategy for adaptively changing the window value according to conditions, including a multi-backoff strategy of a contention window adjustment strategy based on the Q matrix and the collision probability. By introducing the Q matrix to iterate the window value, the contention window can be adaptively adjusted to shorten the backoff time. Compared with the traditional improved BEB algorithm, the error calculated by the Q matrix loop in the present invention is small enough, so that the algorithm is not affected by the size of the network transmission delay, and the situation of too small updated basic backoff window and too many same-level collisions will not occur. Therefore, it can better avoid the decrease of network throughput caused by too long backoff time. The value of the state-action pair of the Q matrix is directly stored in a table, which is easy to understand and implement, and the optimal strategy effect can be obtained by directly selecting the action with the largest Q value from each state; through the prior probability calculation, the collision probability caused by too short backoff time is reduced and avoided, and the response rate between communication systems can be improved in a complex battlefield environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 It is a schematic diagram of a distributed photovoltaic communication system to which the distributed photovoltaic communication method based on the multi-backoff strategy proposed by the present invention is applied.

[0066] Figure 2 It is a schematic diagram of multi-terminal device transmission conflict in the distributed photovoltaic communication method based on the multi-backoff strategy proposed by the present invention.

[0067] Figure 3 It is a schematic diagram of the Q-learning matrix update and optimization process of the distributed photovoltaic communication method based on the multi-backoff strategy proposed by the present invention;

[0068] Figure 4 It is a flowchart of the distributed photovoltaic communication method based on the multi-backoff strategy proposed by the present invention;

[0069] Figure 5 It is a structural diagram of the distributed photovoltaic communication system based on the multi-backoff strategy proposed by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0070] To make the objectives, technical solutions and advantages of the present invention more clear, the following will clearly and completely describe the technical solutions of the present invention in conjunction with the accompanying drawings in the embodiments of the present invention. The embodiments described in this application are only a part of the embodiments of the present invention, rather than all embodiments. Based on the spirit of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

[0071] As Figure 1 shown, the distributed photovoltaic communication system mainly includes a fusion terminal, a transmission bus, and multiple distributed photovoltaic terminals. Among them, the fusion terminal is responsible for managing all distributed photovoltaic terminal devices, including device registration, data collection, and action control.

[0072] As Figure 2 shown, since the device terminals share a single bus, information conflicts often occur during device data transmission and reception, resulting in information transmission delay and failure problems. Specifically, in the substation area communication system, two pieces of information sent at the same time will collide. When two distributed terminals send registration requests M3 simultaneously, or when the fusion terminal sends an active request M1 while a distributed terminal sends a request M3 or M5, etc., data collisions will occur.

[0073] As Figure 4 shown, Embodiment 1 of the present invention provides a backoff algorithm based on multiple distributed photovoltaic devices. The method includes the following steps:

[0074] Step 1, monitor the channel of the distributed photovoltaic communication system to determine whether there is data in the channel. If there is data transmission, determine whether the present terminal needs to perform backoff based on whether the common channel is idle;

[0075] Specifically, the distributed photovoltaic communication system includes multiple distributed photovoltaic device terminals, and each terminal transmits data through the channel;

[0076] If there is no data in the channel, end the process. If there is data transmission in the channel, further determine whether the common channel is idle;

[0077] If the common channel is not idle, it means that backoff needs to be performed;

[0078] If the common channel is idle, it means that there is no data in the current channel. At this time, the present terminal directly sends data and ends the process.

[0079] Step 2, when it is determined that the present terminal needs to perform backoff, determine whether the backoff of other terminals in the channel has ended. If not, return to channel monitoring. Otherwise, construct a Q matrix, and calculate the contention window value, contention window threshold value, and collision probability of the present terminal according to the Q matrix;

[0080] AsFigure 3 As shown in Figure 3 , the threshold value of the contention window is calculated by the Q-learning algorithm, specifically including:

[0081] The constructed Q matrix is in the following form:

[0082]

[0083] In the formula, n is the total number of terminals in the distributed photovoltaic communication system. Each element in the Q matrix represents a different state. The element value of the next state is calculated according to the element values of the previous state and the current state:

[0084]

[0085] R(i,j) ∈ {Q(i - 1,j), Q(i,j - 1), Q(i - 1,j - 1)}

[0086]

[0087] In the formula, R(i,j) represents the element value of the previous state, and its value is randomly one of Q(i - 1,j), Q(i,j - 1), and Q(i - 1,j - 1);

[0088] is the element value of the current state;

[0089] Q(i + 1,j + 1) is the element value of the next state;

[0090] γ is the state parameter of learning;

[0091] Set the initial state R(1,1) = 0;

[0092] According to the above calculation, the elements on the diagonal of the Q matrix can be obtained. For the Q values in the other directions except the diagonal elements, RL decision-making and value selection are adopted, specifically as follows:

[0093]

[0094] For the constructed Q matrix, its structure can be regarded as a two-dimensional table. The rows of the Q matrix represent all possible states s, the columns represent the actions in each state, and the value of each element represents the Q value of the corresponding state-action pair.

[0095] Preferably, each node in the multi-distributed photovoltaic device channel calculates the size of the corresponding backoff window CW. The nodes work in a distributed manner. After the node transmits the decision threshold value, the contention window value is updated, and then the Q(i,j) value of the Q matrix is transmitted and updated. Without any centralized training and calculation, optimization is achieved through continuous matrix update.

[0096] Further, obtain the current contention window value of this terminal according to the Q matrix:

[0097]

[0098] In the formula, CW current represents the current contention window value of this terminal. In the initial state, i = 1 and j = 1, and the values of i and j increase by 1 after each iteration.

[0099] According to whether the data transmission of the distributed photovoltaic terminal that transmitted the previous data on the common channel is successful, and in combination with the current contention window value of this terminal, update the contention window threshold value of this terminal by using different methods. Specifically:

[0100] When the transmission is successful, the backoff window CW changes to a positive value, and the contention window threshold value Th CW is calculated as follows:

[0101]

[0102] In the formula, v represents the learning direction parameter. This value is not a fixed constant, but gradually decreases as the training progresses, and is determined by the ε-greedy policy algorithm;

[0103] CWmin represents the minimum contention window. In the initial state, i = 1 and j = 1, and the values of i and j increase by 1 after each iteration.

[0104] CWcurrent represents the current state of the contention window;

[0105] In the algorithm, ε determines the learning direction, including exploration and exploitation. Exploration is to take the action with the most rewards from the learning experience so far. However, since it is difficult to optimize the Q training dataset in the early stage of learning, exploration takes random actions and rewards with a high probability. As the number of learning times increases, the present invention pays more attention to exploration, and the more optimized the Q matrix is, the closer the threshold value is to the optimal value.

[0106] When the transmission fails, adjust the contention window threshold value to Th CW as follows:

[0107] Th CW = CW current - 1

[0108] Further, calculating the collision probability specifically includes:

[0109] For the probability p of collision generated by multi-terminal random data transmission, when the distributed photovoltaic device terminal is n, the throughput is an extreme value when the channel transmits only one data message;

[0110] The collision probability refers to the probability that a node has to experience N collisions before successfully sending a piece of data, and the calculation formula is as follows:

[0111]

[0112] In the formula, P N is the collision probability;

[0113] P C is the probability that a certain node collides with other nodes;

[0114] n is the total number of terminals;

[0115] Assume that the successful transmission probability of each active node of the terminal is not affected by the number of collisions. Then, when sending data, the probability that a certain node collides with other nodes is P C as follows:

[0116]

[0117] In the formula, represents the probability that node X does not collide with other nodes when sending data; p0 represents the initial collision probability:

[0118]

[0119] From the Bayesian normal distribution λ~N(k,σ 2 ), it can be seen that when adaptively taking values for p, the most appropriate value should be If the initial collision probability p0 is used as the prior probability of the Bayesian formula, then the total number of terminals n in the channel is considered to be the optimal probability estimate of the Bayesian normal distribution instantaneously λ, and λ>0 is satisfied. Among them, k represents the mean value, and σ 2 represents the variance.

[0120] If the successful transmission probability of each active node of each distributed terminal is not affected by the number of collisions, then when sending data, the probability that node X of this terminal collides with other nodes is P C can be expressed as:

[0121]

[0122] In the formula, represents the probability that node X does not collide with other nodes when sending data.

[0123] Then, the probability P N that node X of this terminal has to experience N collisions before successfully sending a piece of data is

[0124]

[0125] Step 3, this terminal sends data and performs backoff during data transmission according to the current contention window value;

[0126] Step 4, update the contention window value by combining the contention window threshold and the collision probability according to the size relationship between the current contention window value and the contention window threshold, and whether the data transmission of this terminal is successful.

[0127] Further, in step 3, according to the collision probability p and the contention window threshold Th CW , different backoff strategies are executed respectively:

[0128] Let CWcurrent, if the current contention window value CWcurrent ≤ Th CW , it is considered that there is no intense collision in the current network environment, and backoff strategy 2 is executed at this time; if CWcurrent > Th CW then it is considered that there are many collisions in the network. Different backoff strategies are executed for the above two size relationships, and backoff strategy 1 is executed at this time.

[0129] Specifically as follows:

[0130] 1) Backoff strategy 1:

[0131] Since the contention window is greater than the threshold, it is judged that there are many collisions in the network. In this environment, when the data transmission fails, in order to quickly resolve the conflict. The original contention window value T CW is increased at double speed. When the node successfully sends data, its decreasing rate is reduced, and at this time the new contention window value will be reduced to p times the original value to reduce the occurrence of collisions.

[0132]

[0133] In the formula, T′ cw is the updated contention window value.

[0134] 2) Backoff strategy 2:

[0135]

[0136] When CWcurrent ≤ Th CW , it is considered that the competition in the current network environment is not intense. In this case, when the data transmission fails, the original window value T CW is increased by multiplying by double p. When successfully sent, the contention window value T CW is halved and decreased.

[0137] Such as Figure 5As shown in the figure, the present invention also proposes a distributed photovoltaic communication system based on a multi-backoff strategy for the above-mentioned distributed photovoltaic communication method based on a multi-backoff strategy. The system includes: a channel listening module, a matrix construction module, a calculation module, and a backoff update module;

[0138] The channel listening module is used to listen to the channel of the distributed photovoltaic communication system, including determining whether there is data in the channel, whether the common channel is idle, and whether the backoff of other terminals in the channel has ended;

[0139] The matrix construction module is used to construct a Q matrix when this terminal needs to perform backoff and the backoff of other terminals in the channel has ended;

[0140] The calculation module is used to calculate the current contention window value, contention window threshold value, and collision probability of this terminal according to the Q matrix;

[0141] The backoff update module is used to perform backoff during data transmission according to the current contention window value when this terminal sends data, and update the contention window value in combination with the contention window threshold value and the collision probability according to the size relationship between the current contention window value and the contention window threshold value, and whether the data transmission of this terminal is successful.

[0142] The beneficial effect of the present invention is that, compared with the prior art, by introducing the Q-learning matrix to iterate the window value, the present invention can adaptively adjust the contention window to shorten the backoff time. Since the error of the cyclic calculation of the Q matrix is small enough, the algorithm is not affected by the size of the network transmission delay, and there will be no situation where the updated basic backoff window is too small and there are too many peer conflicts. Therefore, it can better avoid the situation of network throughput decline caused by too long backoff time.

[0143] This disclosure may be a system, method, and / or computer program product. The computer program product may include a computer-readable storage medium having computer-readable program instructions thereon for causing a processor to implement various aspects of this disclosure.

[0144] A computer-readable storage medium can be a tangible device that can hold and store instructions for use by an instruction execution device. A computer-readable storage medium may be, for example—but not limited to—an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer-readable storage medium include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), a static random access memory (SRAM), a portable compact disc read-only memory (CD-ROM), a digital versatile disc (DVD), a memory stick, a floppy disk, a mechanically encoded device, such as a punched card or raised structures in grooves storing instructions thereon, and any suitable combination of the foregoing. The computer-readable storage medium used herein is not construed as an instantaneous signal itself, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagated through a waveguide or other transmission medium (e.g., an optical pulse through an optical fiber cable), or an electrical signal transmitted through a wire.

[0145] The computer-readable program instructions described herein can be downloaded from a computer-readable storage medium to various computing / processing devices, or downloaded to an external computer or external storage device through a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network may include a copper transmission cable, an optical fiber transmission, a wireless transmission, a router, a firewall, a switch, a gateway computer, and / or an edge server. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in a computer-readable storage medium in each computing / processing device.

[0146] The computer program instructions for performing the operations of the present disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-related instructions, microcode, firmware instructions, state-setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages such as Smalltalk, C++, etc., and conventional procedural programming languages such as the "C" language or similar programming languages. The computer-readable program instructions may be executed entirely on the user's computer, partially on the user's computer, executed as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider). In some embodiments, by using the state information of the computer-readable program instructions to customize an electronic circuit, such as a programmable logic circuit, a field-programmable gate array (FPGA), or a programmable logic array (PLA), the electronic circuit can execute the computer-readable program instructions to implement various aspects of the present disclosure.

[0147] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: it is still possible to modify the specific embodiments of the present invention or make equivalent substitutions, and any modification or equivalent substitution that does not depart from the spirit and scope of the present invention shall be covered by the protection scope of the claims of the present invention.

Claims

1. A distributed photovoltaic communication method based on multiple backoff strategies, characterized in that: The steps include: Monitor the channels of the distributed photovoltaic communication system to determine whether there is data in the channel. If there is data transmission, determine whether the terminal needs to perform backoff based on whether the public channel is idle; When it is determined that the terminal needs to perform backoff, it determines whether the backoff of other terminals in the channel has ended. If not, it returns to channel monitoring. Otherwise, it builds a Q matrix and calculates the current contention window value, contention window threshold value and conflict probability of the terminal based on the Q matrix. The terminal sends data and performs backoff in data transmission according to the current contention window value; The contention window value is updated according to the size relationship between the current contention window value and the contention window threshold value and whether the data transmission of the terminal is successful, combined with the contention window threshold value and the conflict probability.

2. The distributed photovoltaic communication method based on multiple backoff strategies according to claim 1 is characterized in that: Monitor the channels of the distributed photovoltaic communication system to determine whether there is data in the channel. If there is data transmission, determine whether the terminal needs to perform backoff based on whether the public channel is idle. Specifically, it includes: If there is no data on the channel, the process ends. If there is data transmission on the channel, it further determines whether the public channel is idle; If the public channel is not idle, it means that backoff needs to be performed; If the public channel is idle, it means that there is no data in the current channel. At this time, the terminal directly sends the data and ends the process.

3. The distributed photovoltaic communication method based on multiple backoff strategies according to claim 1 is characterized in that: The constructed Q matrix is ​​as follows: In the formula, n is the total number of terminals in the distributed photovoltaic communication system, each element in the Q matrix represents a different state, and the element value of the next state is calculated based on the element values ​​of the previous state and the current state: R(i,j)∈{Q(i-1,j), Q(i,j-1), Q(i-1,j-1)} In the formula, R(i,j) represents the element value of the previous state, and its value is a random one among Q(i-1,j), Q(i,j-1), and Q(i-1,j-1); is the element value of the current state; Q(i+1,j+1) is the element value of the next state; γ is the state parameter of learning; Set the initial state R(1,1)=0; According to the above calculation, the diagonal elements in the Q matrix can be obtained. For the Q values ​​in the directions other than the diagonal elements, the RL decision is made and the values ​​are selected, as follows:

4. The distributed photovoltaic communication method based on multiple backoff strategies according to claim 3 is characterized in that: The current contention window value of the terminal is calculated according to the Q matrix, as follows: Where CW current Indicates the current contention window value of this terminal. In the initial state, i=1, j=1. The values ​​of i and j are increased by 1 after each iteration.

5. The distributed photovoltaic communication method based on multiple backoff strategies according to claim 4 is characterized in that: The calculation of the contention window threshold value specifically includes: According to whether the data transmission of other distributed photovoltaic terminals in the channel is successful, this terminal combines the current contention window value and uses different methods to calculate the contention window threshold value: When other distributed photovoltaic terminals transmit data successfully, the backoff window CW changes to a positive value, and the contention window threshold Th is calculated based on the Q matrix. CW : In the formula, ε represents the direction parameter of learning; CWcurrent represents the current state of the contention window; CWmin represents the minimum value of the contention window, satisfying: In the formula, i=1, j=1 in the initial state, and the values ​​of i and j are increased by 1 after each iteration. When the transmission fails, the contention window threshold Th cW The calculation formula is as follows: Th CW =CW current -1。 6. The distributed photovoltaic communication method based on multiple backoff strategies according to claim 1, characterized in that: The calculation of the conflict probability specifically includes: The collision probability refers to the probability that a node will experience N collisions before successfully sending a data, and the calculation formula is as follows: Where P N is the conflict probability; P C is the probability of a node colliding with other nodes; n is the total number of terminals; Assuming that the successful transmission probability of each active terminal node is not affected by the number of collisions, the probability of a node colliding with other nodes when sending data is P C for: In the formula, It represents the probability that node X does not collide with other nodes when sending data; p0 represents the initial collision probability:

7. The distributed photovoltaic communication method based on multiple backoff strategies according to claim 6 is characterized in that: According to whether the data transmission of the terminal is successful, the contention window value is updated in combination with the contention window threshold value and the conflict probability, specifically including: If the current contention window value CW current Meet: CW current ≤Th CW , Th CW is the contention window threshold value, then according to whether the data transmission of this terminal is successful, the contention window value of this terminal is updated as follows: If the current contention window value of this terminal meets: CW current >Th Cw , then according to whether the data transmission of this terminal is successful, the contention window value of this terminal is updated as follows: In the formula, T′ cw Indicates the updated contention window value of this terminal.

8. A distributed photovoltaic communication system based on multiple backoff strategies, used to implement the distributed photovoltaic communication method based on multiple backoff strategies described in any one of claims 1 to 7, characterized in that: include: Channel monitoring module, matrix building module, calculation module, backoff update module; The channel monitoring module is used to monitor the channel of the distributed photovoltaic communication system, including determining whether there is data in the channel, whether the public channel is idle, and whether the backoff of other terminals in the channel has ended; The matrix construction module is used to construct the Q matrix when the terminal needs to perform backoff and other terminals in the channel have finished backoff; The calculation module is used to calculate the current contention window value, contention window threshold value and conflict probability of the terminal according to the Q matrix; The backoff update module is used to perform backoff in data transmission according to the current contention window value when the terminal sends data, and to update the contention window value in combination with the contention window threshold value and the conflict probability according to the size relationship between the current contention window value and the contention window threshold value and whether the data transmission of the terminal is successful.

9. A terminal comprising a processor and a storage medium; characterized in that: The storage medium is used to store instructions; The processor is configured to operate according to the instructions to execute the steps of the method according to any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images