A method for joint optimization of deployment location and working state of a wireless access point

Abstract

This invention provides a method for jointly optimizing the deployment location and operating status of wireless access points (APs). The method includes acquiring a set of access points (APs) in a wireless local area network (WLAN), the location vectors of each AP, and its operating status; dividing the WLAN into regions with multiple unit cubes to obtain the user density in each sub-region; determining the target points to be covered in each sub-region to form a target point set; and calculating the signal strength from the APs. i To the target point T j The path loss; constrained by the overall coverage of target points and the user correlation rate in the region of each sub-cube, a path loss with each AP is constructed. i The objective function is related to the TX power; a single-objective optimization algorithm is used to find the optimal solution for the objective function and output it, thus obtaining the optimal solution for the working state and location of all APs. Implementing this invention can minimize the total AP TX power while ensuring coverage and considering the differences in user density between different areas and the temporal fluctuations of user density in the WLAN.

Description

Technical Field

[0001] This invention relates to the field of wireless network access technology, and in particular to a method for jointly optimizing the deployment location and working status of wireless access points. Background Technology

[0002] With the increasing demand for broadband internet access and the rapid development of wireless network technology, a large number of wireless local area networks (WLANs) have been deployed, bringing great convenience to network access, but also consuming a lot of energy and resulting in a lot of resource waste.

[0003] Currently, the two main reasons for energy waste in wireless LANs are as follows: First, most access points (APs) in WLANs operate at maximum transmission (TX) power (typically 100mW for indoor APs), working continuously 24 hours a day. While this maximizes network coverage, it also leads to energy waste. Furthermore, excessive TX power exacerbates signal interference. Second, all WLANs are designed and deployed based on user capacity requirements. However, in many cases, user capacity varies over time. For example, in a busy shopping mall WLAN, user density often exhibits significant periodic fluctuations, especially during shutdown periods when most APs operate under very low load, or even zero load, yet still continuously consume energy at maximum power. Given these circumstances, optimizing AP configurations—including TX power, deployment location, and operating status—has attracted considerable attention in recent years.

[0004] Regarding TX power optimization, H. Briantoro et al. proposed a method to determine the optimal AP power based on the signal-to-noise ratio of the received signal strength, thereby reducing the overall energy consumption of WLAN. M. Hmila et al. proposed a two-stage semi-distributed solution to reduce energy consumption. In the first stage, they used a cooperative alliance game framework for channel allocation. In the second stage, they used fractional programming to determine the AP's TX power. J. Guo et al. studied in detail the relationship between three key factors in wireless networks, such as energy consumption, sensor quality, and connectivity. W. Wu et al. discussed the energy efficiency of WLAN by considering congestion avoidance and migration constraints. They proposed an AP joint management framework and channel width selection for active APs in WLAN, which can effectively save energy.

[0005] Regarding AP deployment location optimization, S. Karimi-Bidhendi et al. proposed a heterogeneous two-layer Lloyd-like algorithm to optimize AP deployment in wireless networks. Taking the overall signal strength of terminal devices as the optimization objective, J. Du et al. proposed an automatic AP deployment location optimization mechanism using a heuristic algorithm. BPTewari effectively solved the problem of efficient AP placement and frequency allocation by considering two important parameters: power adjustment and partially overlapping channel allocation. T. Wen et al. proposed a method based on a brute-force search algorithm to optimize AP deployment in a communication-based train control system.

[0006] Furthermore, some works have considered both AP TX power and location optimization. To maximize the overall throughput of the wireless network, X. Zhang et al. jointly considered power allocation and AP placement. Many researchers have also chosen to use swarm intelligence optimization (SIOA) algorithms to adjust AP locations to reduce the network's overall TX power. P. Liu et al. further eliminated redundant APs through traversal checks. While these studies are effective in reducing system energy consumption, their optimization scenarios are two-dimensional (2D), and the aforementioned algorithms may not be applicable to practical three-dimensional optimization scenarios (3D). In contrast, Q. Hu et al. proposed a novel algorithm that optimizes the overall TX power of the network while ensuring effective coverage by jointly optimizing the TX power and location of each AP in a 3D scenario.

[0007] In WLANs, energy saving by adjusting the operating status of access points (APs) is a hot topic. C. Xu et al. proposed a resource reassociation scheduling algorithm based on Benders decomposition, which reduces system energy consumption by shutting down some APs while maintaining system coverage and not affecting user experience. RG Garroppo et al. proposed a method that reduces WLAN power consumption by only turning on a portion of APs and associating users with the turned-on APs. GHApostolo et al. combined network clustering and machine learning models to determine the AP status throughout the day. To ensure load balancing among APs and WLAN energy saving, MHDwijaksara et al. proposed an efficient user association scheme by adjusting the operating status of APs. RG Garroppo et al. aimed to minimize WLAN energy consumption during periods of low user density by shutting down some APs and adjusting the TX power level of each AP.

[0008] While the aforementioned work has made significant strides in optimizing AP configurations, several issues remain unresolved. Specifically, it does not account for differences in user density across different areas or temporal fluctuations in user density within the WLAN. Summary of the Invention

[0009] The technical problem to be solved by the embodiments of the present invention is to provide a method for jointly optimizing the deployment location and working status of wireless access points, which can minimize the total AP TX power while ensuring coverage, taking into account the differences in user density between different areas and the temporal fluctuations of user density in WLAN.

[0010] To address the aforementioned technical problems, embodiments of the present invention provide a method for jointly optimizing the deployment location and operational status of wireless access points, the method comprising the following steps:

[0011] Step S1: Obtain information from the wireless local area network (WLAN) that has... AP set Position vectors of each AP and work status And divided into The region of a unit cube To obtain the user density in each sub-region in, S i = 1 or 0, where 1 indicates that the AP is active and 0 indicates that the AP is dormant;

[0012] Step S2: Determine the region D of each sub-cube. j Target point T j To form a target point set And through the formula Calculate the signal from AP i To the target point T j Path loss;

[0013] Among them, T j Representing region D j The geometric center is taken as the target point to be covered, which represents region D. j The location of all users in T, and T j The position is represented as ||AP i ,T j ||2 represents AP i and T j The Euclidean distance between them; γ represents the path loss factor, which is related to the surrounding environment; x σ d represents a normally distributed random variable with standard deviation σ; o α represents the reference distance; d represents the distance. o Received power; β o Indicates the obstacle pair T j Signal attenuation;

[0014] Step S3: For each sub-cube's region D j Target point T j Overall coverage Relevance with users As constraints, construct a condition related to each AP. i TX power The relevant objective function is as follows:

[0015]

[0016]

[0017]

[0018] C1 constrains the full coverage of the target point; C2 guarantees the user capacity requirements within the service area. and P min and P max R(T) represents the minimum and maximum TX power of the same AP, respectively; j ) represents T j Can all users in the WLAN be connected to the WLAN? C(T j ) represents T j The joint coverage, and C(AP i ,T j ) represents T j Is it AP i Coverage, and Indicates AP i TX power, P 0,j T represents j The lowest received signal strength that can be covered;

[0019] Step S4: Use a single-objective optimization algorithm to find the optimal solution for the objective function and output it, that is, obtain the optimal solution for the working state and position of all APs.

[0020] A method for jointly optimizing the deployment location and operational status of wireless access points, characterized in that step S4 specifically includes:

[0021] Step 1: Initialize parameters: population size N, current iteration count k, maximum iteration count k max and network parameters; wherein the network parameters include user density. and the signal attenuation β by obstacles o ;

[0022] Step 2: Based on the individual coding method, randomly generate the initial population Y(0) = [X1,X2,…,X]. N ] T and initial solution X i The uniform distribution X in the search space i =X min +r1·(X max -X min ); where X i ={[s i,1 ,x i,1 ,y i,1 ,z i,1 ],[s i,2 ,x i,2 ,y i,2 ,z i,2 ],…,[s i,n ,x i,n ,y i,n ,z i,n ]};i represents the i-th individual in Y(0);n represents the number of APs;s represents the state of APs;x,y,z represent the positions of APs;X min and X max They represent solutions X respectively. i The search space is defined by its lower and upper bounds; r1 represents a random number between (0,1);

[0023] Step 3: According to the formula Calculate the fitness of each individual and select the individual X with the lowest fitness in the population. M and the smallest individual X FM The total energy consumption of AP in each individual is the individual's fitness.

[0024] Step 4: Control the population's exploration and utilization through escape energy (EE), and through the formula... Calculate the value of EE; if |EE|≥1, proceed to Step 5; otherwise, go to Step 6; r2 represents a random number in (0,1); c1 = 1.5;

[0025] Step 5: Exploration Phase: Obtain a random value within the range [0,1]. If this value is greater than 0.5, then use the formula... Update all X positions i (k+1); otherwise, use the formula and formula Update all X positions i (k+1);

[0026] Where r3, r4, and r5 represent random numbers between (0, 1); ⊙ represents the dot product operator, i.e., calculating R. LThe sum of the products of each component of the two vectors Y(k); R L It is an N-dimensional vector, where each element R L,i The i = 1, 2, ..., N follow the distribution as follows: And λ = 1.5 and b ~ N(0,1), and

[0027] Step 6: Utilization Phase: First, select X M and X FM As an intersecting individual, its calculation formula is: in, Indicates an interleaving operation; This represents the portion of the gene randomly selected from each individual for the crossover operation;

[0028] Then, obtain a random value within the range [0,1]. If this value is greater than 0.5, update each position using the formula X(k+1)=V(k+1)+GS1·||V(k+1),X(k)||2; otherwise, use the formula... Update each position; where ||V(k+1),X(k)||2 represents the Euclidean distance between individuals V and X; GS1~N(0,0.333) indicates that vector GS1 follows a Gaussian distribution with mean 0 and standard deviation 0.333; CL is the center of the search space, and ||CL,X M (k)||2 represents CL and X M The Euclidean distance between (k); GS2~N(0,1) indicates that the vector GS2 follows a Gaussian distribution with a mean of 0 and a standard deviation of 1;

[0029] Step 7: k = k + 1, repeat Step 2 to Step 7 until k = k max ;

[0030] Step 8: Output the final X M This means obtaining the optimal solution for all AP locations and states.

[0031] The geometric center is taken as the target point to be covered; implementing the embodiments of the present invention has the following beneficial effects:

[0032] This invention constructs a mapping relationship between each AP and the overall coverage of all user locations (i.e., target points to be covered) and user association rates within the region of each sub-cube. iThe objective function associated with the TX power is determined, and a single-objective optimization algorithm is used to find the optimal solution to obtain the optimal solution for the working state and location of all APs. This allows for minimizing the total AP TX power while ensuring coverage and considering the differences in user density between different areas and the temporal fluctuations of user density in the WLAN. Attached Figure Description

[0033] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, obtaining other drawings based on these drawings without creative effort still falls within the scope of the present invention.

[0034] Figure 1 A flowchart illustrating the method for jointly optimizing the deployment location and working status of wireless access points according to embodiments of the present invention;

[0035] Figure 2 The flowchart of step S5 of the method for jointly optimizing the deployment location and working status of wireless access points provided in the embodiments of the present invention, which uses a single-objective optimization algorithm to find the optimal solution of the objective function;

[0036] Figure 3 The simulation experiment of the method for jointly optimizing the deployment location and working status of wireless access points provided in the embodiments of the present invention is shown as a schematic diagram of the user density distribution from 7:00 AM to 11:00 PM and from 11:00 PM to 7:00 AM the next day.

[0037] Figure 4 A comparison of total TX power of different OAC methods based on SIOA (Swarm Intelligence Optimization Algorithm) in a simulation experiment of the method for jointly optimizing the deployment location and working status of wireless access points provided in the embodiments of the present invention;

[0038] Figure 5 A comparison chart of coverage rates of different SIOA-based OAC methods in a simulation experiment of the method for jointly optimizing the deployment location and working status of wireless access points provided in the embodiments of the present invention.

[0039] Figure 6 A comparison of the number of APs during non-working hours in a simulation experiment of the method for joint optimization of wireless access point deployment location and working status provided in the embodiments of the present invention;

[0040] Figure 7The simulation experiment of the method for jointly optimizing the deployment location and working status of wireless access points provided in the embodiments of the present invention shows the heat map of the received signal strength at different times; wherein, (a) is the heat map of the received signal strength during working hours from 7:00 am to 11:00 pm; (b) is the heat map of the received signal strength during rest hours from 11:00 pm to 7:00 am the next day.

[0041] Figure 8 A comparison chart of total TX power under various AP user capacities in a simulation experiment of the method for jointly optimizing the deployment location and working status of wireless access points provided in the embodiments of the present invention;

[0042] Figure 9 A comparison chart of coverage under various AP user capacities in a simulation experiment of the method for jointly optimizing the deployment location and working status of wireless access points provided in the embodiments of the present invention. Detailed Implementation

[0043] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings.

[0044] like Figure 1 As shown in the figure, this invention proposes a method for jointly optimizing the deployment location and working status of wireless access points. The method includes the following steps:

[0045] Step S1: Obtain information from the wireless local area network (WLAN) that has... AP set Position vectors of each AP and work status And divided into The region of a unit cube To obtain the user density in each sub-region in, S i = 1 or 0, where 1 indicates that the AP is active and 0 indicates that the AP is dormant;

[0046] Step S2: Determine the region D of each sub-cube. j Target point T j To form a target point set And through the formula Calculate the signal from AP i To the target point T j Path loss;

[0047] Among them, T j Representing region D j The geometric center is taken as the target point to be covered, which represents region D. jThe location of all users in T, and T j The position is represented as ||AP i ,T j ||2 represents AP i and T j The Euclidean distance between them; γ represents the path loss factor, which is related to the surrounding environment; x σ d represents a normally distributed random variable with standard deviation σ; o α represents the reference distance; d represents the distance. o Received power; β o Indicates the obstacle pair T j Signal attenuation;

[0048] Step S3: For each sub-cube's region D j Target point T j Overall coverage Relevance with users As constraints, construct a condition related to each AP. i TX power The relevant objective function is as follows:

[0049]

[0050]

[0051]

[0052] C1 constrains the full coverage of the target point; C2 guarantees the user capacity requirements within the service area. and P min and P max R(T) represents the minimum and maximum TX power of the same AP, respectively; j ) represents T j Can all users in the WLAN be connected to the WLAN? C(T j ) represents T j The joint coverage, and C(AP i ,T j ) represents T j Is it AP i Coverage, and Indicates AP i TX power, P 0,j T represents j The lowest received signal strength that can be covered; the geometric center of which is taken as the target point to be covered;

[0053] Step S4: Using a preset single-objective optimization algorithm, find the optimal solution for the objective function and output it, thus obtaining the optimal solution for the working state and position of all APs.

[0054] The specific process is as follows: In step S1, obtain the wireless local area network (WLAN) with... The set of APs, the location vectors of each AP, and their operating states are divided into... The region is divided into units of a cube to obtain the user density in each sub-region.

[0055] In step S2, firstly, the region D of each sub-cube is determined. j Target point T j This is to form a target point set, that is, to form the set of locations of all users to be covered in each sub-region.

[0056] Secondly, the calculation signal is from AP i To the target point T j The path loss is reduced to meet the requirements of subsequent objective function construction.

[0057] In step S3, for each sub-cube region D j Target point T j Given the overall coverage rate and user association rate as constraints, construct a framework for each AP. i TX power The associated objective function.

[0058] In step S4, as Figure 2 As shown, a single-objective optimization algorithm is used to find and output the optimal solution of the objective function. The specific process is as follows:

[0059] Step 1: Initialize parameters: population size N, current iteration count k, maximum iteration count k max and network parameters; wherein the network parameters include user density. and the signal attenuation β by obstacles o ;

[0060] Step 2: Based on the individual coding method, randomly generate the initial population Y(0) = [X1,X2,…,X]. N ] T and initial solution X i The uniform distribution X in the search space i =X min +r1·(X max -X min ); where X i ={[s i,1 ,x i,1 ,y i,1 ,zi,1 ],[s i,2 ,x i,2 ,y i,2 ,z i,2 ],…,[s i,n ,x i,n ,y i,n ,z i,n ]};i represents the i-th individual in Y(0);n represents the number of APs;s represents the state of APs;x,y,z represent the positions of APs;X min and X max They represent solutions X respectively. i The search space is defined by its lower and upper bounds; r1 represents a random number between (0,1);

[0061] Step 3: According to the formula Calculate the fitness of each individual and select the individual X with the lowest fitness in the population. M and the smallest individual X FM The total energy consumption of AP in each individual is the individual's fitness.

[0062] Step 4: Control the population's exploration and utilization through escape energy (EE), and through the formula... Calculate the value of EE; if |EE|≥1, proceed to Step 5; otherwise, go to Step 6; r2 represents a random number in (0,1); c1 = 1.5;

[0063] Step 5: Exploration Phase: Obtain a random value within the range [0,1]. If this value is greater than 0.5, then use the formula... Update all X positions i (k+1); otherwise, use the formula and formula Update all X positions i (k+1);

[0064] Where r3, r4, and r5 represent random numbers between (0, 1); ⊙ represents the dot product operator, i.e., calculating R. L The sum of the products of the components of two vectors, Y(k); R L It is an N-dimensional vector, where each element R L,i The i = 1, 2, ..., N follow the distribution as follows: And λ = 1.5 and b ~ N(0,1), and

[0065] Step 6: First, select X M and X FM As an intersecting individual, its calculation formula is: in, Indicates an interleaving operation; This represents the portion of the gene randomly selected from each individual for the crossover operation;

[0066] Then, obtain a random value within the range [0,1]. If this value is greater than 0.5, update each position using the formula X(k+1)=V(k+1)+GS1·||V(k+1),X(k)||2; otherwise, use the formula... Update each position; where ||V(k+1),X(k)||2 represents the Euclidean distance between individuals V and X; GS1~N(0,0.333) indicates that vector CS1 follows a Gaussian distribution with mean 0 and standard deviation 0.333; CL is the center of the search space, and ||CL,X M (k)||2 represents CL and X M The Euclidean distance between (k); GS2~N(0,1) indicates that the vector GS2 follows a Gaussian distribution with a mean of 0 and a standard deviation of 1;

[0067] Step 7: k = k + 1, repeat Step 2 to Step 7 until k = k max ;

[0068] Step 8: Output the final X M This means obtaining the optimal solution for all AP locations and states.

[0069] like Figures 3 to 9 The figure shows a simulation experiment and result comparison of a method for jointly optimizing the deployment location and working status of a wireless access point according to an embodiment of the present invention.

[0070] Figure 3 The diagram shows the distribution of user density from 7 a.m. to 11 p.m. and from 11 p.m. to 7 a.m. the next day.

[0071] Figure 4 A comparison of total TX power for different SIOA-based OAC methods; Figure 5 A comparison chart of coverage for different SIOA-based OAC methods; Figure 6 A graph comparing the number of APs during the vacation period for different SIOA-based OAC methods.

[0072] Figure 7 The heatmaps of received signal strength at different times are top views from the XY angle; where (a) is a heatmap of received signal strength during working hours from 7:00 AM to 11:00 PM; and (b) is a heatmap of received signal strength during rest hours from 11:00 PM to 7:00 AM the next day.

[0073] Figure 8 A comparison chart of total TX power under various AP user capacities.

[0074] Figure 9 This is a comparison chart of coverage under various AP user capacities.

[0075] As can be seen from the above experimental and result comparison figures, the proposed IEGJO-OAC method is highly effective. It can significantly reduce the total AP TX power of WLAN while still ensuring complete and effective coverage of the service area.

[0076] Implementing the embodiments of the present invention has the following beneficial effects:

[0077] This invention constructs a mapping relationship between each AP and the overall coverage of all user locations (i.e., target points to be covered) and user association rates within the region of each sub-cube. i The objective function associated with the TX power is determined, and a single-objective optimization algorithm is used to find the optimal solution to obtain the optimal solution for the working state and location of all APs. This allows for minimizing the total AP TX power while ensuring coverage and considering the differences in user density between different areas and the temporal fluctuations of user density in the WLAN.

[0078] Those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, such as ROM / RAM, disk, optical disk, etc.

[0079] The above description is merely a preferred embodiment of the present invention and should not be construed as limiting the scope of the invention. Therefore, any equivalent variations made in accordance with the claims of the present invention are still within the scope of the present invention.

Claims

1. A method for jointly optimizing the deployment location and working status of wireless access points, characterized in that, The method includes the following steps: Step S1: Obtain information from the wireless local area network (WLAN) that has... AP set Position vectors of each AP and work status And divided into The region of a unit cube To obtain the user density in each sub-region in, S i = 1 or 0, where 1 indicates the AP is active and 0 indicates the AP is dormant; j = D1, D2, ... Step S2: Determine the region D of each sub-cube. j Target point T j To form a target point set And through the formula Calculate the signal from AP i To the target point T j Path loss; Among them, T j Representing region D j The geometric center is taken as the target point to be covered, which represents region D. j The location of all users in T, and T j The position is represented as ||AP i ,T j ||2 represents AP i and T j The Euclidean distance between them; γ represents the path loss factor, which is related to the surrounding environment; x σ d represents a normally distributed random variable with standard deviation σ; o α represents the reference distance; d represents the distance. o Received power; β o Indicates the obstacle pair T j Signal attenuation; Step S3, for each sub-cube region D j Target point T j Overall coverage Relevance with users As constraints, construct a condition related to each AP. i TX power The relevant objective function is as follows: C1 constrains the full coverage of the target point; C2 guarantees the user capacity requirements within the service area. And P j APi =min{max{β i,j +P 0,j ,P min },P max }, P min and P max R(T) represents the minimum and maximum TX power of the same AP, respectively; j ) represents T j Can all users in the WLAN be connected to the WLAN? C(T j ) represents T j The joint coverage, and C(AP i ,T j ) represents T j Is it AP i Coverage, and Indicates AP i TX power, P 0,j T represents j The lowest received signal strength that can be covered; Step S4: Use a single-objective optimization algorithm to find the optimal solution for the objective function and output it, that is, obtain the optimal solution for the working state and position of all APs.

2. The method for jointly optimizing the deployment location and working status of a wireless access point as described in claim 1, characterized in that, The specific steps of step S4 include: Step 1: Initialize parameters: population size N, current iteration count k, maximum iteration count k max and network parameters; wherein the network parameters include user density. and the signal attenuation β by obstacles o ; Step 2: Based on the individual coding method, randomly generate the initial population Y(0) = [X1,X2,…,X]. N ] T and initial solution X i The uniform distribution X in the search space i =X min +r1·(X max -X min ); where X i ={[s i,1 ,x i,1 ,y i,1 ,z i,1 ],[s i,2 ,x i,2 ,y i,2 ,z i,2 ],…,[s i,n ,x i,n ,y i,n ,z i,n ]};i represents the i-th individual in Y(0);n represents the number of APs;s represents the state of APs;x,y,z represent the positions of APs;X min and X max They represent solutions X respectively. i The search space is defined by its lower and upper bounds; r1 represents a random number between (0,1); Step 3: According to the formula Calculate the fitness of each individual and select the individual X with the lowest fitness in the population. M and the second smallest individual X FM The total energy consumption of AP in each individual is the individual's fitness. Step 4: Control the population's exploration and utilization through escape energy (EE), and through the formula... Calculate the value of EE; if |EE|≥1, proceed to Step 5; otherwise, go to Step 6; r2 represents a random number in (0,1); c1 = 1.5; Step 5: Exploration Phase: Obtain a random value within the range [0,1]. If this value is greater than 0.5, then use the formula... Update all X positions i (k+1); otherwise, use the formula and formula Update all X positions i (k+1); Where r3, r4, and r5 represent random numbers between (0, 1); ⊙ represents the dot product operator, i.e., calculating R. L The sum of the products of each component of the two vectors Y(k); R L It is an N-dimensional vector, where each element R L,i The i = 1, 2, ..., N follow the distribution as follows: And λ = 1.5 and b ~ N(0,1), and Step 6: Utilization Phase: First, select X M and X FM As an intersecting individual, its calculation formula is: in, Indicates an interleaving operation; This represents the portion of the gene randomly selected from each individual for the crossover operation; Then, obtain a random value within the range [0,1]. If this value is greater than 0.5, update each position using the formula X(k+1)=V(k+1)+GS1·||V(k+1),X(k)||2; otherwise, use the formula... Update each position; where ||V(k+1),X(k)||2 represents the Euclidean distance between individuals V and X; GS1~N(0,0.333) indicates that vector GS1 follows a Gaussian distribution with mean 0 and standard deviation 0.333; CL is the center of the search space, and ||CL,X M (k)||2 represents CL and X M The Euclidean distance between (k); GS2~N(0,1) indicates that the vector GS2 follows a Gaussian distribution with a mean of 0 and a standard deviation of 1; Step 7: k = k + 1, repeat Step 2 to Step 7 until k = k max ; Step 8: Output the final X M This means obtaining the optimal solution for all AP locations and states.

Images