A graph coloring energy efficiency balancing method based on competitive AP selection

Abstract

This invention discloses a graph coloring-based energy efficiency balancing method for AP selection based on competitive AP selection. The invention selects APs in descending order of their large-scale fading coefficients. After competing with other users in the same set for an AP, users with lower large-scale fading coefficients are eliminated and their APs are blacklisted. Users within different AP sets are sequentially filtered by index to ensure no user is overlooked. Furthermore, graph coloring is used for pilot allocation to reduce pilot pollution. This invention ensures that users with lower large-scale fading coefficients are not overlooked during AP selection, and that an AP is available to serve them. The graph coloring-based pilot allocation further reduces pilot pollution among users, improving both frequency efficiency and energy efficiency of the system.

Description

Technical Field

[0001] This invention belongs to the field of information and communication engineering technology. It designs a decellularized massive multiple input multiple output (MIMO) technology in wireless communication systems, specifically a graph coloring energy efficiency balancing method based on contention for AP selection. Background Technology

[0002] In recent years, with the increasing demands of communication equipment for network capacity and data throughput, traditional cellular communication systems have shown limitations and are gradually failing to meet higher-quality communication requirements. Decellularized massive MIMO technology has emerged to address this need, improving the system's spectral efficiency (SE) and energy efficiency (EE) by fully utilizing spatial resources. In decellularized massive MIMO systems, users can establish connections with multiple access points (APs) and collaboratively process signals from multiple APs, significantly improving communication quality. However, this collaborative processing can lead to inter-user interference and excessive power loss. Therefore, to optimize system performance and reduce the load on the backhaul link, it is necessary to select the AP subset with the best service quality for each user.

[0003] To address the above issues, some scholars have proposed using the signal-to-interference-plus-noise ratio (SINR) as a performance indicator to screen the subset of APs with the best service quality. However, this approach easily overlooks users with poor channel quality, resulting in no APs serving these users. Others have used K-means to cluster users in a static state, but this approach clearly does not conform to the actual conditions in real life where users are in motion. Summary of the Invention

[0004] To address the interference problem among users in decellularized massive MIMO systems, this invention provides a graph coloring energy efficiency balancing method based on contention for access points (APs). It ensures that each user can access at least one AP to obtain network service, with a total of M APs. It assumes that the number of users each AP can connect to is subject to a pre-set limit, i.e., the number of users served by each AP is a fixed value U. max The competitive strategy is that when user k' needs to find a nearby AP with better channel conditions, if the AP is already serving user U... max One user, user k' needs to be with U. max Several users compete for AP service, k * Let A represent the subset of users served by the m-th AP. m Users with the worst channel conditions will inevitably be subset A. m middle user k* Remove and add the m-th AP to the subset The purpose is to prevent k * The omission resulted in no AP serving it, subset The AP in the middle is no longer for user k * Service, when This indicates that only one AP remains to serve user k. * Service. Initially, each user will look for the AP with the best nearby channel conditions, i.e. This AP is called the backup AP for user k', and M k' This represents the set of access points (APs) that serve user k'.

[0005] The specific steps of the technical solution adopted by the present invention to solve its technical problem are as follows:

[0006] Step (1) For user k', first determine the selectable AP subset G based on the large-scale fading coefficient. Its elements are M APs that remove the AP serving user k' and APs that user k' has blacklisted.

[0007] G = {1…M} / (M) k' ∪B k' (1)

[0008] Among them, B k' This indicates the AP that has been blacklisted by user k';

[0009] Step (2) User k' searches for the AP with the largest large-scale fading coefficient in subset G;

[0010]

[0011] The two cases are as follows: If the m-th AP, i.e., has already served U... max One user, i.e., |A m |=U max Then proceed to step (3); if |A m |<U max , directly stored in set M k' in.

[0012] Step (3) When a new user joins, they need to interact with user subset A. m When other users within the channel compete for access, priority is given to users with better channel conditions, and the user with the smallest large-scale fading coefficient, k, is removed. * The details are as follows:

[0013]

[0014] in, This represents the user served by the m-th AP. Large-scale fading coefficient. If Return to step (2) and continue searching for the next AP that can be added in subset G; otherwise, k * ∈A m User k * Put the m-th AP into middle.

[0015] like This indicates that all APs have been selected, and If the user k' to be removed has m as the backup AP, then it cannot be removed from subset A. m Remove user k' from the list.

[0016] Step (4) returns to step (1) and continues searching for the next user.

[0017] The set of APs e serving the k'-th user is determined through AP selection. k' e k' ={e k'1 ,…,e k'M}, set e k' Place it in the corresponding row of the service matrix Φ, the k'th row of the service matrix Φ. k' ={θ1,…,θ k' Let} represent the AP serving user k'. From the service matrix Φ, we can determine whether any two users share the same AP. An interference matrix D∈C is generated based on the service matrix between users. K×K .

[0018]

[0019] If any two users within the interference matrix D share a common AP, their corresponding element is 1, and they need to be assigned orthogonal pilots during pilot allocation. Conversely, if they do not share an AP, it means that the two users do not share an AP and can be assigned the same pilot.

[0020] To minimize pilot pollution, a graph coloring algorithm is used to allocate pilots. The main principles are: (1) If two users share an AP, they are assigned different colors and orthogonal pilots are allocated. (2) All users in the area are covered using the fewest possible color types. The optimization of selecting the fewest color types is expressed as:

[0021]

[0022] In the above formula, n * ν represents the number of colors used for coloring, i.e., the orthogonal pilot number. kThis represents the color assigned to the k-th user, i.e., the pilot. This can lead to a pilot being reused multiple times, resulting in other pilots not being fully utilized. Furthermore, n may be generated during the coloring process. * >τ p Situation, τ p This represents the total number of pilot signals. To address the potential scenarios described above, a constraint value n' is used to limit the maximum number of times a pilot signal can be used, and n' ≤ K / τ. p K represents the total number of users.

[0023] The beneficial effects of this invention are as follows:

[0024] The advantages of this invention are that it does not overlook users with small large-scale fading coefficients when competing for AP services, and ensures that there are APs to serve them; furthermore, it reduces pilot pollution among users by using graph coloring for pilot allocation, thereby improving the frequency efficiency and energy efficiency of the system. Attached Figure Description

[0025] Figure 1 This is the decellularized large-scale MIMO system model structure of the algorithm of this invention.

[0026] Figure 2 The given values ​​are M=100, K=30, N=1, and τ. p CDF curves of uplink throughput for each user under different pilot allocation algorithms when =10.

[0027] Figure 3 M=100, N=1, τ p When the number of users is 10, the impact of different numbers of users on the system frequency efficiency.

[0028] Figure 4 M=100, N=1, τ p When the number of users is 10, the impact of different numbers of users on system energy efficiency. Detailed Implementation

[0029] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0030] This invention discloses an energy-efficient balancing method for competitive AP selection in a decellularized massive MIMO system. APs compete for users in descending order of their large-scale fading coefficients. After competing with other users in the same set, users with lower large-scale fading coefficients are eliminated and their APs are blacklisted. Users within different AP sets are sequentially filtered by index to ensure no user is missed. Furthermore, pilot allocation using graph coloring reduces pilot pollution. This invention not only suppresses pilot pollution but also achieves superior energy efficiency balancing.

[0031] like Figure 1 As shown, the method of the present invention is specifically implemented as follows:

[0032] Step (1) For user k', first determine the selectable AP subset G based on the large-scale fading coefficient. Its elements are M APs that remove the AP serving user k' and APs that user k' has blacklisted.

[0033] G = {1…M} / (M) k' ∪B k' (1)

[0034] Among them, B k' This indicates the AP that has been blacklisted by user k';

[0035] Step (2) User k' searches for the AP with the largest large-scale fading coefficient in subset G;

[0036]

[0037] The two cases are as follows: If the m-th AP, i.e., has already served U... max One user, i.e., |A m |=U max Then proceed to step (3); if |A m |<U max , directly stored in set M k' in.

[0038] Step (3) When a new user joins, they need to interact with user subset A. m When other users within the channel compete for access, priority is given to users with better channel conditions, and the user with the smallest large-scale fading coefficient, k, is removed. * The details are as follows:

[0039]

[0040] in, This represents the user served by the m-th AP. Large-scale fading coefficient. If Return to step (2) and continue searching for the next AP that can be added in subset G; otherwise, k * ∈A m User k * Put the m-th AP into middle.

[0041] like This indicates that all APs have been selected, and If the user k' to be removed has m as the backup AP, then it cannot be removed from subset A. m Remove user k' from the list.

[0042] Step (4) returns to step (1) and continues searching for the next user.

[0043] The set of APs e serving the k'-th user is determined through AP selection. k' e k' ={e k'1 ,…,e k'M}, set e k' Place it in the corresponding row of the service matrix Φ, the k'th row of the service matrix Φ. k' ={θ1,…,θ k' Let} represent the AP serving user k'. From the service matrix Φ, we can determine whether any two users share the same AP. An interference matrix D∈C is generated based on the service matrix between users. K×K .

[0044]

[0045] If any two users within the interference matrix D share a common AP, their corresponding element is 1, and they need to be assigned orthogonal pilots during pilot allocation. Conversely, if they do not share an AP, it means that the two users do not share an AP and can be assigned the same pilot.

[0046] To minimize pilot pollution, a graph coloring algorithm is used to allocate pilots. The main principles are: (1) If two users share an AP, they are assigned different colors and orthogonal pilots are allocated. (2) All users in the area are covered using the fewest possible color types. The optimization of selecting the fewest color types is expressed as:

[0047]

[0048] In the above formula, n * ν represents the number of colors used for coloring, i.e., the orthogonal pilot number. k This represents the color assigned to the k-th user, i.e., the pilot. This can lead to a pilot being reused multiple times, resulting in other pilots not being fully utilized. Furthermore, n may be generated during the coloring process. * >τ p Situation, τ p This represents the total number of pilot signals. To address the potential scenarios described above, a constraint value n' is used to limit the maximum number of times a pilot signal can be used, and n' ≤ K / τ. p K represents the total number of users.

[0049] Experimental parameter results:

[0050] like Figure 2 The given values ​​are M=100, K=30, N=1, and τ. p CDF curves of uplink throughput for each user under different pilot allocation algorithms when =10.

[0051] like Figure 3 M=100, N=1, τp When the number of users is 10, the impact of different numbers on system frequency efficiency is shown. Figure 4 M=100, N=1, τ p The impact of different numbers of users on system energy efficiency when the system capacity is 10. Figure 3 and Figure 4 It is evident that the proposed method represents a significant improvement over the competition-free AP algorithm.

Claims

1. A graph coloring energy efficiency balancing method based on competitive AP selection, characterized in that... Includes the following steps: Step (1) For user k', first determine the selectable AP subset G based on the large-scale fading coefficient. Its elements are M APs that serve user k' and APs that have been blacklisted by user k'. Step (2) User k' searches for the AP with the largest large-scale fading coefficient in subset G; Step (3) When a new user joins, they need to interact with user subset A. m When other users within the channel compete for access, priority is given to users with better channel conditions, and the user with the smallest large-scale fading coefficient, k, is removed. * ; Step (4) returns to step (1) and continues searching for the next user.

2. The graph coloring energy efficiency balancing method based on competitive AP selection according to claim 1, characterized in that... To ensure that each user can access at least one access point (AP) to obtain network service, the total number of APs is M; assuming that the number of users each AP can connect to is subject to a pre-set limit, i.e., the number of users served by each AP is a fixed value U. max The competitive strategy is that when user k' needs to find a nearby AP with better channel conditions, if the AP is already serving user U... max One user, user k' needs to be with U. max Several users compete for AP service, k * Let A represent the subset of users served by the m-th AP. m Users with the worst channel conditions will inevitably be subset A. m middle user k * Remove and put the m-th AP into subset B k* The purpose is to prevent k * The omission resulted in no AP serving it, subset B k* The AP in the middle is no longer for user k * Service, when This indicates that only one AP remains to serve user k. * Service; initially, each user will look for the AP with the best nearby channel conditions, i.e. This AP is called the backup AP for user k', and M k' This represents the set of APs that serve user k'.

3. The graph coloring energy efficiency balancing method based on competitive AP selection according to claim 2, characterized in that... Step (1) is implemented as follows: G={1,…,M} / (M k' ∪B k' ) (1) Among them, B k' This indicates the AP that has been blacklisted by user k'; 4. The graph coloring energy efficiency balancing method based on competitive AP selection according to claim 3, characterized in that... Step (2) is implemented as follows: If the m-th AP, i.e., has already served U max One user, i.e., |A m |=U max Then proceed to step (3), if |A m |<U max , directly stored in set M k 'in.

5. The graph coloring energy efficiency balancing method based on competitive AP selection according to claim 4, characterized in that... Step (3) is implemented as follows: in, This represents the user served by the m-th AP. Large-scale fading coefficient; if Return to step (2) and continue searching for the next AP that can be added in subset G; otherwise, k * ∈A m User k * Put the m-th AP into middle.

6. The graph coloring energy efficiency balancing method based on competitive AP selection according to claim 5, characterized in that... like This indicates that all APs have been selected, and If the user k' to be removed has m as the backup AP, then it cannot be removed from subset A. m Remove user k' from the list.

7. The graph coloring energy efficiency balancing method based on competitive AP selection according to claim 6, characterized in that... The set of APs e serving the k'-th user is determined through AP selection. k' e k' ={e k'1 ,…,e k'M }, subset e k' Place it in the corresponding row of the service matrix Φ, the k'th row of matrix Φ. k' ={θ1,…,θ k' Let} represent the AP serving user k'. From the matrix Φ, we can determine whether any two users share the same AP. An interference matrix D∈C is generated based on the service matrix between users. K×K ; Within the interference matrix D, any two users served by a shared AP have a corresponding element of 1, and orthogonal pilots must be assigned during pilot allocation. Conversely, if two users do not share an AP, they can be assigned the same pilot. To minimize pilot pollution, a graph coloring algorithm is used for pilot allocation. The optimal solution for selecting the fewest color types is as follows: In the above formula, n * ν represents the number of colors used for coloring, i.e., the orthogonal pilot number. k This represents the color assigned to the k-th user, i.e., the pilot. This can lead to a pilot being reused multiple times, resulting in underutilization of other pilots. Furthermore, n may be generated during the coloring process. * >τ p Situation, τ p This represents the total number of pilot signals. To address the possible scenarios described above, a constraint value n' is used to limit the maximum number of times a pilot signal can be used, and n' ≤ K / τ. p K represents the total number of users.

Images