Design And Implementation Of AP Selection Algorithm For Decellularized Massive MIMO System With Maximum Signal-to-interference-noise Rati

Cell free massive MIMO system eliminates the concept of a cell and has multiple wireless access points(APs),providing higher high-frequency spectral efficiency and energy efficiency.Compared with traditional cellular networks,it can not only solve the problems of edge cells,frequent user switching,and inter cell co frequency interference,but also improve the system capacity limited by cell splitting limitations.The existing research on cell free massive MIMO technology focuses on AP selection,pilot pollution,and power allocation,among which AP selection is particularly important.There are a large number of APs in a cell free massive system,and each AP serves all users,which will increase the burden on the forward link.Additionally,due to the power limitations of the APs themselves,adopting a full-service approach is not feasible.Therefore,how to choose AP to achieve higher data rates for user services is a key issue.This article addresses this issue and proposes two AP selection algorithms considering both static and dynamic user scenarios.This paper improves the static AP selection algorithm based on user grouping and signal-to-noise ratio(SINR)for static user scenarios in cell free massive MIMO systems.Firstly,the K-means clustering algorithm is used to group users within the region,and during grouping,the elbow method is proposed to determine the number of clusters to ensure a more reasonable number of user groups;Then,based on the location information and required service quality of each user group,select an optimal AP service group.During the selection process,the signal quality between the AP and the user is judged based on the signal-to-noise ratio;Finally,this paper also proposes a load balancing strategy,setting a threshold for the number of AP connected to avoid situations where an AP is overloaded and to ensure stable and reliable signal transmission.Simulation shows that compared with the large-scale fading AP selection algorithm based on user grouping,this proposed algorithm can improve spectral efficiency by about 10%.This paper improves the AP dynamic selection algorithm based on collaborative benefits to provide services for mobile user scenarios in cell free massive MIMO systems.This algorithm updates the Access Point Group(APG)that serves users in realtime based on their mobile status and surrounding environment.Firstly,initialize the AP group that serves the user;Secondly,during the user movement process,the APG is updated,and a two-level candidate AP group selection method is proposed for the update of the AP.Firstly,a circular neighborhood candidate AP set is generated based on distance constraints,and then the Candidate Access Points Grouping(CAPG)in the users’ movement direction is further selected based on angle constraints and received signal power threshold in the middle;Then update the APG group based on the collaboration gain in the candidate AP set CAPG,add new APs that meet service requirements,and delete APs that do not meet the conditions and may cause interference;Finally,delete the identifier and members of the APG for users who have left the communication network.The experimental results show that the algorithm proposed in this paper can better adapt to user mobility compared to fully connected algorithms and AP selection algorithms based on whale swarm reinforcement learning,and has improved system performance in terms of user rate and spectral efficiency.