Research And Design Of Pilot Assignment Schemes For Cell-free Massive MIMO Systems

The evolution from 1G(First Generation)to 5G(Fifth Generation)is based on cellular network.In order to improve the system capacity,we need to reduce the cell area and deploy base stations intensively.However,due to the lack of cooperation between base stations,inter-cell interference is becoming more and more serious,especially for users close to the edge of the cell.The wireless communication system needs to provide unified and good services for all users in the designated area.Therefore,Cell-Free Massive MIMO(multiple-input multiple-output)system as a 6G candidate technology has attracted great attention.Cell-Free Massive MIMO has completely changed the cellular network architecture and can effectively solve the bottleneck of spectrum efficiency and energy efficiency faced by mobile communication.Like Massive MIMO system,due to the limited channel coherence time and a large number of users,it is impossible to assign completely orthogonal pilot sequences for all users in Cell-Free Massive MIMO.Users have no choice but to reuse pilot resources,which results in pilot contamination.However,the existing pilot decontamination methods can not make a compromise between system performance and complexity.This paper aims to study the pilot decontamination problem of Cell-Free Massive MIMO systems,two pilot assignment strategies are proposed in order to reduce the pilot contamination and improve the accuracy of channel estimation,so as to improve the spectral efficiency of Cell-Free Massive MIMO system.The main contributions and innovations of this paper are listed as follows.First,in order to effectively reduce pilot contamination in Cell-Free Massive MIMO,this paper proposes a novel pilot assignment strategy based on quantum bacterial foraging optimization to maximize the average achievable downlink throughput.A Cell-Free Massive MIMO with conjugate beamforming on the downlink and matched filtering on the uplink is considered.We derive rigorous closed-form capacity lower bounds for the Cell-Free Massive MIMO downlink and uplink with finite numbers of APs and users.In this scheme,we first utilize adaptive quantum rotation gate to imitate chemotaxis.Then,the health of bacterium is introduced to record the cumulative fitness sum of each bacterium.Moreover,we consider elimination-dispersal operation in order to avoid local optimum.Finally,the quantum bacterial population collapses to the ground state population by measuring to obtain the pilot assignment solutions.Numerical results demonstrate the superiority of the proposed scheme to other traditional schemes and validate that the QBFO-based scheme can significantly improve system’s average downlink throughput.This scheme achieves a good throughput-complexity trade-off for the Cell-Free Massive MIMO systems and is feasible in practice.Second,in order to effectively reduce pilot contamination in Cell-Free Massive MIMO,this paper proposes a pilot assignment scheme based on K-means clustering and soft pilot reuse to maximize the total capacity of the system.A Cell-Free Massive MIMO system with conjugate beamforming on the downlink and matched filtering on the uplink is analyzed.In this scheme,we first utilize K-means clustering method to group users according to their location coordinates.Then,calculate the threshold of each cluster and divide the users into central users and edge users.Finally,the soft pilot reuse strategy is used for pilot assignment.Numerical results manifest that the scheme based on K-means clustering and soft pilot reuse can significantly improve the overall system performance.This scheme achieves a good throughput-complexity trade-off for the Cell-Free Massive MIMO systems and is feasible in practice.