Research On Uplink Multi-Cell Virtual MIMO System

Multiple input multiple output(MIMO)technology can improve the spectrum utilization of the system effectively and satisfy the communication needs of people.However,in the SC-FDMA uplink system,MIMO technology cannot be applied to the system effectively due to the limitations of the volume of the user terminal and the capacity of the battery.So the researchers proposed a virtual multipleinput multipleoutput(VMIMO)technology.By combining multiple users with one antenna,VMIMO can regard multiple users with one antenna as one device with multiple antennas.The antennas of the communication entity is combined with multiple antennas of the base station to form multiple antenna arrays,which can get the effect of MIMO technology.The key research point of VMIMO technology is to perform user pairing.In the SC-FDMA uplink system,the performance of the communication system is related to the allocation of spectrum resources closely.As the cell radius decreases,the cellular base station deployed more densely,the system spectrum utilization can be improved significantly by using multi-cell cooperation technology.However,VMIMO technology also faces some difficult problems.That is,when the number of users that can be paired increases,the complexity of the pairing process increases significantly.Therefore,the main content of this paper is to design the multi-cell joint user pairing and resource allocation problem model and propose an efficient algorithm to solve the problem.This paper first gives a multi-cell VMIMO system model.Next,this paper introduces several basic user pairing algorithms for VMIMO systems and the resource allocation algorithms for SC-FDMA uplink system.And this paper analyzes the advantages and disadvantages of these user pairing algorithms and resource allocation algorithms.Then,based on the existing technology,this paper proposes a multi-cell joint user pairing and resource allocation mathematical model,which the BER proformace is considered.This model takes into account user pairing of the VMIMO system,resource allocation in the SC-FDMA uplink system and multi-cell cooperation technolgy.By using the proposed mathematical model,the spectral efficiency while ensuring the BER performance of the system can be maximized.Then,inspired by the K-Means clustering algorithm in machine learning,this paper proposes an iterative K-Groups algorithm which can quickly solve multi-cell joint user pairing and resource allocation problems.The algorithm uses the K-Means algorithm clustering algorithm as the theoretical basis,decomposes the large-scale problem of multi-cell joint user pairing and resource allocation into several small-scale sub-problems to solve in parallel,the search space for solving the problem is greatly reduced.However,since the K-Groups algorithm can not always obtain the optimal solution,there will be some reductions in spectral efficiency,so in order to compensate for the reduced spectral efficiency,an ordered successive interference cancellation technique based on minimum mean square error(MMSE-OSIC)is used for detection.So,an iterative K-Groups algorithm which based on minimum mean square error and ordered successive interference cancellation technology(IKG_OSIC)is obtained.Finally,in the SC-FDMA uplink system,this paper sets some simulations to compare the proposed IKG_OSIC algorithm with some existing joint user pairings and resource allocation algorithms.The simulation results show that the proposed IKG_OSIC algorithm can quickly solve the problem model and obtain higher spectral efficiency while ensuring the BER proformancce.