Research On Resource Allocation Schemes In Cellfree Massive MIMO System

In the past few years,due to the rapid development of emerging technologies such as Internet of Things,cloud computing,big data and virtual reality,the demands for large connections,low latency,and high-speed data traffic by individuals and enterprises have also greatly increased.The fifth generation(5G)mobile communication system provides strong support for solving these problems.However,the cellular network architecture has been used from the first generation mobile communication system to the current 5G.From macro cells to micro cells,with the area of cells gradually shrinking,the problems of inter-cell interference and hand-off are more serious,which bring challenges to improve system performance.The proposal of the cell-free network architecture makes a breakthrough for solving these bottlenecks,and it is expected to be one of the key technologies for the future mobile communication system.Cell-free massive multiple input multiple output(MIMO)system is an implementation of cell-free architecture.In cell-free massive MIMO system,a large number of access points(APs)are deployed and they connect to a central processing unit(CPU)through backhaul links.All APs can use the same time and frequency resources to serve each user simultaneously.If the system resources are simply allocated to these APs,they can't be utilized efficiently and the high quality of service can't also be guaranteed.In order to utilize system resources more efficiently and further improve system performance,this thesis studies the resource allocation schemes of cell-free MIMO system from the perspective of user rate.Firstly,for the full-pilot zero-forcing(fpZF)receiver,the problem of analog-to-digital converters(ADCs)equal quantization bits allocation in the uplink cell-free massive MIMO system is studied.In order to reduce the overhead of backhaul link,the quantization process is introduced in the system.With low-resolution ADCs at the APs,these APs can quantify the received signals to carry out the tractable analysis on multiple quantization bits.Different from the conventional zero-forcing(ZF)receiver,the fpZF receiver is adopted to detect the data sent by users in the system.The fpZF receiver can perform signal detection tasks at the APs with no need to transmit the received signals to the CPU via backhaul link for centralized processing.During the research,a closed-form expression of achievable uplink rate is derived,which helps us to make a comparison with the conjugate beamforming(CB)receiver counterpart.The simulation results show that when the equal quantization bits allocation scheme is used,the fpZF receiver is more sensitive to quantization operations than the CB receiver.Secondly,for fpZF and normalized conjugate beamforming(NCB)precoders,the ADC quantization bits allocation scheme and power allocation scheme in the downlink cell-free massive MIMO system are studied.In order to analyze the impact of the quantization process on the downlink user rate,the low-resolution ADCs are configured both at the APs and the users.The ADC quantization bits allocation problem is first studied.Unlike the equal quantization bits allocation scheme,a new quantization bits allocation scheme is proposed under the assumption that perfect ADCs are used at the users and the total number of quantization bits allocated to all ADCs at the APs is fixed.According to this scheme,the quantization bit allocated to each ADC can be optimized to further increase the total downlink user rate.Then the weighted max-min power optimization problem is studied,which aims to increase the lower user rate and to provide users with uniform and good quality of service.It also can set different weighted figure to meet the higher priority users with higher rate requirements.Finally,the simulation results verify the effectiveness of these two resource allocation schemes,which can significantly improve the performance of cell-free massive MIMO system.