Research On Resource Allocation Strategies In Massive MIMO Enabled SWIPT Networks

With the continuous improvement of the level of social information,the wireless communication business is explosively growing and thus the demand for wireless communication service is also increasing rapidly.However,the existing wireless communication networks have failed to meet the growing communication requirements and so their network performance needs to be improved urgently.On the one hand,massive multi-input multi-output(MIMO)can greatly improve spectral efficiency and energy efficiency without increasing transmitted power and network bandwidth.As a result,it has received wide attention and research in academia and industry and is considered as one of key technologies in future wireless communication networks.On the other hand,the power supply of terminals directly affects the normal operation of the network and thus users need to recharge or replace the batteries frequently for terminals.However,it is impossible for all cases such as military communications,submarines,and medical implants.To solve this problem,simultaneous wireless information and power transfer(SWIPT)is nominated as the good choice because of its feasibility,controllability and stability.SWIPT can carry both information and power using the same radio frequency signal and thus has gained strong research interest from researchers.With the objective to realize common advantages of massive MIMO and SWIPT,many key issues need to be addressed.Among them,how to efficiently optimize network resources such as power,frequency,time and space,and achieve a further improvement of network performance is one of valuable research topics.This is also the research problem of this thesis.This thesis takes massive MIMO and SWIPT as the core,and aims to improve network performance.Regarding network resource optimization as the key point of breaking through,the work in this thesis mainly studies network resource allocation in four scenarios of massive MIMO SWIPT.The main contributions are summarized as follows:Firstly,resource allocation strategy in the relay-assisted massive MIMO SWIPT networks is investigated.The common assumption from the majority of recent studies on SWIPT is that the base station(BS)can communicate with terminals directly.The scenario that there are no direct links between BS and terminals because of huge obstacles is ignored.Aiming at this problem,this thesis considers the scheme of relay-assisted SWIPT transmission and analyzes the optimization problem of the rate fairness among terminals.When the time switching(TS)protocol works at the relay node,the closed-form expression of achievable rate for each terminal is first derived,and then the joint optimization problem of power allocation coefficients and TS factor is formulated.Similarly,when the power splitting(PS)protocol works at the relay node,the joint optimization problem of power allocation coefficients and PS coefficient is formulated.Finally,as the two optimization problems are non-convex and hard to solve directly,two iteration optimization algorithms are proposed for solving them respectively.Numerical results manifest that the scheme of relay-assisted SWIPT transmission is feasible,and that the two proposed algorithms can guarantee the rate fairness among terminals.Moreover,compared with TS based SWIPT,PS based SWIPT gets a higher quality of service in the considered network.Secondly,resource allocation strategy in the distributed massive MIMO SWIPT networks is investigated.Aiming at the problem that the energy transfer efficiency of SWIPT is very low due to the long-distance transmission when terminals are very far from BS,this thesis analyzes the performance of the distributed massive MIMO SWIPT networks.As the remote radio heads(RRHs)are more arbitrarily distributed over the network,and the distance from any given terminal to its nearby RRH is much smaller,the above-mentioned problem will disappear.First of all,the closed-form expression on the ergodic capacity for each terminal is derived.Then,the ergodic sum capacity maximization optimization problem with respect to RRH-terminal association,channel estimation duration,PS coefficients and power allocation is designed.The designed optimization problem is proved to be a mixed-inter nonlinear programming with combinatorial variables,which is generally hard to solve due to its NP-hardness.Finally,an iteration optimization algorithm utilizing decomposition technique is proposed.Numerical results show that the proposed algorithm can maximize the ergodic sum capacity.Compared with the centralized version of massive MIMO networks,the distributed version of massive MIMO networks can significantly improve the energy transfer efficiency of SWIPT.Thirdly,resource allocation strategy based on the hybrid protocol transmission in the massive MIMO SWIPT networks is investigated.Based on the existing TS protocol and PS protocol,this thesis introduces a novel hybrid wireless energy harvesting protocol,which is a combination of TS protocol and PS protocol.As a result,a general network model is developed and a general analytical framework is formulated for the single-cell networks.Then the joint spectral efficiency of uplink and downlink maximization problem with respect to channel estimation duration,power allocation,TS factor,PS coefficients and energy allocation ratios is formulated.As the formulated optimization problem is non-convex,it is hard to solve directly.To provide a solution,an iteration optimization algorithm is proposed.Numerical results manifest that the proposed algorithm can maximize the joint spectral efficiency of uplink and downlink,and that this work can help to study three energy harvesting protocols(TS,PS,and hybrid protocol)and three transmission modes(uplink,downlink,and joint uplink and downlink)by the setting of the relevant parameters.This brings great convenience for the comprehensive network performance analysis in the single-cell massive MIMO SWIPT networks.Finally,resource allocation strategy based on the hybrid signal transmission in the massive MIMO SWIPT networks is investigated.The majority of recent studies on SWIPT are based on wireless energy harvesting protocol.The attention to SWIPT of the hybrid signal transmission is very little.Based on this observation,this thesis investigates the efficiency of information transfer(Eo IT)for information terminals and the efficiency of power transfer(Eo PT)for energy terminals in massive MIMO networks.First of all,the ergodic capacity expression for each information terminal and the harvested energy expression for each energy terminal are derived respectively.According to these expressions,two optimization problems with the purposes of maximizing the Eo IT and maximizing the Eo PT are formulated respectively.As the two formulated optimization problems belong to classical fractional programming ones and hard to solve directly,two iteration optimization algorithms are proposed to solve them by a series of transforms.Numerical results manifest that the two iteration optimization algorithms can provide the optimal Eo IT and Eo PT,which is in line with the development trend of green communications.