Resource Allocation For SWIPT-Enabled NOMA Systems

Recently,the great development of the communication technology industry has prompted the explosive growth of mobile devices and the rapid rise of new application scenarios such as Internet of Things,which has led to the contradiction between the growing demands for communication and the increasingly lack of radio resources.Therefore,the wireless communication systems are required to provide much greater performance.On the one hand,spectrum resource is limited,and the low frequency band that is suitable for wireless communication is extremely scarce.Hence,it is a crucial issue to improve the spectrum efficiency(SE)of communication systems.On the other hand,the enormous energy consumption and environment pollution caused by the communication industry will become increasingly serious.Therefore,it is urgent that immediate action is taken to enhance the energy efficiency(EE)of communication systems.Compared to the traditional orthogonal multiple access(OMA),non-orthogonal multiple access(NOMA)is capable of improving SE and expanding connectivity of systems,resulting from the elimination of the requirement for orthogonality of transmission channel among different users.On the other hand,simultaneous wireless information and power transfer(SWIPT)is able to achieve wireless information transmit(WIT)and meanwhile perform wireless energy transfer(WPT)by exploiting the energy that carried in the radio frequency(RF)signal,thereby reducing the actual energy consumption and enhancing the EE of systems.Hence,with the combination of spectrum-efficient NOMA and energy-efficient SWIPT,the resource allocation algorithm for SWIPT-enabled NOMA systems are investigated in this thesis,and the main content and achievement are summarized as follows:(1)Joint resource allocation optimization problem is addressed to maximize the EE of SWIPT-enabled single-carrier NOMA(SC-NOMA)system.Firstly,the system model of the downlink of SWIPT-enabled SC-NOMA is constructed,in which the total transmit power of base station is limited.To maximize the EE of the considered system,the joint resource allocation optimization problem is mathematically formulated with the constraints of the minimum demands for communication quality and energy harvesting.In particular,considering the difference between the SWIPT-enabled system and the traditional system,that is,the harvested energy is able to compensate the energy consumption,a more practical definition of power consumption as well as EE of the considered system is given in this thesis.To deal with the EE maximization problem which is of coupled variables and non-convex,a dual-layer iterative algorithm is proposed to optimize the resource allocation,and on the basic of this,Dinkelbach's method is applied to solve the fractional programming sub-problem.Numerical results verify the convergence and effectiveness of the proposed algorithm,and also prove the superiority of SWIPT-enabled NOMA system to SWIPT-enabled OMA or traditional OMA without SWIPT in terms of EE.(2)Joint resource allocation optimization problem is studied to maximize the throughput of SWIPT-enabled multiple-carrier NOMA(MC-NOMA)system.In order to further improve the performance of the system,based on the advantage of Orthogonal Frequency Division Multiple Access(OFDMA),we construct SWIPT-enabled OFDMA-NOMA system,i.e.,SWIPT-enabled MC-NOMA system.Considering the downlink of the aforementioned system model,the joint resource allocation optimization problem is formulated to maximize the achievable data rate of the system whilst satisfying the constraints of the maximum transmit power of base station and the minimum requirement for harvested energy of each user.Owing to the presence of the coupled variables and the inter-user interference,the considered optimization problem is non-convex and hence is challenging to be solved directly.To tackle the problem,a dual-layer iterative algorithm is developed to optimize the resource allocation,which is verified to be convergent and effective.In addition,an alternative approach based on deep belief network(DBN)is proposed to approximate the optimal solution of the considered optimization problem.Based on the larger number empirical data and the well-trained DBN,the deep learning-based approach is of critical significance for the communication systems that are expected to achieve low latency.The effectiveness of the learning-based approach is also verified in the simulation part.More importantly,the simulation results demonstrate the superiority of the SWIPT-enabled MC-NOMA system to the SWIPT-enabled SC-NOMA and SWIPT-enabled OMA systems in terms of improving the achievable data rate and SE.