Optimal Energy-efficient Resource Allocation For Wireless-powered Information Transmission In D2D-aided NOMA Networks

In order to overcome the “doubly near-far” obstacle of wireless power transfer(WPT)and guarantee the Qo S of the far-away UEs,this paper investigates energy-efficient resource allocation for device-to-device(D2D)communications with non-orthogonal multiple access(NOMA)technique in wireless cellular networks,where all uplink user equipments(UEs)transmit information based on energy harvested from a base station(BS)via WPT.To achieve high system energy efficiency,a far-away UE exploits D2 D communication to communicate with a near UE and near UE transmits data of both UEs to the BS with NOMA technique.In a single-pair D2 D communication system,the formulated target problem is non-convex due to self-interference of NOMA scheme.Exploiting the fractional programming theory and difference of convex(DC)programming,we obtain a sub-optimal resource allocation policy.Our solution reveals that the optimal energy-efficient D2 D communication should occur if the far D2 D user transmits information to the near D2 D user with the minimal required rate.For more general cellular and D2 D communication hybrid network systems,we also get the problem formulated into a non-convex optimization problem(bits/Joule)with the goal of maximizing energy efficiency of the D2 D system.In order to ensure the basic communication performance of the UEs,we also take into account the minimum required harvested energy and the minimum required quality-of-service(Qo S)of far-away UEs.Exploiting the fractional programming theory and monotonic optimization,we characterize the optimal solution structure of joint time,frequency,and power allocation and to derive an iterative algorithm based on Branch and Reduced Bound(BRB)for obtaining the optimal solution.In order to reduce the computational complexity of the proposed optimal algorithm,we also propose a low-complexity suboptimal algorithm to achieve a good balance between calculation time and system performance.The simulation results show that compared with two baseline schemes,both the proposed optimal and suboptimal algorithm can improve the energy efficiency substantially,while the suboptimal algorithm performs closely to the optimal performance.