Resource Allocation For Wireless Power Transfer(WPT)-Based Internet Of Things

The popularization and application of Io T(Internet of Things)technology provides high requirement to the performance of wireless communication system.The development of 5G communication technology has become the major driver of the growth of Io T application.For example,wireless power transfer(WPT)technology proposes an effective solution for the energy supply problem of Io T nodes in complex and diverse environments.With WPT technology,Io T devices receive energy from radio-frequency signal to prolong their battery lifetime.Non-orthogonal multiple access(NOMA)technology allows multiple users sharing a common channel,so as to improve spectrum efficiency effectively.On the other hand,thanks to the excellent flexibility and low cost of unmanned aerial vehicles(UAVs)communication technology,we can avoid the difficulty and high cost of traditional base station deployment.In this paper,we apply NOMA and UAV communication technology in WPT-based Io T system and study the resource allocation strategy in different scenarios.The main content and achievement are summarized as follows:(1)We study the resource allocation problems of NOMA-SWIPT enabled Io T system.Firstly,we build an Io T downlink communication system that can achieve collaborative information and energy transmission by combing NOMA with SWIPT technology.On the constraints of the maximum transmitted power of base station and minimum requirements of data-transfer rate and energy harvesting(EH)rate of Io T devices,we propose a resource optimization problem that aims to maximize total transmission rate and total harvested power simultaneously by jointly optimizing the power allocation scheme of base station and power splitting strategy of receivers.Since the optimization objectives are conflict with each other and their dimensional are inconsistent,we transform the harvested power into data rate equivalently using Shannon formula,and then reformulate the considered multi-objective optimization(MOO)problem into a single-objective optimization(SOO)problem with scalarization method.The optimization variables of the considered problem are coupling and non-convex with respect to the optimization objective.To this end,we decompose the SOO problem into two concave maximization problems with single variable,and then put forward a joint resource allocation algorithm based on alternating optimization.These two sub-problems are solved iteratively through Lagrange duality theorem and the properties of the convex function respectively.Simulation results verify the convergence and effectiveness of the proposed algorithm.It also illustrates that significant performance gain can be achieved by considering the optimization of multiple performance objects simultaneously.(2)We research on the trajectory planning problem of unmanned aerial vehicles(UAVs)-assisted wireless powered communication(WPC)Io T system.Firstly,we build an Io T communication system where an UAV is applied as a mobile access point to process data collection and energy transmission for ground devices.On this basis,a MOO problem that aims to simultaneously maximize sum data-rate as well as total harvested energy of Io T system and minimize total flight energy consumption of UAV over a mission period is proposed.Since the Io T system is dynamic and uncertain,a MODDPG(Multi-objective DDPG)algorithm is proposed to realize the real-time trajectory planning of UAV under multiple optimization objectives based on the framework of deep reinforcement learning(DRL).Finally,the evaluation results of trained network prove that the proposed algorithm achieves joint optimization of three objectives.Besides,optimal policies can be adjusted according to weight parameters among optimization objectives.