Optimal Operating Mode Selection Of Ambient Backscatter Communication Systems

With the booming development of the Internet of Things(Io T),at present,there are more and more application scenarios of wireless sensor networks.However,Io T devices mostly rely on external batteries to power communication with the network nodes.This not only causes high expense on network maintenance but also limits their working environment.As an emerging communication technology,ambient backscatter realizes communication by reflecting the existing radio frequency(RF)signals in the nearby environment as carriers and can supplement energy consumption by harvesting RF energy.Through this special communication design,it can effectively conserve the energy consumptions of small sensor network nodes.However,in the ambient backscatter communication system,when a backscatter device runs in information reflection mode,it is not able to harvest energy,and vice versa.Therefore,it is critical to choose an appropriate operating mode under the changing conditions of ambient signals.This paper studies the optimal operating mode selection of the ambient backscatter system under time-varying ambient RF signals.We consider two cases where the distribution of ambient signal is either known or not.Given the channel distribution,this paper formulates a backscatter mode selection problem based on channel probability distribution and the problem is converted to a 0-1 knapsack problem.An optimal algorithm based on dynamic programming(DP)is proposed,and according to the communication characteristics of ambient backscatter,a greedy algorithm based on a cost-effective function is proposed to obtain the approximate solution that significantly reduces its complexity while maintains satisfying performance.In addition,in the case of unknown channel distribution,this paper studies the operation model selection problem base on the reinforcement learning method.Specifically,a strategy learning algorithm base on the model-free Q-learning is proposed.Finally,the advantages of the proposed algorithms are verified by simulations,and the effectiveness and stability of the proposed algorithms are confirmed in different environments.We show that under all environment,the proposed algorithms are able to successfully improve the average communication rate under the condition that the backscatter device meets a certain average energy constraint.