A Research Of Power Domain Non-orthogonal Multiple Access Technology For Industrial Wireless Networks

Industrial wireless networks(IWNs),which realize the collection,processing and transmission of industrial production data,are currently facing stringent requirements such as low power consumption and low delay.Non-Orthogonal Multiple Access(NOMA)technology,which superimposes multiple user information in same resource space at the transmitters,while successive interference cancellation is employed to separate them at the receiver,can realize the parallel accesses of multiple users.This thesis focuses on the applications of power domain NOMA in industrial wireless networks.Through scheduling access technologies presented in this thesis,the power consumption and the delay performance are optimized.The innovative works of this thesis are in the following two aspects;(1)Given the system uplink transmission delay bound,and under the model of imperfect SIC(Imperfect Successive Interference Cancellation),we study the problem of minimizing aggregate power consumption of transmitters by means of the joint user scheduling and power allocation strategy.Based on the problem formulation,we first present an explicit sufficient and necessary condition for the existence of the optimal solution,which is based on three key parameters of network system:residual coefficient,decoding threshold and SIC parallel degree.According to the necessary and sufficient condition,an optimal algorithm with complexity of(9)7)2)9))is proposed.Furthermore,an approximation algorithm with complexity of(9)~2)in the case of discrete transmit powers is presented,and a close approximation ratio is presented.Performance evaluations reveal that the delay bound requirement has tremendous impacts on both the aggregate power consumption and the maximum transmit power.Relative to the perfect SIC,the residual error introduced by imperfect SIC results in extra power consumption of transmitters.Fortunately,with the further relaxation of the delay bound,the total power consumption decreases exponentially.(2)Given the traffic loads of wireless sensors,the Shortest Uplink Scheduling(SUS)problem based on the joint power allocation and wireless Sensor(WS)scheduling is studied.A key term named Maximum Decoded Level(MDL),which models the transmitting capabilities of wireless sensors under SIC and thus lays the foundation for revealing a sufficient and necessary condition for successful transmissions under SIC,is presented in the first step.Then,guided by the theoretical condition which decouples WS scheduling from power allocation,a two-step greedy algorithm for the SUS problem in the case of continuous transmit powers is proposed.And,the optimality under two regular cases:2-SIC with random loads and k-SIC with unit load,is proved.Further,an optimal algorithm under 2-SIC and an approximation algorithm under k-SIC are proposed by adapting the above greedy algorithm in the case of discrete transmit powers.Experimental evaluations reveal the average frame length increases linearly with the traffic loads of WSs,and the algorithms can effectively decrease the uplink scheduling access delay than the traditional ones.