Research On Dynamic Decode Ordering SIC Receiver For Uplink NOMA

To satisfy the demand of high spectral efficiency,there has been tremendous interest in developing non-orthogonal multiple access(NOMA)techniques.The key idea of NOMA is to exploit the power domain multiple access via multiplexing multiple users at the same resource.It is proved feasible to obtain higher spectrum efficiency and to make more users accessed simultaneously.However,new challenges occur,where strong inter-user interference is introduced and how to accommodate more users to meet the quality of service(e.g.the outage performance).Therefore,it is of great importance to excavate the potential of NOMA systems in these two aspects.Numerous research activities have been initiated to conquer the serious inter-user interference in the received signal.A successive interference canceller(SIC)is utilized to separate superimposed multiuser messages in NOMA systems.Whereas,a fixed decoding order based on the long term statistic channel state information is adopted for uplink NOMA systems in many previous recearches.From the principle of the SIC receiver,the user information with higher signal strength should be detected first by treating the other users' information as the interference.However,the fixed decoding order adopted is not optimal for uplink NOMA which needs further investigation.To overcome these drawbacks,we propose a dynamic ordered SIC decoding mechanism,based on the instantaneous received signal strength,in uplink NOMA systems.By this mechanism,the signal that brings the highest inter-user interference can be removed as early as possible.An outage balancing among different users as well as lower outage probabilities can be achieved compared with the fixed ordered SIC,as shown in the simulation results.We theoretically analyze the dynamic decoding mechanism in terms of the outage performance.Modeling the fading channel as independent but non-identically distributed which is more practical.Accordingly,we obtain not only the conditioned outage probabilities under certain orders,but also the overall outage performance for each specific user with different required data rate considering all possible decoding orders.Moreover,most of the previous literatures assume perfect knowledge of the channel state information(CSI).This is not practical,because the channel estimation with high accuracy will cause significant system overhead.For our proposed scheme,to estimate the instantaneous received signal power for each user,the estimation of CSI is necessary.Therefore,the influence of the incorrect CSI estimation in uplink NOMA system is critical problem to be further analyzed.For the above reason,we derive the closed-form expressions of the outage probability for the dynamic-ordered decoding scheme based on the instantaneous signal strength with imperfect CSI.Shown by the simulation,this mechanism is better than the fixed-ordered scheme with the same channel estimation quality,especially when the channel estimation error is not too severe.We investigate the impact of imperfect CSI on uplink NOMA system by categorizing the effect into two classes: introducing of extra interference and disturbance of decoding order determination.As shown in the simulation results,introducing extra interference is the main impact of imperfect CSI,but the incorrect decoding order cannot be ignored,especially in poor channel estimation cases where the effect of the incorrect decoding order becomes dominant.