Abstract:
Non-orthogonal multiple access (NOMA) has been considered as one of the key technique for 5G wireless network systems. Here similar to the previous generation multiple access techniques, the available transmission bandwidth is divided into bands of different frequency. All the users in the system are allotted into the sub bands and transmitted by being superimposed and a power allocation is carried out. And at the receiver end, a special type of processing is carried out called Successive interference cancellation (SIC) to detect the desired signal which has been differentiated by the power allotted to it. In this dissertation, a sub-optimal user grouping scheme is being implemented by exploiting the differences in channel conditions between the base station and various users in a cluster and then allots them into sufficient number of clusters to maximize the throughput. This is done with the objective of enhancing the sum-throughput of the system. After cluster formation, we then derive the optimal power allocation scheme that enhances the sum-throughput per NOMA cluster and therefore in turn maximizes the overall throughput of the NOMA system. Closed-form solutions for optimal power allocations are then derived for any cluster size, for downlink NOMA systems using Karush-Kuhn-Tucker optimality conditions.
