Capacity Enhancement in 5G Downlink through Non-orthogonal Spreading and Scrambling based Multiple Access (SSMA)

Abstract

Future wireless networks and 5G communication technology will have to support wide variety of applications with diverse requirement of quality of service along with the enhanced capacity. The aim is to provide seamless connectivity for geographical areas having thousands of users within hundreds of meter radius. Emerging communication technologies, specifically, multiple access techniques have to provide enhanced capacity in order to meet this objective. Non-orthogonal multiple access (NOMA) is a prominent 5G wireless communication technique which ensures higher spectral efficiency, and user fairness. Individually, each of the NOMA schemes have certain limitations on the capacity or their performance badly degrades by increasing the number of active users. Mainly, power domain NOMA fails to distinguish equidistant users around base station on the basis of power factors. To overcome these limitations and to further improve the capacity (or the number of users), a novel hybrid multiple access technique is proposed in this thesis. The technique is named as Spreading and Scrambling based Multiple Access (SSMA). The proposed scheme combines power domain NOMA (a type of scrambling-based NOMA) and Multiuser shared access (a type of spreading-based NOMA). The main objective is to provide improved capacity and better performance without compromising spectral efficiency and non-orthogonality of the resources. Since both the techniques require interference cancellation in their respective receivers, therefore, one of the major challenge while proposing this hybrid scheme is to develop efficient interference cancellation algorithm. On of the major contributions of this thesis is the development of joint interference cancellation algorithm combining both successive interference cancellation and parallel interference cancellation. The performance of the proposed scheme is analyzed and compared with MUSA and PD-NOMA using MATLAB simulations to show its superiority over both, in terms of BER and spectral efficiency.