Coding and Multi-User Schemes for Future Wireless Networks

Passcode: 899261

The ability to support variety of emerging applications in next generations of wireless networks introduces critical design challenges, in terms of satisfying stringent requirements on throughput, latency, reliability, fairness, complexity, security, massive connectivity, and power efficiency, among others. The research in this thesis aims at, first, fundamentally understanding and modeling these constraints, and, second, designing efficient protocols to address these requirements.

In the first part, we explore design of efficient coding schemes for emerging communication and computing applications. While decades of theoretical research have led to the invention of many landmark codes, we demonstrate that the design of channel coding can be automated by incorporating machine learning tools. To this end, we introduce ProductAE – a new class of neural channel codes, i.e., a family of deep-learning driven (encoder, decoder) pairs – that enables designing large neural codes with remarkable gains over state-of-the-art neural and classical codes. Next, we present a machine learning aided soft recursive algorithm for efficient and low-complexity decoding of Reed-Muller (sub-) codes. We also present efficient coding schemes for communication over emerging low-capacity channels as well as distributed computing systems.

The second part is devoted to the design of efficient multi-user networking solutions for future wireless communication applications. We first propose a low-complexity recursive approach for code-domain nonorthogonal multiple access (NOMA) with particular applications to massive communications and low-capacity scenarios. We then study uplink power-domain NOMA in the context of relaying networks with a high-throughput free-space optics (FSO) link as the backhaul link to cope with the users’ ever-increasing desire for higher data rates. Finally, we introduce covert communication in the context of millimeter-wave (mmWave) systems, not only to increase the security level of wireless applications involving critical data but also to enable secure transmission solutions for emerging mmWave communication networks.

Chair: Professor Hessam Mahdavifar