A Low-Complexity Recursive Approach Toward Code-Domain NOMA for Massive Communications

TL;DR

This paper introduces a recursive, low-complexity method for code-domain NOMA that uses Kronecker product factorization to reduce detection complexity and support massive connectivity in future wireless networks.

Contribution

It proposes a novel pattern matrix factorization technique that simplifies detection and enables highly overloaded systems with minimal complexity increase.

Findings

Reduced detection complexity through matrix factorization

Supports massive user connectivity with low latency

Maintains system performance with lower computational demands

Abstract

Nonorthogonal multiple access (NOMA) is a promising technology to meet the demands of the next generation wireless networks on massive connectivity, high throughput and reliability, improved fairness, and low latency. In this context, code-domain NOMA which attempts to serve $K$ users in $M\leq K$ orthogonal resource blocks, using a pattern matrix, is of utmost interest. However, extending the pattern matrix dimensions severely increases the detection complexity and hampers on the significant advantages that can be achieved using large pattern matrices. In this paper, we propose a novel approach toward code-domain NOMA which factorizes the pattern matrix as the Kronecker product of some other factor matrices each with a smaller dimension. Therefore, both the pattern matrix design at the transmitter side and the mixed symbols' detection at the receiver side can be performed over much smaller dimensions and with a remarkably reduced complexity and latency. As a consequence, the system can significantly be overloaded to effectively support the requirements of the next generation wireless networks without any considerable increase on the system complexity.