Fast Uplink Grant-Free NOMA with Sinusoidal Spreading Sequences

TL;DR

This paper introduces sinusoidal spreading sequences for uplink grant-free NOMA in mMTC, enabling efficient active user detection and data decoding with lower complexity and performance guarantees, validated through extensive simulations.

Contribution

It proposes sinusoidal codes as a novel spreading sequence for grant-free NOMA, facilitating non-iterative detection algorithms with improved practicality and performance.

Findings

Sinusoidal codes enable non-iterative detection algorithms.

Performance guarantees are established for the proposed method.

Simulation results confirm effectiveness in realistic mMTC scenarios.

Abstract

Uplink (UL) dominated sporadic transmission and stringent latency requirement of massive machine type communication (mMTC) forces researchers to abandon complicated grant-acknowledgment based legacy networks. UL grant-free non-orthogonal multiple access (NOMA) provides an array of features which can be harnessed to efficiently solve the problem of massive random connectivity and latency. Because of the inherent sparsity in user activity pattern in mMTC, the trend of existing literature specifically revolves around compressive sensing based multi user detection (CS-MUD) and Bayesian framework paradigm which employs either random or Zadoff-Chu spreading sequences for non-orthogonal multiple access. In this work, we propose sinusoidal code as candidate spreading sequences. We show that, sinusoidal codes allow some non-iterative algorithms to be employed in context of active user detection, channel estimation and data detection in a UL grant-free mMTC system. This relaxes the requirement of several impractical assumptions considered in the state-of-art algorithms with added advantages of performance guarantees and lower computational cost. Extensive simulation results validate the performance potential of sinusoidal codes in realistic mMTC environments.