Signal Scrambling-Aided ODMA for Pilot-Free Unsourced Random Access Over Fading Channel

Abstract

This work investigates unsourced random access (URA) over quasi-static fading channels, which is well suited for massive connectivity. In most existing URA schemes for fading channels, message recovery typically requires channel estimation, and these schemes therefore rely on pilot-assisted designs, resulting in limited spectral efficiency and degraded performance. In this context, on–off division multiple access (ODMA) has emerged as a promising URA framework due to its super-sparse structure; however, existing ODMA-based schemes still rely on a single large pattern book and exhibit high computational complexity. To overcome these limitations, this paper proposes a novel URA scheme that integrates signal scrambling with an ODMA-based framework, referred to as SS-ODMA. The key idea of SS-ODMA is to introduce a user-specific polarity scrambling pattern as an auxiliary signature, which decouples the original large pattern book into two compact components and mitigates the exponential growth in detection complexity. Building on this structure, we develop a low-complexity three-stage iterative detection algorithm that enables activity detection, joint channel estimation and scrambling pattern detection, and multi-user data recovery without relying on pilot signals. Simulation results demonstrate that the proposed SS-ODMA scheme outperforms existing URA schemes over a wide range of active users while reducing computational complexity by approximately two orders of magnitude.