A Bayesian Tensor Approach to Enable RIS for 6G Massive Unsourced Random Access

TL;DR

This paper introduces a Bayesian tensor-based method for joint device detection and channel estimation in RIS-assisted 6G unsourced random access, leveraging tensor decomposition and probabilistic modeling for improved performance.

Contribution

It proposes a novel probabilistic model and a coupled tensor-based algorithm that automatically detects active devices and estimates channels efficiently in large-scale RIS scenarios.

Findings

The proposed CTAD algorithm achieves high detection accuracy.

It automatically learns the number of active devices.

Simulation results show significant performance improvements.

Abstract

This paper investigates the problem of joint massive devices separation and channel estimation for a reconfigurable intelligent surface (RIS)-aided unsourced random access (URA) scheme in the sixth-generation (6G) wireless networks. In particular, by associating the data sequences to a rank-one tensor and exploiting the angular sparsity of the channel, the detection problem is cast as a high-order coupled tensor decomposition problem. However, the coupling among multiple devices to RIS (device-RIS) channels together with their sparse structure make the problem intractable. By devising novel priors to incorporate problem structures, we design a novel probabilistic model to capture both the element-wise sparsity from the angular channel model and the low rank property due to the sporadic nature of URA. Based on the this probabilistic model, we develop a coupled tensor-based automatic detection (CTAD) algorithm under the framework of variational inference with fast convergence and low computational complexity. Moreover, the proposed algorithm can automatically learn the number of active devices and thus effectively avoid noise overfitting. Extensive simulation results confirm the effectiveness and improvements of the proposed URA algorithm in large-scale RIS regime.