User-Centric Stream Sensing for Grant-Free Access: Deep Learning with Covariance Differencing

TL;DR

This work tackles collisions in grant-free uplink access under overloaded conditions where the number of active streams exceeds the sensing antennas. It introduces a user-centric framework that reframes the problem as detecting the incremental change $d$ in active streams using covariance differencing across two windows and a deep learning classifier to mitigate finite-sample noise, avoiding total-stream enumeration. Theoretical analysis provides a channel-correlation-based bound on SCM deviation, guiding window sizing, while a two-window sensing pipeline and a DL classifier fuse raw and differential spectral features. Empirical results demonstrate robust, high-accuracy detection across i.i.d. Rayleigh and 3GPP TR 38.901 TDL-A channels, with significant gains over non-DL baselines and strong resilience to overload and frequency-selective fading.

Abstract

Grant-free (GF) access is essential for massive connectivity but faces collision risks due to uncoordinated transmissions. While user-side sensing can mitigate these collisions by enabling autonomous transmission decisions, conventional methods become ineffective in overloaded scenarios where active streams exceed receive antennas. To address this problem, we propose a differential stream sensing framework that reframes the problem from estimating the total stream count to isolating newly activated streams via covariance differencing. We analyze the covariance deviation induced by channel variations to establish a theoretical bound based on channel correlation for determining the sensing window size. To mitigate residual interference from finite sampling, a deep learning (DL) classifier is integrated. Simulations across both independent and identically distributed flat Rayleigh fading and standardized channel environments demonstrate that the proposed method consistently outperforms non-DL baselines and remains robust in overloaded scenarios.

Figures (6)

• Figure 1: System model for user-centric stream sensing in a GF network by monitoring the number of active GF UEs.
• Figure 2: The proposed stream sensing method based on covariance differencing and a deep learning model.
• Figure 3: Time-frequency resource grid structure for sampling two consecutive sensing windows.
• Figure 4: Frequency-domain correlation of the 3GPP TR 38.901 TDL-A channel versus subcarrier index difference $\Delta f$.
• Figure 5: Sensing accuracy versus the SNR with $K_t = 4$.