Search papers, labs, and topics across Lattice.
This paper introduces CARNet, a channel-adaptive neural receiver network that leverages a mixture-of-experts (MoE) framework to enhance signal detection in diverse communication scenarios. By employing multiple expert networks optimized for specific channel conditions and a lightweight routing mechanism for efficient expert selection, CARNet significantly improves generalization capabilities over traditional static networks. Experimental results indicate that CARNet outperforms existing methods in various channel conditions, showcasing its robustness and adaptability for next-generation communications.
CARNet achieves superior signal detection performance across diverse channel conditions by dynamically selecting specialized expert networks based on real-time channel estimates.
Neural receivers have been recognized as a promising paradigm for the next-generation (NextG) communications. However, due to the reliance on a static network optimized for specific channel conditions, their generalization capability across diverse scenarios remains a significant challenge. To address this issue, this paper proposes a novel channel-adaptive neural receiver network (CARNet) based on the mixture-of-experts (MoE) framework. The proposed architecture employs multiple expert networks together with an efficient routing mechanism to enable signal detection in various scenarios. The experts are constructed via stacked ResNet blocks and specialize in robust signal detection within specific channel conditions, while the routing mechanism incorporates a lightweight representation learning module, which projects the coarse channel estimate into a low-dimensional latent embedding. The learned embedding characterizes task-relevant channel conditions and provides efficient guidance for accurate expert selection. Link-level simulation experiments demonstrate that the proposed CARNet achieves superior performance across diverse channel conditions.
