Enhancing Communication Efficiency of Semantic Transmission via Joint Processing Technique

TL;DR

This work targets improving semantic spectral efficiency in wireless networks by enabling multiple cooperative base stations to transmit semantic information to multiple users via joint processing. It introduces a DRL-guided dynamic semantic mapping and resource allocation (DSMRA) framework, which uses a generalized logistic surrogate $ ilde{\xi}_k( au_k^t,oldsymbol{w}_k^t;oldsymbol{h}_k^t)$ to approximate semantic similarity and optimizes the surrogate objective $ ilde{R}$. The approach decomposes the problem into Learning, Beamforming, and Update modules, with a bi-convex reformulation and alternating optimization to achieve a stationary solution while handling the integer mapping variable $oldsymbol{ au}^t$. Simulation results with $N=3$, $K=3$, and $M=3$ demonstrate that DSMRA attains near-Optimal Mapping Scheme performance and significantly surpasses random or fixed-mapping baselines, highlighting its practical viability for semantic communications under JP.

Abstract

This work presents a novel semantic transmission framework in wireless networks, leveraging the joint processing technique. Our framework enables multiple cooperating base stations to efficiently transmit semantic information to multiple users simultaneously. To enhance the semantic communication efficiency of the transmission framework, we formulate an optimization problem with the objective of maximizing the semantic spectral efficiency of the framework and propose a lowcomplexity dynamic semantic mapping and resource allocation algorithm. This algorithm, based on deep reinforcement learning and alternative optimization, achieves near-optimal performance while reducing computational complexity. Simulation results validate the effectiveness of the proposed algorithm, bridging the research gap and facilitating the practical implementation of semantic communication systems.

Figures (4)

• Figure 1: System model with $K=N=3$.
• Figure 2: The schematics of the proposed DSMRA algorithm.
• Figure 3: The convergence for the DSMRA algorithm.
• Figure 4: Performance evaluation of the proposed algorithm and baselines.