Multi-source Scheduling and Resource Allocation for Age-of-Semantic-Importance Optimization in Status Update Systems

TL;DR

This work targets semantic-aware status updates by introducing Age of Semantic Importance (AoSI), a metric that combines information freshness with semantic loss through $\Delta^{AoSI}(t)=\Delta^{AoI}(t)\cdot\psi(u(t))$. The authors formulate a long-term AoSI minimization problem for multi-source scheduling and semantic-resource allocation and solve it with a DQN-based policy deployed at an edge server, where the state includes per-source AoSI, AoI, and estimated SNR, and actions specify the scheduled source and per-word semantic symbol count $k_{m,n}$. The proposed AoSI-aware joint SRA method converges and outperforms baselines such as Random, Round-robin, AoI-aware, and AoSI-aware scheduling, with performance affected by the number of sources and the sampling interval; an optimal sampling interval is observed near $\tau=0.1$ s. These results demonstrate that integrating semantic similarity into freshness-focused metrics yields substantial gains for timely and reliable semantic communication in edge-enabled networks.

Abstract

In recent years, semantic communication is progressively emerging as an effective means of facilitating intelligent and context-aware communication. However, current researches seldom simultaneously consider the reliability and timeliness of semantic communication, where scheduling and resource allocation (SRA) plays a crucial role. In contrast, conventional age-based approaches cannot seamlessly extend to semantic communication due to their oversight of semantic importance. To bridge this gap, we introduce a novel metric: Age of Semantic Importance (AoSI), which adaptly captures both the freshness of information and its semantic importance. Utilizing AoSI, we formulate an average AoSI minimization problem by optimizing multi-source SRA. To address this problem, we proposed a AoSI-aware joint SRA algorithm based on Deep Q-Network (DQN). Simulation results validate the effectiveness of our proposed method, demonstrating its ability to facilitate timely and reliable semantic communication.

Figures (5)

• Figure 1: Semantic-aware multi-source status update system.
• Figure 2: Evolution of AoI and AoSI.
• Figure 3: Convergence performance of the proposed AoSI-aware joint SRA.
• Figure 4: Average AoSI versus different number of sources.
• Figure 5: Average AoSI versus different sampling interval.