AI & ML
Drift-Aware Adaptive Semantic Codec Update Strategy for Dynamic Semantic Communications
arXiv math.IT0 views
arXiv:2611.23722v1 Announce Type: new Abstract: Semantic communication reduces transmission overhead by conveying task-relevant semantic features, but its performance depends on the consistency between training and target-domain data distributions. In dynamic environments, distribution drift can make offline-trained semantic codec models stale and degrade transmission performance. Frequent updates alleviate this degradation but increase communication, computation, and queueing costs, whereas infrequent updates delay model adaptation. To address this tradeoff, we propose a drift-aware adaptive update strategy for dynamic semantic communication systems. Drift intensity is quantified in the semantic feature space, and Drift-Aware Age of Model (DA-AoM) is introduced to jointly capture drift-induced performance degradation, drift-dependent update service delay, and conventional update latency. Closed-form average DA-AoM expressions are derived for edge-layer M/M/1 and cloud-layer D/M/1 update scenarios, based on which the update frequency is optimized under the queueing-stability constraint. Simulations show that DA-AoM captures the joint effects of drift intensity, computing rate, bandwidth, and update frequency on model freshness beyond conventional Age of Information (AoI). The proposed strategy reduces average DA-AoM by 9.2\%--42.4\% compared with representative baselines under the same resource constraints.
Source:https://arxiv.org/abs/2611.23722PreviousThe Double Entropy Fit Index (DEFI): An Information-Theoretic Approach to Dimensionality Reduction and ClusteringNextThe Geometry of Mode Sampling in Reconfigurable MIMO Channel Extrapolation
Related posts
The draft is not the send
0dev.to Discuss

My First GitHub Issue & My First Forem (DEV.to) Issue 🚀
0dev.to Discuss
Multimodel AI for Business: How to Build a Reliable Video Content Workflow
0dev.to Productivity
Your Thanksgiving flight home is on the busiest travel day of the year
0dev.to Productivity
Audit Week: When Your AI Assistant Knows Your Website Better Than You
0dev.to Productivity

Telegram เตรียม Dictator: AI Agent บน Mac ที่รายงานว่ารันบน Hermes Agent
0dev.to Machine Learning