Semantic-Aware Power Allocation for Generative Semantic Communications with Foundation Models

Chunmei Xu,Mahdi Boloursaz Mashhadi,Yi Ma,Rahim Tafazolli

2024-10-08

Abstract:Recent advancements in diffusion models have made a significant breakthrough in generative modeling. The combination of the generative model and semantic communication (SemCom) enables high-fidelity semantic information exchange at ultra-low rates. A novel generative SemCom framework for image tasks is proposed, wherein pre-trained foundation models serve as semantic encoders and decoders for semantic feature extractions and image regenerations, respectively. The mathematical relationship between the transmission reliability and the perceptual quality of the regenerated image and the semantic values of semantic features are modeled, which are obtained by conducting numerical simulations on the Kodak dataset. We also investigate the semantic-aware power allocation problem, with the objective of minimizing the total power consumption while guaranteeing semantic performance. To solve this problem, two semanticaware power allocation methods are proposed by constraint decoupling and bisection search, respectively. Numerical results show that the proposed semantic-aware methods demonstrate superior performance compared to the conventional one in terms of total power consumption.

Signal Processing

What problem does this paper attempt to address?

This paper aims to solve the problem of how to minimize the total power consumption during the transmission process while ensuring the semantic performance of images. Specifically, the paper proposes a generative semantic communication framework for image tasks based on pre - trained base models. This framework utilizes pre - trained base models as semantic encoders and decoders to perform semantic feature extraction and image regeneration respectively. By modeling the mathematical relationship between transmission reliability and the perceived quality of the regenerated image and defining the semantic value of the extracted features, the paper further explores the power allocation problem of semantic perception. The goal is to minimize the total power consumption while ensuring semantic performance. To this end, the paper proposes two semantic - perception - based power allocation methods: the constraint decoupling method and the bisection search method. Numerical results show that the proposed semantic - perception methods are superior to traditional methods in terms of total power consumption.