Batch Denoising for AIGC Service Provisioning in Wireless Edge Networks

TL;DR

The paper addresses the challenge of provisioning AIGC services over wireless edge networks where content is diffusion-model generated on-edge and transmitted to mobile devices under end-to-end delay constraints. It introduces a batch denoising framework and a two-stage optimization: first, a STACKING-based batch denoising algorithm that exploits parallelism while prioritizing early denoising steps, and second, a PSO-based bandwidth allocation to balance service quality and latency. The problem is reformulated into a fixed-bandwidth batch-denoising subproblem (P2) and a bandwidth-optimization subproblem (P1), enabling tractable solutions; the batch-denoising model uses $g(X_n)=aX_n+b\|X_n\|_0$ and defines $D_k^{cg}$ accordingly. Results show the approach yields superior mean Fréchet Inception Distance (FID) and end-to-end latency performance across varying numbers of services and delay requirements, validating its practical impact for scalable AIGC provisioning at the network edge.

Abstract

Artificial intelligence-generated content (AIGC) service provisioning in wireless edge networks involves two phases: content generation on edge servers and content transmission to mobile devices. In this paper, we take image generation as a representative application and propose a batch denoising framework, followed by a joint optimization of content generation and transmission, with the objective of maximizing the average AIGC service quality under an end-to-end service delay constraint. Motivated by the empirical observations that (i) batch denoising effectively reduces per-step denoising delay by enhancing parallelism and (ii) early denoising steps have a greater impact on generation quality than later steps, we develop the STACKING algorithm to optimize batch denoising. The STACKING operates independently of any specific form of the content quality function and achieves lower computational complexity. Building on the batch solution, we further optimize bandwidth allocation across AIGC services. Simulation results demonstrate the superior performance of our algorithm in delivering high-quality, lower-latency AIGC services.

Figures (2)

• Figure 1: Measured data and fitting curves: (a) Denoising delay vs. batch size; (b) FID score vs. denoising steps.
• Figure 2: Simulation results: (a) End-to-end delay illustration; (b) Mean FID sore vs. service number; (c) Mean FID score vs. minimum delay requirement.