Saved in:
Bibliographic Details
Main Authors: Li, Muqing, Li, Ning, Yuan, Xin, Xu, Wenchao, Chen, Quan, Guo, Song, Zhang, Haijun
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2508.09208
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912535558488064
author Li, Muqing
Li, Ning
Yuan, Xin
Xu, Wenchao
Chen, Quan
Guo, Song
Zhang, Haijun
author_facet Li, Muqing
Li, Ning
Yuan, Xin
Xu, Wenchao
Chen, Quan
Guo, Song
Zhang, Haijun
contents The proliferation of large language models (LLMs) has driven the adoption of Mixture-of-Experts (MoE) architectures as a promising solution to scale model capacity while controlling computational costs. However, deploying MoE models in resource-constrained mobile edge computing environments presents significant challenges due to their large memory footprint and dynamic expert activation patterns. To address these challenges, we propose a novel dynamic resource-aware collaborative optimization framework that jointly optimizes expert aggregation granularity and offloading strategies based on real-time device resource states, network conditions, and input characteristics in mobile edge environments, denoted as CoMoE. In CoMoE, we first systematically analyze existing expert aggregation techniques, including expert parameter merging,knowledge distillation,and parameter sharing decomposition, identifying their limitations in dynamic mobile environments.We then investigate expert offloading strategies encompassing expert prediction and prefetching, expert caching and scheduling, and multi-tier storage architectures, revealing the interdependencies between routing decisions and offloading performance.The CoMoE incorporates adaptive scheduling mechanisms that respond to user mobility and varying network conditions, enabling efficient MoE deployment across heterogeneous edge devices. Extensive experiments on real mobile edge testbeds demonstrate that CoMoE achieves approximately 70% reduction in memory usage compared to baseline methods, 10.5% lower inference latency than existing expert offloading techniques, while maintaining model performance stability. For large-scale MoE models (e.g,7.4B-parameter Switch-Base-128), the CoMoE reduces memory requirements from 15.6GB to 4.7GB, enabling deployment on resource-constrained mobile edge devices that previously could only support much smaller models.
format Preprint
id arxiv_https___arxiv_org_abs_2508_09208
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle CoMoE: Collaborative Optimization of Expert Aggregation and Offloading for MoE-based LLMs at Edge
Li, Muqing
Li, Ning
Yuan, Xin
Xu, Wenchao
Chen, Quan
Guo, Song
Zhang, Haijun
Networking and Internet Architecture
Artificial Intelligence
The proliferation of large language models (LLMs) has driven the adoption of Mixture-of-Experts (MoE) architectures as a promising solution to scale model capacity while controlling computational costs. However, deploying MoE models in resource-constrained mobile edge computing environments presents significant challenges due to their large memory footprint and dynamic expert activation patterns. To address these challenges, we propose a novel dynamic resource-aware collaborative optimization framework that jointly optimizes expert aggregation granularity and offloading strategies based on real-time device resource states, network conditions, and input characteristics in mobile edge environments, denoted as CoMoE. In CoMoE, we first systematically analyze existing expert aggregation techniques, including expert parameter merging,knowledge distillation,and parameter sharing decomposition, identifying their limitations in dynamic mobile environments.We then investigate expert offloading strategies encompassing expert prediction and prefetching, expert caching and scheduling, and multi-tier storage architectures, revealing the interdependencies between routing decisions and offloading performance.The CoMoE incorporates adaptive scheduling mechanisms that respond to user mobility and varying network conditions, enabling efficient MoE deployment across heterogeneous edge devices. Extensive experiments on real mobile edge testbeds demonstrate that CoMoE achieves approximately 70% reduction in memory usage compared to baseline methods, 10.5% lower inference latency than existing expert offloading techniques, while maintaining model performance stability. For large-scale MoE models (e.g,7.4B-parameter Switch-Base-128), the CoMoE reduces memory requirements from 15.6GB to 4.7GB, enabling deployment on resource-constrained mobile edge devices that previously could only support much smaller models.
title CoMoE: Collaborative Optimization of Expert Aggregation and Offloading for MoE-based LLMs at Edge
topic Networking and Internet Architecture
Artificial Intelligence
url https://arxiv.org/abs/2508.09208