Mobile edge computing system improves task accuracy and handles faults
Fault-Tolerant Offloading Framework for Real-Time Applications in Mobile Edge Computing
Distributed, Parallel, and Cluster Computing
Summary
Offloading tasks from mobile devices to nearby computing servers can speed up apps but faces limits on network and computing power. The authors study how to best assign tasks while meeting strict deadlines and resource constraints to improve accuracy in real time. They show the problem is very hard to solve exactly, so they create a fast approximation algorithm with guaranteed performance. They also design a way to detect and recover from network and server failures to keep the system reliable. Tests show their method improves accuracy more than others and keeps working well even when faults occur.
What this means in practice
- •For mobile app developers: Improve real-time app performance by efficiently offloading tasks to edge servers with guaranteed accuracy and fault tolerance.
- •For network schedulers: Optimize bandwidth and computational resources allocation in edge networks to maximize timely task completions and reduce performance loss during failures.
Authors
Chuanchao Gao, Yiyang Gao, Michael Yuhas, Arvind Easwaran
Abstract
This paper studies a Deadline-constrained task Offloading and resource Allocation Problem (DOAP) in a backhaul-network-enabled Mobile Edge Computing (MEC) system with both bandwidth and computational resource constraints. The goal is to maximize the total inference accuracy improvement across all tasks. We formulate this problem, denoted as $\mathbf{P}$, as an Integer Linear Programming (ILP) problem and show it is NP-hard via a reduction from the maximum-weight three-dimensional matching problem. To address $\mathbf{P}$, we develop $\mathtt{FastSA}$, a linear-time $\frac{1}{10}$-approximation algorithm based on the local-ratio technique. To the best of our knowledge, $\mathtt{FastSA}$ provides the first constant-factor performance guarantee for backhaul-network-enabled DOAPs. In addition, we introduce a fault detection and recovery framework that enables robust MEC operation under backhaul link and edge server failures. Experiments using profiled image classification workloads and real-world taxi traces show that $\mathtt{FastSA}$ achieves an average accuracy improvement of $25.6\%$ over benchmark algorithms under fault-free conditions while maintaining the lowest runtime. Under fault-injected scenarios, our framework reduces performance degradation by an average of $12.8\%$, demonstrating its effectiveness in fault handling.