Hypergraph-Aided Task-Resource Matching for Maximizing Value of Task Completion in Collaborative IoT Systems

TL;DR

This paper introduces a hypergraph-based model for task-resource matching in IoT systems, unifying multiple system aspects to improve task completion efficiency amid heterogeneity and resource constraints.

Contribution

It proposes a novel hypergraph model and a game-theoretic matching algorithm to enhance task-resource allocation in IoT systems, addressing heterogeneity and complexity.

Findings

The TRM-hypergraph model outperforms existing algorithms in matching accuracy.

The game-theoretic algorithm effectively balances computational complexity and matching quality.

Numerical results confirm improved task completion performance.

Abstract

With the growing scale and intrinsic heterogeneity of Internet of Things (IoT) systems, distributed device collaboration becomes essential for effective task completion by dynamically utilizing limited communication and computing resources. However, the separated design and situation-agnostic operation of computing, communication and application layers create a fundamental challenge for rapid task-resource matching, which further deteriorate the overall task completion effectiveness. To overcome this challenge, we utilize hypergraph as a new tool to vertically unify computing, communication, and task aspects of IoT systems for an effective matching by accurately capturing the relationships between tasks and communication and computing resources. Specifically, a state-of-the-art task-resource matching hypergraph (TRM-hypergraph) model is proposed in this paper, which is used to effectively transform the process of allocating complex heterogeneous resources to convoluted tasks into a hypergraph matching problem. Taking into account computational complexity and storage, a game-theoretic hypergraph matching algorithm is proposed via considering the hypergraph matching problem as a non-cooperative multi-player clustering game. Numerical results demonstrate that the proposed TRM-hypergraph model achieves superior performance in matching of tasks and resources compared with comparison algorithms.