Convergecast scheduling and cost optimization for industrial wireless sensor networks with multiple radio interfaces

Abstract

Industrial wireless sensor networks have been widely deployed in many industrial systems. The main communication paradigm of such systems, known as convergecast, is to converge sensing data to a centralized manager. The rapid and reliable data convergecast is essential to the industrial production. Multiple radio interfaces on a network device and convergecast scheduling algorithms can effectively reduce convergecast delay. Existing works confine to the convergecast based on linear- and tree-based routing. Compared to the two routing schemes, graph routing is more reliable. Although the graph routing gains more popularity in industrial networks due to its better reliability, few works have addressed its temporality performance. On the other hand, the number of radio interfaces also impacts on the convergecast delay. In this paper, we present a holistic framework to solve how to use multiple radio interfaces to converge data. First, we propose a convergecast scheduling algorithm for industrial wireless sensor networks with multiple radio interfaces. Second, based on our proposed scheduling algorithm, we propose an optimal algorithm and a fast heuristic algorithm to minimize the number of radio interfaces under the temporality constraint of industrial production. Evaluations show that all our algorithms perform closely to the optimal solution.