Toward Performant and Energy-Efficient Network Queries: A Parallel and Stateless Approach

Abstract

In many edge computing applications (e.g., wireless sensor networks, WSNs) where nodes are mostly battery-powered, queries’ energy consumption, and response time are two of the most important metrics, representing the network’s sustainability and performance, respectively. Conventional techniques used to focus on only one of the two metrics and did not attempt to optimize both in a coordinated manner. This work aims to achieve both high sustainability and high performance of these queries at the same time. To that end, a new mechanism is proposed to construct the topology of a three-tier WSN. The proposed mechanism eliminates the conventional stateful routing tables and employs a stateless and efficient addressing scheme inspired by the Chinese remainder theorem (CRT). The CRT-based topology allows for query parallelism, an unprecedented feature in the WSNs. On top of the new topology encoded by CRT, a new protocol is designed to parallelly preprocess collected data on sensor nodes by effectively aggregating and deduplicating data in a cluster of neighborhood nodes. At the same time, the hibernating mechanism is proposed to prolong the network life cycle. Moreover, a new algorithm is devised to allow the queries and results to be transmitted through low-power and fault-tolerant paths using recursive elections over a subset of the entire power range. With all these new techniques taken together, the system presented in this work outperforms approximate algorithms from various perspectives: (i) the query response is improved by up to 21.6%; (ii) the energy consumption is reduced by up to 16.8%; and (iii) the reliability is increased by up to 18.3%.