Design and Analysis of Routing and Scheduling Algorithms for Industrial Wireless Sensor Networks
Rajith Delanka Karunaratne Madduma-Bandarage · 2020
The oil and refinery industry has greatly benefited from process automation with Industrial Wireless Sensor Networks (IWSNs). However, the requirements are particularly strict in these environments. Network designers expect the reachability of all nodes throughout the operation. Depending on the application and stability requirements, they expect a certain end-to-end latency level. Since nodes are battery powered, it is necessary to optimize energy consumption. Thus, reliability, latency, and energy-efficiency are major concerns for IWSNs. Deployment of low-cost sensor nodes makes these networks error-prone. Due to the harsh environment, nodes are difficult to access, and it takes time to detect, isolate, and repair node failures. Thus, the impact of node failures on network performance must be assessed. This thesis designs an efficient routing/scheduling algorithm to improve the performance of IWSNs. The proposed routing algorithm, Frame Level Optimized Reliable Graph (FLO-RG), is a graph routing algorithm, which consists of primary and backup paths. A Medium Access Control (MAC) protocol is introduced to reduce the wasted energy of idle listening in backup slots. A petroleum refinery wireless sensor network model complying with WirelessHART/ ISA100.11a industrial standards is developed using the ns-3 package to evaluate the proposed FLO-RG algorithm with recent existing algorithms on the post-failure/ pre-repair scenario. Based on the results, the FLO-RG algorithm performs well on reliability, latency, energy-efficiency domains, and makes a better compromise between reliability and energy-efficiency. This thesis also investigates the impact of engineering guidelines and best practices on the performance results.