Linear system design with application in wireless sensor networks

Abstract Wireless sensor networks (WSN) are essential technologies of modern smart industry and their reliability is crucial for industrial information integration. This paper contributes by modeling the reliability of a linear multi-state consecutively connected system (LMCCS), which consists of a WSN providing the signal source, some intermediate nodes, and a sink node. The wireless sensors (WS) constituting the WSN are deployed around monitored objects to sense their conditions and transmit the sensed signal to a base station. The base station of the WSN serves as the source node of the LMCCS. The intermediate nodes of the LMCCS are sequentially arranged along a line to provide coverage and transmission of signal from the WSN base station to the sink node. Each node may assume multiple states, characterized by different signal coverage and transmission ranges of the connection element (CE) deployed in the node. As the state of each node is random, the signal coverage of the entire LMCCS is also random. Existing works on LMCCSs have mainly focused on studying the continuity of signal transmission (in particular, the probability that the sink node or signal receiver can successfully receive the signal from the source node or signal emitter). However, in some practical applications (e.g., risk source monitoring systems), the signal coverage area is a more important performance metric than the signal receiving probability. This paper proposes a universal generating function-based method to model and analyze the reliability and expected signal coverage of the LMCCS with WSN. The optimal CEs allocation problem is then solved with the objective of minimizing the cost of the entire system while meeting pre-specified requirements on the expected signal coverage and system reliability. The CEs allocation policy obtained should strike the balance among the system reliability, signal coverage and cost.