Abstract
•Linear consecutively-connected system is considered.•Each node of the system has a resource generator and storage.•An algorithm for evaluating instantaneous availability of nodes is suggested.•The instantaneous connectivity of the system is obtained.•The optimal storage allocation problem is formulated and solved.
Motivated by real-world applications like wireless sensor networks powered by photovoltaic sources, this paper models a linear consecutively connected system whose nodes form a linear sequence. To provide the connectivity between the first (source) and last (destination) nodes, each non-destination node hosts a connecting element characterized by a different connection range and time-to-failure distribution. To supply resource needed for the connecting element's operation, each node also contains a resource generating subsystem and a storage, which is used for saving surplus resource when productivity of the resource generator exceeds the connecting element's demand and can also supply resource to the connecting element when the resource generator fails or its productivity becomes insufficient to meet the demand. A numerical algorithm is first put forward to evaluate the instantaneous availability of an individual connecting element. A universal generating function-based approach is further proposed to evaluate the instantaneous connectivity of the considered system with unreliable resource generators and storages. The optimal storage allocation problem and impacts of several parameters on the system connectivity and optimal solutions are investigated through a detailed case study of a wireless sensor network with four types of storages, characterized by different time-to-failure distributions, initial and maximum capacities, and maximum uploading and downloading rates.