Power-saving queueing model of a wireless sensor node operating in two modes subject to random occurrences of sleep orders

This paper analyzes the pursuance of a wireless sensor node that works in two modes: fast and slow. The node is also subject to random occurrences of sleep orders from the base station. The service rate in the slow phase is less than in fast mode. Customers arrive at the node only when the server is in fast mode. In fast or slow mode, the available customers in the server are served on a first-come, first-served discipline. If no further customers are present for service, the server immediately enters sleep mode and resides in it for a random time. All arriving customers are lost during the sleep mode. At the end of sleep mode, the server moves immediately to fast mode to serve the arriving customers. The server in fast mode becomes sick randomly, and the sick server immediately switches to slow mode and available customers are served. The sensor node experiences random occurrences of sleep orders in slow mode only, and at the occurrence of a sleep order, all customers are washed out, and the sensor moves to sleep mode. Steady-state probabilities for system size and some system performance measures are explicitly obtained. A numerical illustration of the impact of sleep orders on system performance is provided.