Dynamic Node Lifetime Estimation for Wireless Sensor Networks

Wireless sensor networks (WSNs) consist of a large number of nodes each with limited battery power. As networks of these nodes are usually deployed unattended, network lifetime becomes an important concern. This paper proposes a novel, feasible, dynamic approach for node lifetime estimation that works for both static and dynamic loads. It covers several factors that have an impact on node lifetime, including battery type, model, brand, self-discharge, discharge rate, age, and temperature. The feasibility of the proposed scheme is evaluated by using the real testbed experiments with two wireless sensor platforms: Mica2 and N740 NanoSensor, two operating systems: TinyOS and Contiki, and different brands of alkaline and nickel-metal-hydride batteries. The deviation of the proposed estimation is in the range of -3.5%-2.5%. Three major contributions are presented in this paper: 1) the impact factors on node lifetime; 2) lifetime equations for any starting voltage, ageing, charge cycles, and temperatures; and 3) the dynamic node lifetime estimation technique, which is proposed and implemented on real hardware and software platforms in WSNs.