Research article

Optimal power consumption control of sensor node based on (N, D)-policy discrete-time queues

¹ School of Science, Chongqing University of Technology, Chongqing 400054, China
² Department of Electrical, Electronic and Computer Engineering, University of Pretoria, 0002 Pretoria, South Africa
³ Department of Electrical and Computer Engineering, University of Manitoba, R3T 5V6 Winnipeg, Manitoba, Canada
⁴ School of Science, Sichuan University of Science and Engineering, 643000 Zigong, Sichuan, China

^* Corresponding author: liurb888@126.com

Received: 19 October 2016
Accepted: 19 April 2017

Abstract

In this paper, we consider two types of power consumption control policies for the long lifetime of wireless sensor node based on the discrete-time Geo/G/1 queue. One is the max(N, D)-policy, which triggers transmission mode of radio server when the N and D policies are met simultaneously, and another is the min (N, D)-policy, which restarts transmission function of radio server when either of the N and D policies is first satisfied. Under two control policies, the steady-state queueing analysis of sensor node is mathematically carried out. The mean queueing measures of sensor node, such as the mean number of data packets, mean transmission time backlog, mean waiting time, mean busy period, mean busy cycle period, and so on, are derived. Two power consumption functions are constructed through the queueing measures obtained. Numerical experiments validate that two policies are feasible and efficient for power consumption control of sensor node. At a minimum power consumption, the superiority of the N-policy, D-policy, and two dyadic (N, D) policies is numerically compared. Some practical insights on the operation of two (N, D) polices in power consumption control of sensor node are obtained.