Minimizing Power Consumption in Sensor Networks with Quality of Service Requirement Navid Ehsan Dept. of Electrical and Computer Engineering University of California, San Diego, CA 92093 nehsan@ucsd.edu Mingyan Liu Dept. of Electrical Engineering and Computer Science University of Michigan, Ann Arbor, MI 48109 mingyan@eecs.umich.edu Abstract Minimizing energy and power consumption in wireless sensor networks is critical in extending the lifetime of the network. In this paper we consider efficient power/rate allocation strategies with certain quality of service con- straints. The transmission rate achieved and the power used are related by a power-rate function for a given channel state. Specifically, two problems are formulated, one seeking to minimize the average power consumption sub- ject to an average rate constraint, while the other seeking to maximize the average data rate subject to an average power constraint. We consider two scenarios of power management of a wireless node/device. In the first case the node always stays on thereby consuming power in circuitry as well as in transmission/reception. In the second case the node turns its transceivers off from time to time, and by doing so while avoiding circuitry power during sleep periods, consumes power in switching from the off state to the on state. For both problems stationary and deterministic policies are considered. We first study the optimum power/rate allocation when the nodes are always active and provide algorithms that produce optimal solutions to these two problems. We then consider a two-state channel model and study optimal joint power allocation and sleep scheduling for both problems. 1 Introduction Energy and power efficiency is an essential factor in the design and operation of wireless sensor networks, since many sensor devices are battery powered and thus energy/power constrained. On the other hand, it is also desirable to be able to satisfy certain performance guarantee when transmitting data over a wireless sensor network, e.g., delay, throughput, loss requirements. How to meet such requirements while limited by energy/power is the central focus of this paper. Specifically, we consider the problems of optimally assigning power (rate) to sensor nodes so as to minimize (maximize) the average power (rate) subject to an