ABSTRACT
Technological advances have led to the development of small, low-cost, and low-power devices that integrate microsensing with on-board processing and wireless communications. A node can typically have multiple sensors for collecting data, such as temperature, pressure, acoustic signals, video, etc. A large number of sensor nodes can be deployed or dropped in an area, and these nodes can self-organize to form a sensor network to perform useful computations on the sensed data in their environment. Sensor networks can be deployed in a variety of scenarios to collect information from the field, for example, target detection and tracking, military surveillance, building security, habitat monitoring, and for scientific investigations on other planets.2,3,12,21
Sensor networks can perform distributed computations in the network for a number of interesting applications, including signal and image processing to detect object signatures, tracking detected objects in the field, and performing statistical computations. However, sensor nodes have limited processing capability, memory resource, and communication bandwidth. Because they operate in the field using battery power, sensor nodes are severely constrained by energy resources. To operate a sensor network for a long period, energy resources must be carefully managed. There has been significant research progress in developing techniques for power-aware and energy-efficient computations and communications. These techniques
include the use of voltage scaling, turning nodes off when not in use, operating with different power levels, turning on and off selective parts, novel energy efficient protocols, etc. In addition, application-specific techniques can also be developed to reduce energy consumption in sensor networks.
