Resource Efficiency In Green Sensor Networks

A wireless sensor network (WSN) consists of a plenty of sensor nodes that arescattered in the Field of Interest (FoI) to monitor physical or environmental condi-tions. WSN is one kind of wireless ad-hoc networks. WSN achieves data sampling,processing and transmitting, which are corresponding to the sensing technology, com-puting technology and communication technology respectively. These technologiesare exactly the three fundamental information technologies today.WSN helps human to extend the ability of information capture. It connects thephysical information in real world and the digital information in cyber space. WSNis also the significant basis of next generation network: internet of things. WSN hasa wide application prospect including military defense, processing control, intelligentcity, health care, environment detection, emergent rescue, dangerous area surveillance,and remote control.Since the energy of a sensor node is limited, conventional works usually focuson the energy efficiency problem in order to prolong the life time of WSN. However,except the energy, other resources such as time, storage, and hardware are also limited.This paper extends the research of the green sensor networks from special energyefficiency to general resource efficiency. We propose novel research directions ongreen time resource, green storage resource, and green hardware resource (minimizingthe number of nodes).1. Green time resource. This paper studies a novel WSN application: mobilebarrier coverage (MBC) as an example for time efficiency. In the conventional barriercoverage, sensor nodes construct a barrier surrounding a static object, so that anyintruder would be detected by the sensor nodes. In the proposed MBC, the objects aredynamic. Thus, the mobile sensor nodes need to build a mobile barrier to surround theobjects adaptively. In the real world, several dynamic objects can benefit from MBC. For example, marching troop can detect any adversary intrusion without blind spotby MBC. However, the movement of objects are unpredictable, so the sensor nodeshave very limited time to react. Green time resource studies MBC by convex analysisand then proposes a fully distributed algorithm for sensor nodes to cooperatively move.The target is to minimize the traveling distance of sensor nodes, then the mobile barriercould be achieved as fast as possible.2. Green storage resource. A long-term and large scale WSN generates hugedata, which demand massive storage resource. A paper in SCIENCE investigates thatthe increment of data generated worldwide (dominated by sensor data) is growing31%faster than the increment of storage device produced. Consequently, storage efficiencyis worthy to study. Otherwise, data deluge will be a big problem for WSN. Our studyof green storage resource focuses on the environment surveillance application. Wepropose a dynamic data gathering (DDG) method based on compressive sensing the-ory and time series estimation. DDG adjusts the sensing rate and duty-cycle of allnodes according to the change of environment. The goal of DDG is to gather the leastdata, which can reconstruct the environment satisfying the accuracy requirement. Theevaluation performance shows that DDG is much better than existing low-duty-cycleor opportunistic data gathering method on storage and energy efficiency.3. Green hardware resource. Every sensor node has its manufacturing cost. In thelarge-scale WSN applications, it is cost saving to study how to use the least number ofsensors for covering the FoI. This is the goal of green hardware resource. Conventionalworks study the coverage problem in ideal2D plane or3D space. However, the FoIis usually a bounded complex surface in real applications. Green hardware researchfocuses on the surface coverage. Based on integral geometry theory, we study the fullcoverage on a surface with the least number of sensors.Combining the above three directions, we construct the basic research architec-ture on resource efficiency in green sensor networks.