Grid-based Approach for Working Node Selection in Wireless Sensor Networks Haining Chen, Hongyi Wu, and Nian-Feng Tzeng Center For Advanced Computer Studies University of Louisiana at Lafayette P.O. Box 44330 Lafayette, LA 70504 E-mail: {hxc5633,wu,tzeng}@cacs.louisiana.edu Abstract—In this paper, we propose a grid-based working node (WN) selection approach for wireless sensor networks. Due to coverage redundancy, it is highly desirable to identify a minimum subset of sensors in a wireless sensor network to serve as WNs, while the remaining sensors are deactivated to save power and reduce potential interference. The basic idea of our solution approach is to represent the coverage of the sensors by a number of sample points, i.e., the intersection points of the established grid. A simple approximation algorithm and a linear programming method are employed to select as few sensors as possible to cover all sample points. In order to reduce the computational time, clusters are formed and WN selection is performed within each cluster. The performance of the proposed WN selection schemes is quantified and the tradeoff among accuracy, communication overhead and computational time is evaluated via analyses and simulations. Keywords: Clustering, linear programming, node and network coverage, sensor networks, set selection, working nodes (WNs). I. INTRODUCTION With the advances in integrated circuits, micro-mechanism, and radio technologies, the low-power, low-cost wireless inte- grated sensors have become available for various monitoring, assessing, and control applications [1] [2]. Due to the limited computing power, sensing range, and transmission range of individual sensors, the sensor network is formed to detect the indicated phenomenon and to deliver the collected data to one (or several) processing/aggregating sensor(s) via pos- sible multiple hops. For some applications (such as forest environment monitoring), a large-scale sensor network can be deployed, with thousands or more sensors densely distributed in the interested area and working cooperatively for data collection and transmission. Additionally, to support effective monitoring, the sensors usually have location information through Global Positioning System [3] or local positioning systems [4] [5]. Various issues, such as sensor hardware design, location management, medium access control (MAC) and routing, have been discussed in the literature (see [6], [7], [8], [9]). This paper focuses on the topology management of sensor networks. Given the limited sensing range, each sensor has a coverage area, within which it can collect useful data. The union of the coverage areas of all sensors gives rise to the system coverage. For applications employing the sensor network in hostile environments, accurate sensor placement is usually impossible or non-cost-effective, and thus a large number of sensors are randomly distributed in the field to ensure the desired coverage and connectivity. The random sensor distribution may result in possible coverage redundancy (i.e., overlaps between the coverage areas of adjacent nodes), and it is highly desirable to identify a minimum subset of sensors, called working nodes (WNs), which together provide the same system coverage; any node outside the subset can be deactivated not only to save power for future use but also to reduce potential interference in wireless channels and to lower contention in the medium access, thereby prolonging the total life span and enhancing the performance of the sensor network. In this paper, we propose a grid-based WN selection ap- proach. In brief, a grid with proper density is established and the coverage area of a sensor is represented by a set of inter- section points of the grid. The WN selection problem can thus be converted into a set-covering problem (whose solution is NP-hard) and be solved by a simple approximation algorithm or a linear programming method. The grid density and its effects on accuracy are investigated and evaluated via analyses and simulations. In order to scale to large sensor networks, a distributed approach is developed, where the sensors are clustered by using the cluster formation algorithm [10] with the WN selection performed within each cluster. Extensive simulations are carried out to quantify the performance of the grid-based WN selection scheme, in terms of efficiency, communication overhead, and computational time. Our results show that the proposed approach can effectively select a small subset of sensors as WNs. The number of WNs selected by the linear programming approach is about 25% less than that obtained by the approximation algorithm. There is a tradeoff between the accuracy and the overhead. Specifically, with an increase in the cluster size, fewer WNs are selected (i.e., more accurate) at the expense of increased communication overhead and computing time. Our experiments show that the rational cluster size varies with the size of the network and the sensing range of the sensors. The running time of the proposed WN selection algorithm ranges from seconds to hundreds of seconds, depending on the cluster size and the grid density. The rest of this paper is organized as follows. Section II 0-7803-8533-0/04/$20.00 (c) 2004 IEEE IEEE Communications Society