Locating Objects in a Sensor Grid Buddhadeb Sau 1 and Krishnendu Mukhopadhyaya 2 1 Department of Mathematics, Jadavpur University, Kolkata - 700032, India buddhadebsau@indiatimes.com 2 Advanced Computing and Microelectronics Unit, Indian Statistical Institute, 203, B. T. Road, Kolkata - 700108, India krishnendu@isical.ac.in Abstract. Finding the location of an object, other than the sensor in a sensor network is an important problem. There is no good technique available in the literature to find the location of objects. We propose a technique to find the location of objects in a sensor grid. The locations of the sensors are assumed to be known. A sensor can only sense the number of objects present in its neighborhood. This small information injects low traffic in the network. The computations are carried out completely in the base station. 1 Introduction Micro-sensor is a small sized and low powered electronic device with limited computational and communicational capability. A Sensor Network [1] is a net- work containing some ten to millions of such micro-sensors (or simply sensors). Location finding is an important issue [4, 5, 6, 7] in a sensor network. It usually involves transmission of huge information and a lot of complex computations. Heavy transmission load drains the energy and shortens the life of the network [4]. An efficient location finding technique is required that involves little infor- mation passing, like only the counts, angles or distances of objects. Counting of objects needs simpler mechanism and hardware than those for angles or distances etc. Several techniques are available for finding the location of a sensor with un- known position [2, 8, 10, 11]. But finding the location of objects has not received much attention. We propose a technique for finding locations of objects based only on the counts of objects sensed by different sensors. We formulate a system of linear equations for this purpose. The variables in the system are binary in nature. Standard techniques for solving a system of linear equations takes poly- nomial time whenever the system has a unique solution. Otherwise, the system gives an infinite number of solutions. But in our case, usually the system of equa- tions does not possess a unique solution. But the number of solutions should be finite, as the variables are binary in nature. One can determine the positions of A. Sen et al. (Eds.): IWDC 2004, LNCS 3326, pp. 526–531, 2004. c  Springer-Verlag Berlin Heidelberg 2004