Abstract: Wireless Sensor Network has most widely used applications ranging from health care to military. Wireless Sensor Networks (WSNs) have many tempting features but due to the lack of any defense mechanism, the security becomes an issue in such networks. To operate WSNs in a secure way, any kind of intrusions should be detected before malicious attackers can harm the network. The intrusion detection system identifies the legitimate attackers in the network area. In this paper we have optimized the intrusion detection capability in a wireless sensor network using Gaussian distribution and uniform distribution method. Furthermore, the performance of Gaussian-distributed WSNs is compared with uniformly distributed WSNs. The effect of various network and system parameters such as sensing range of the sensors, number of sensor deployed, intruderís starting distance and maximum allowable intrusion distance on the intrusion detection probability of the network has been observed. Hence detection probability can be analytically formulated in a random WSN and an appropriate deployment strategy can thus be selected to determine critical network parameters.

Keywords: Gaussian and uniform distribution, intrusion detection, WSNs.