Target Surveillance Using Mobile Wireless Sensor Networks: Models, Theories, and Applications

Sunday, May 10, 2009 - 2:00 pm to 5:30 pm. Salmiah Room.

Presented by:

Prof. Noureddine Boudriga, Dr. Mohamed Hamdi
Communication Networks and Security Research Lab.
SUPCOM, University of Carthage, Tunis
URL: http://www.cnas.org.tn

Abstract:

Wireless Sensor Networks (WSNs) are among the technologies that will probably shape the first decades of the twenty first century. These networks are mainly cost-effective, easy to deploy, and multi-purpose. In fact, WSNs have been used in various contexts including mobile target detection, healthcare, water resource monitoring, and virtual reality. However, they are also characterized by severe memory, CPU, and (most importantly) energy limitations that harden their deployment in environments where voluminous data should be processed and high-speed networks are to be used to transmit these data. For instance, existing WSNs devised for military surveillance often provide coarse data about the hostile target(s) moving in the battlefield. This constraint, mainly fixed by the sensor node cost, considerably affects the efficiency of the WSN.

The proposed tutorial consists of four parts. First, we cover the main challenges associated with the development of a WSN. In fact, the fact that processing, memory, and energy resources are scarce in a network composed of tiny motes and should be necessarily taken into consideration. Moreover, data loss, conflicts, and communication latency render the communication unreliable. The second part addresses several issues related to the deployment of a sensor network.  An emphasis is made on activity scheduling mechanisms which are implemented to extend the lifetime of network. Sensor nodes can switch between different states (active, inactive, idle) in order to adapt its energy consumption to the WSN performance. An interesting point around this aspect is the impact of activity scheduling on the efficiency of the WSN.  We also consider the coverage problem which is of great importance in target tracking applications. The objective is to assess the area of the regions where the mobile target cannot be detected by the WSN. The third part is dedicated to the study of secure tracking algorithms. A set of filtering-based techniques will be discussed. Some of these filters will serve to predict future positions of the target given a history of position measurements while other filters, referred to as smoothing filters, are rather used to interpolate the target position. Furthermore, threshold signature and group key exchange algorithms are presented as alternatives to enforce security policies in a sensor network. Finally, two major applications will illustrate the practical concepts addressed in the tutorial. The first is military tracking where the target is a vehicle. Acoustic and image-based techniques are the most techniques in this context. The second is pollution tracking where the surveillance should take place over a wide area since the propagation of the ‘target’ clearly differs from the former case.