Robust Location Detection and Routing in Wireless Sensor Networks - 05/24/2010

Niloofar Fazlollahi

Abstract: With the advances in sensor production technology, a sensor with multiple functionalists (e.g. radio transmission, sensing, data processing, routing, etc.) has become a light and small unit which can easily be placed on room ceilings. Large number of sensors reaching thousands can form a sensor network which is useful in applications requiring unattended operations.

Location detection in indoor environments is an important application of sensor networks. A recent work by Saikat Ray and others uses robust identifying codes (ID codes) for location detection using sensor networks which is more robust compared to previous proximity based methods. Identifying codes where originally introduced by professor Karpovsky (from BU) for fault detection in mutli-processor networks. Later, Moshe Laifenfeld and others proposed use of robust identifying codes for joint location detection and routing in sensor networks. I will explain these in my presentation.

However, Moshe does not get involved in details like: how to distribute labels based on ID codes for routing? how to handle dynamic environment (maintenance)? what if an identifying code does not exist for the given graph?

My goal is to address some of these issues in more detail and possibly create more efficient schemes. For example, I have figured out that using dominating sets may be better than identifying codes for routing. So we are thinking of a mixed strategy. I also want to explore if the ID codes or dominating sets can be used to monitor network nodes (sensors) as well. Our other project on the table is to analyze performance of partial ID codes, i.e. codes that identify only a fraction of all nodes. There is some work on this by Moshe but he has not considered "robust" ID codes.

-- JiaxiJin - 13 Jun 2010