Wireless Sensor Networks is an area of research that has been evolving in many dimensions over the last few decades. A significant amount of work has already been done by the US Department of Defense in setting up radar imagery, target tracking, topology mapping and monitoring. Yet, in a commercial sense there is still a lot of work to be done and many new research avenues to be fully investigated. This site is designed to help fellow researchers with their ideas by facilitating the availability of information resources and simulation framework to use.

I am currently in the process of wrapping up my research on data routing and clustering in sensor networks. Along the way, I have had to spend a lot of time realizing some of the issues in a typical Wireless Sensor Network, that are not immediately obvious.

I have created this website to help fellow researchers in this area to avoid such problems. Also, I intend to include a simulation software (WSNSim) so others around the globe can also start enriching it with new content as well as getting a jump start on their research through the use of a framework that already emulates a typical wireless environment.

A bit about WSNSim:
The simulation tool uses a framework that attempts to emulate a true wireless environment capable of hosting multiple sensor nodes. The manner in which the nodes are deployed has been abstracted out, allowing users to use their own algorithms for deployment, rather than being restricted to what's available. This is also the case with the sensor nodes as well. Each node can be fitted with an engine that adheres to a common interface, but can be retrofitted to use any communication protocol to communicate between the nodes.

Being a wireless environment emulation tool, the framework also looks for packet collisions. So, unlike other simulation tools, WSNSim will not deliver packets to neighbouring nodes, if more than one node happens to transmit at the same time.

Each node is equipped with a clock and a processor - the clock frequency is relatively slow compared to a real node, but this is intentional to help understand node behaviour in a simulated environment. The nodes also carry a battery and for this I have used the Vrudhulla-Rakhmatov model - it is NOT a linear model, so a better representation of the battery life of a node can be achieved.

Also, there are several types of charts and node engines already included in the WSNSim, thus allowing users to quickly adopt the model.



Email: souren.sinha@wsnpedia.net