Twente Conference on IT: proposal madness
Well, the end of my bachelor is in sight. After six conferences are held, I've entered the track Ubiquitous Computing to write a paper and present it at the seventh conference on IT held as a final presentation of the bachelors Computer Science and Business & IT.
As usual these few years, it wasn't easy. Today I talked to the PhD student who will overview the research I will be doing in the field of Wireless Sensor Networks. She had some very useful comments on my draft and it is up to me to incorporate them into the draft and email my final proposal to the heads of the track. It was a bumpy road, being ill, stolen wallet and repairing everything (explaining yourself like I've had to do the last two years gets to you at some point). But now I'm there, it feels a bit unreal, like it's not ment to.
I'm actually quite pleased with the subject, but it will take some digging getting into it. What i'll be doing for the next nine weeks....? I'll try to make a simple introduction :)
Wireless Sensor Networks are quite dense networks of nodes. These nodes are very small and relatively cheap electronics with a limited battery, a radio and some calculational power. A this point, they are about the size of a matchbox. They are called sensor nodes because they have one or more sensors on them, which monitor their surroundings. For example: it has a heat sensor and registers that your living room has a nice and comfy temperature. Because of their relative low cost, you can deploy a lot of them, having multiple nodes per room. They have a radio, so they can talk to each other and relay information to each other or to an external computer. This way, you could monitor your room, keeping the temperature comfortably warm, with the network instructing your (de)central heating. But you could also equip them with small cameras and detect intrusions in your warehouses. Or put them on crates so the container in which they are placed can detect if something is wrong. There are a lot of applications and these networks run unattended, only interacting when needed so they don't bother you (well, that is the goal :P)
While such a network could be semi-static (the nodes are attached on surfaces like walls or stationary objects), you could put them on dynamic objects as well, making a network that is constantly moving, rearranging itself. For example: equip cars with a node and gridlocks are constantly creating new networks of cars trapped together. My research will be done in these dynamic networks, exploiting nodes which have movement sensors. Using nodes with an accelerometer and a compass I try to let the network discover which nodes are traveling as a group. This can be applied in logistics (are those goods moving together supposed to leave together?) or more fun things like dancing (are people moving together?). This is just a very simple explaination and I won't bother you with the complexities of the communication, mathematical calculations and other technical stuff ... maybe later ;-)

