Dear all,

We noticed the discussions in this mailing list regarding multihop performance. Although from a pure theoretical point of view it is normal that errors occur when transmitting large chunks of data across a lossy network, we agree that the current situation can be improved. We will follow up this discussion and help you with our experience on MAC protocols like 802.15.4 and routing protocols like AODV.

In the near future, we would also like to help fixing and avoiding bugs. As announced on the Symposium, we can start theses and research internships around telecom related topics. That's why we have decided to add to our list of possible subjects a thesis about implementing a virtual Solarium radio. The idea is to have a perfectly controllable environment in Solarium to test all kinds of protocols and situations, based on all existing layers above the radio. This way, testing will be possible without (or with minimal) changes to the existing code. Of course, parameters like the channel model and collision would have to be taken into account. Another thesis will be centered around multihop performance in sensor networks. A lot of work has already been done here, it would be very interesting in our opinion to see which results can and should be ported to the SPOTs. We consider improving diffusion of large files over a multihop network an interesting starting case.

Any comments on these topics are welcome, we are always interested to consider topic suggestions as well.

We will soon be posting a list of ongoing work to the forum, so you know what we are working on now.