Hierarchical and distributed control of wireless networks
Choosing how to optimise the behaviour of each node in a wireless network requires normally a global view of the whole network performance. But there is a way to allow each node to make some decisions itself and then to receive directions from controller placed higher in the hierarchy of the network (normally a gateway or a cluster head).
The delicate operation is to find the best strategy to distribute the control. For example each node can choose the queue length (how many incoming and outgoing packets are stored in internal memory) and transmission speed (how many packets per second are transmitted). The node doesn’t know though how many packets are received at the gateway and how long it took them to get there. Feedback form the gateway would enable the node to choose the right parameters.
The goal is to design, implement, and evaluate a distributed control problem for wireless networks.
The students can work on the queuing problem proposed above or can propose any control problem they’d like. They will work on networking simulators so experience with one of these (NS3, OMNET, The ONE) is advisable.
20% Theory, 60% Simulations, 20% Writing
Alessandro Chiumento (firstname.lastname@example.org)