Femtocell timeslot synchronization Keith Briggs 2013-12-19 (This project can interact with the other femtocell self-organization project which I am simultaneously proposing. The first three paragraphs of general background information are identical.) LTE is the technology currently being deployed for 4G wireless services. As well as providing the new generation of mobile phone services, its potential application to other areas is being investigated. A basic principle of LTE is the division of the two dimensions of frequency and time into so-called resource blocks (RB); a RB can be imagined as a rectangular unit capable of carrying a certain amount of data. In the time dimensions, RBs are 1 millisecond long. A possible application is to have a small LTE base-station ("femtocell") in every house, as a replacement for a wifi homehub. But there is a risk of interference, causing reductions in the data rate, and so various kinds of interference mitigation mechanisms are being considered. Roughly speaking, we want to achieve a self-organizating network (SON), in which each femtocell manages its own resource in response to its local environment, without any central planning or management. In mathematical terms, we have a complex system, with femtocells as nodes, and a graph with edges representing neighbour relationships. This project would use complex systems ideas, rather than specifics of the LTE technology. One difficulty is that the 1 millisecond RBs of one femtocell will not in general be aligned with its neighbours, since there are no known convenient ways of achieving the required accuracy in the internal electronic clocks. So this project would ask for an algorithm to be developed in which femtocells listen to their neighbours, and try to align their timeslots to those of their neighbours. No node has priority. Is there a simple algorithm which converges to global synchronization, despite the behaviour rule being purely local? Some possibly relevant existing work is in firefly swarm synchronization.