Optimization techniques in computer networks.

The rapid growing demand of internet services makes a continuous expansion and enhancement of internet infarstructure and organisatiodn a necessity. The challenge is to ensure that for a certain demand in data traffic the exisiting lines and machines be used effectively. On the other hand, a clever design of the network ad distributed storage of data can decrease the amount of data traffic needed to meet a certain demand of services. There is no doubt that mathematical optimisation techniques are a neecessary tool in solving this kind of problems. The aim of the project is therefore to explore the appropriate mathematical techniques. Important areas of application include: Optimisation of data traffic, given a certain demand, by enhancing routing algoritms: staring from the design of a mathematical model for the network, the routing problem can be solved with mathematical programming and yields an optimal distribution of the data flows in the model. the challenge is then to find out to which extent the existing routing algorithms lead to an optimal distribution, and how these can be refined so as to approach that optimal solution. Distributed data storage: the location of data servers within a network is a key factor in determining the amount of traffic needed for a certain demand of services? The data traffic can be greatly reduced and bottlenecks can be avoided by replication of frequently requested data in different locations. One of the major issues here is to maintain a consistent set of data when update is needed. This problem is closely to the domain of location analysis.