Contents:
The planning of the core network (fixed part) of a mobile radio network is a complex and difficult task. It can be split into several smaller sub-tasks, e.g. the analysis of traffic requirements, the topological outline, the provision of an optimal dimensioning of the links in terms of capacity. The planning process can be seen as an optimization problem: Find a network Topology which satisfies given constraints while minimizing the cost. Another important aspect Is the validation of the calculated results.
Nowadays the design process is often assisted by tools, which instrument approaches like stepwise design, heuristics or non-deterministic algorithms (e.g. genetic algorithms, simulated annealing) to achieve solutions to this task. In this laboratory planning, optimization and validation will performed with the help of the Composis tool.
Objectives:
On the one hand the objectives comprise the knowledge of being able to identify and classify typical network design problems. On the other hand participants should be able to choose, combine and to learn the mode of operation of selected solution techniques. The topological network design can be split into the following sub-problems each of which is present in almost any network planning task and each of which is also solved by one or more algorithms:
* election of the optimal topology from a set of possible ones
* Determination of an optimal routing
* Calculate the needed capacity of the links (dimensioning)
* Connection of nodes to an existing network (Expansion/Assignment problem)
In this laboratory methods of solving the above problems - while minimizing network cost and Fulfilling other given constraints - are presented. These constraints include: traffic requirements, Quality of service (e.g. blocking), potential locations of nodes and ducts and specific hardware cost.
Activities:
All sub-problems involved in the design of networks will be solved using the tool Composis as an example for a whole class of network design tools with a strong emphasis on algorithmic optimization. After you have learned the basic usage, you will design a GSM backbone network. You will have to solve the location problem, the assignment problem and other optimization problems, create valid routing tables and dimension all links appropriately. You will solve each task by using various settings and compare your results in order to find the best solution. Last but not least you will evaluate that network by means of stochastically event driven simulation.