GAWIND · Grid-enabled automatic wireless network design

Final Activity Report Summary - GAWIND (Grid-enabled Automatic Wireless Network Design)

Broadband Wireless Access Networks (BWANs) such as 3G/4G play an important role in ICT and in many other areas that use wireless communications. The rapid developments in BWANs require that many new BWANs to be deployed or expanded and that existing resources to be best utilised. Without proper planning and optimisation, a BWAN can neither be successfully deployed, nor be successfully expanded. Hence, network Planning and Optimisation (P&O) plays a vital role for the deployment and maintenance of the BWANs. The planning and optimisation of BWANs is very complicated. It is very unlikely one can find the optimal network configuration using the traditional manual methods. Hence, automatic network planning and optimisation approaches are highly desirable as they can typically improve the performance of a network by some 20-30%. Automatic planning and optimisation of BWANs is highly interdisciplinary, as it requires expertise in telecommunications, Operations Research (OR) and computing. Due to its complexity and interdisciplinary nature, previous attempts have had limited successes. New methods and approaches are urgently needed. Automatic planning and optimisation of BWANs requires large computing resources, both in terms of CPU and memory. In order to speed up BWAN planning and optimisation, one should take advantage of the recent developments in parallel and distributed computing. The goal of this ToK (Transfer of Knowledge) project was to help the host - Centre for Wireless Network Design CWIND, http://www.cwind.org develop competences that are needed to automate BWAN planning and optimisation. The objectives of the project are to help the host: 1. Develop competences in efficient Grid-enabled POAs (Parallel Optimisation Algorithms) such as Parallel Simulated Annealing (PSA) that can be used in BWAN planning and optimisation tasks. 2. Develop new competences in efficient parallel dynamic simulation that can improve the accuracy of the planning and optimisation and can be used to study RRM (Radio Resource Management) algorithms. 3. Develop competences in applying Grid to solve 3G/4G/WLAN planning and optimisation problems. 4. Acquire expertise on optimal design of drive test routes and on using system identification algorithms to automatically tune propagation models. 5. Establish multidisciplinary research within the host and to improve the host's participation in EU and international research programmes. At the end of the project, the host has obtained all the above mentioned competences and all the project objectives have been achieved. CWIND has been established as one of the best known research groups in wireless network planning and optimisation in the world. The project benefited 5 more experienced researchers, and 11 experienced researchers. These researchers involved in this project published over 50 papers in journals and conferences papers in the proposed topics during and shortly after the project.

Data: CORDIS, © European Union