Final Activity Report Summary - RANPLAN-HEC (Automatic 3G/4G radio access network planning and optimisation - A high end computing approach)
Cellular networks experienced very fast technology innovations and rapid expansion in the last three decades. Third and fourth generation mobile telecommunication (3G/4G) cellular networks will play a very important role in mobile broadband communications and ubiquitous computing.
Network planning is critical when building a new network or expanding an existing one. As radio access network (RAN) accounts for most of the infrastructure cost of a cellular network, careful design and accurate planning of RAN is vital to any operator. While in operation, RAN needs to undergo frequent optimisation, during which the problematic areas are identified and the base station (BS) parameters adjusted.
The planning and optimisation of RAN is very complicated. Traditional planning methods require experienced planners to manually select and configure network elements. As there are many network elements with many parameters to configure, it is very unlikely to find the optimal network configuration using such a method. Approaches and tools that can automate RAN planning and optimisation are urgently needed by the mobile industry.
Automatic RAN planning and optimisation is highly multidisciplinary, as it requires expertise in telecommunications, operational research (OR) and computing. Automatic 3G/4G RAN planning and optimisation requires large computing resources, both in terms of central processing unit (CPU) time and memory demand. In order to find a near-optimal solution for networks of realistic sizes in acceptable time frames, the recent developments in high end computing (HEC) should be used. In this project the way to automate 3G/4G radio network planning and optimisation was investigated. The main objectives of this project were to:
1. develop novel approaches to plan and optimise 3G/4G RAN
2. develop parallel approaches for RAN planning and optimisation, including parallel simulation approaches as well as parallel optimisation approaches
3. develop grid and web services for RAN planning and optimisation tasks
4. train highly skilled researchers in RAN planning and optimisation.
All the above objectives were achieved.
Network planning is critical when building a new network or expanding an existing one. As radio access network (RAN) accounts for most of the infrastructure cost of a cellular network, careful design and accurate planning of RAN is vital to any operator. While in operation, RAN needs to undergo frequent optimisation, during which the problematic areas are identified and the base station (BS) parameters adjusted.
The planning and optimisation of RAN is very complicated. Traditional planning methods require experienced planners to manually select and configure network elements. As there are many network elements with many parameters to configure, it is very unlikely to find the optimal network configuration using such a method. Approaches and tools that can automate RAN planning and optimisation are urgently needed by the mobile industry.
Automatic RAN planning and optimisation is highly multidisciplinary, as it requires expertise in telecommunications, operational research (OR) and computing. Automatic 3G/4G RAN planning and optimisation requires large computing resources, both in terms of central processing unit (CPU) time and memory demand. In order to find a near-optimal solution for networks of realistic sizes in acceptable time frames, the recent developments in high end computing (HEC) should be used. In this project the way to automate 3G/4G radio network planning and optimisation was investigated. The main objectives of this project were to:
1. develop novel approaches to plan and optimise 3G/4G RAN
2. develop parallel approaches for RAN planning and optimisation, including parallel simulation approaches as well as parallel optimisation approaches
3. develop grid and web services for RAN planning and optimisation tasks
4. train highly skilled researchers in RAN planning and optimisation.
All the above objectives were achieved.