The increasing demand for mobile communications leads mobile service providers to look for ways to improve the QoS and to support increasing numbers of users in their systems. Because the amount of frequency spectrum available for mobile communications is very limited, efficient use of the frequency resource is needed. Currently, cellular system design is challenged by the need for a better QoS and the need for serving an increased number of subscribers. Network planning is becoming a key issue in the current scenario, with exceedingly high growth rates in many countries which force operators to reconfigure their networks virtually on a monthly basis. Therefore, the search for intelligent techniques, which may considerably alleviates planning efforts (and associated costs), becomes extremely important for operators in a competitive market.
Cellular network planning is a very complex task, as many aspects must be taken into account, including the topography, morphology, traffic distribution, existing infrastructure, and so on. Things become more complicated because a handful of constraints are involved, such as the system capacity, service quality, frequency bandwidth, and coordination requirements. Nowadays, it is the network planner’s task to manually place base stations (BSs) and to specify their parameters based on personal experience and intuition. These manual processes have to go through a number of iterations before achieving satisfactory performance and do not necessarily guarantee an optimum solution. It could work well when the demand for mobile services was low. However, the explosive growth in the service demand has led to a need for an increase in cell density. This in turn has resulted in greater network complexity, making it extremely difficult to design a high-quality network manually [1–3].
Furthermore, OFDM (orthogonal frequency division multiplexing) technology is emerging as an attractive solution for fast wireless access. It has been adopted for many future wireless networks, e.g., FLASH (fast low-latency access with seamless handoff) OFDM and WiMAX. The UMTS (Universal Mobile Telecommunication System) evolution will go into the direction of OFDM, e.g., LTE (Long Term Evolution). Similar to other technologies, the deployment of OFDM networks poses the problem to select antenna locations and configurations with respect to contradictory goals: low costs versus high performance. A key to successful planning is the fast and accurate assessment of network performance in terms of the coverage, capacity, and QoS [4]. This also makes the conventional design methods insufficient for planning mobile networks in the future. Thus, more advanced and intelligent network planning tools are required. A promising planning tool should be able to aid the human planner by automating the design processes.
