Pianificazione Ottima di Reti di Stazioni Radiobase mediante Algoritmi Genetici a Codifica Binaria

The sudden growth of mobile communications, and more specifically the forthcoming new generations of mobile systems (such as UMTS) require the development of sophisticated tools for automatic and optimum planning and network design. Such systems must include accurate radiopropagation models, as well as efficient global optimizers. Due to the huge complexity of the task, the combined use of more than one radiopropagation model is needed, as well as their integrated use inside an optimization framework. An environment is proposed, based on a Geographical Information System (GIS) framework, inside which a wide variety of radiopropagation models can dynamically interact with a novel binary-coded genetic optimizer using multiple cost functions, so that the positioning, power sizing and tilting of radiobase antennas is optimized. Results on real cases, such as real GSM/DCS subnetworks, demonstrate the affordability and reliability of the system. The use of multiple cost functions, as well as the use of rigorous radiowave models, render the environment appropriate to achieve both radioprotection purposes (controlling the exposure of citizens to electromagnetic fields) and radiocommunication goals (for instance improving network coverage and quality-of-service).