Convergence analysis of genetic algorithm applied for dynamic optimization of terminal to base station assignment in satellite Fed BFWA systems

Abstract

In point-multipoint (PmP) systems the signal to interference and noise ratio (SINR) highly depends on the assignment of terminal stations (TS) to base stations (BS). The broadband fixed wireless access (BFWA) systems operate at high carrier frequencies, i.e. microwave domain. In this frequency range wave propagation is highly influenced by precipitation, especially rain. Feeding of base stations can be solved with fiber optic or e.g. with satellite. In this work satellite feeding is assumed, which avoids land demolition caused by fiber deployment, but its drawback is the precipitation attenuation of Earth-space links. Applying site diversity can mitigate rain attenuation effects; however, known site diversity methods are only considering downlink channel quality and during site diversity terminals do not consider decisions of each other. By site diversity when downlink signal level decreases below a threshold, terminal station can be assigned to an other base station, though in point-multipoint systems this base station re-assignment can raise uplink interferences at other terminals. Present contribution provides a special site diversity which adopts genetic algorithm (GA) to simultaneously optimize downlink and uplink SINR values in BFWA. A convergence analysis method of the applied genetic algorithm is provided using Markov chain model.