Reducing out-of-band radiations for cognitive automatic identification system based on multi-population genetic algorithm

As a maritime monitoring system aiming to minimize traffic accidents and improve traffic efficiency, cognitive automatic identification system (CAIS) employing some promising technologies, such as spectrum sensing and OFDM, has been developing by us in recent five years. However, OFDM suffers from high sidelobes, which results in adjacent channel interference. Reducing out-of-band radiations (OOBR) is thus critical to the CAIS, and other wireless communication systems with the high sidelobes of the subcarriers. To overcome the above issue, this paper investigates a new method utlizing the one among the database on the OOBR of the featured original signal and the optimal weighting factors which are optimized by multi-population genetic algorithm. These special data suppress the sidelobes of the transmit signal, but not affect signal in band given. With this approach, a significant reduction of the OOBR is achieved without waste spectrum but at the cost of an optimum searching procedure.