Alocação de canais em uma rede de rádios cognitivos gerenciada por base de dados preditiva adaptativa

Abstract

This work proposes a channel allocation methodology in a cognitive radio network (CRN) that uses a predictive and adaptive channel database. A spectrummodeling algorithm, that allows simulating channel occupation in cellular technology bands, generates the prediction data. The use of prediction enables the identification of opportunities in the bands of interest, which allows prioritizing the use of those channels with greater probability of being idle. Thus, this work extends the scope of IEEE 802.22, in that proposes the use of prediction data, as well as the use of cellular technology bands, which are commonly used in urban areas. The adaptive nature of the developed system updates the channel list dynamically over the allocation cycles. This feature enables to prioritize the channels with greater success in the allocation process, since the channel list is updated according to the results of secondary users (SUs) allocation. The proposed methodology aims to allocate the SU with the minimum of channel sensing operations required. The adaptive feature of the channel database gives a major contribution to reducing the channel sensing operations in the CRN. The decrease in channel sensing operations reduces the exchange of information between the functions that comprise the cognitive plane of a CRN. Consequently, a reduction in the network signaling traffic is expected, as well as a reduction of delays resulting from the search of available channels.