| Abstract: | In this article we present the Self-Adaptive Universal Receiver (SAUR), a novel wireless reconfigurable receiver architecture. This scheme is based on blind recognition of the system in use, operating on a new radio interface comprising two functional phases. The first phase performs a wideband analysis (WBA) on the received signal to determine its standard. The second phase corresponds to demodulation. Here we only focus on the WBA phase, which consists of an iterative process to find the bandwidth compatible with the associated signal processing techniques. The blind standard recognition performed in the last iteration step of this process uses radial basis function neural networks. This allows a strong analogy between our approach and conventional pattern recognition problems. The efficiency of this type of blind recognition is illustrated with the results of extensive simulations performed in our laboratory using true data of received signals. |
