Echo state network based symbol detection in chaotic baseband wireless communication

The Chaotic Baseband Wireless Communication System (CBWCS) is expected to eliminate the Inter-Symbol Interference (ISI) caused by multipath propagation using optimal decoding threshold that sum of ISI past decoded bits and future transmitting bits. However, current technique only capable removing partial effects ISI, because are available for sub-optimal calculation. unavailability information needed an obstacle further improve Bit Error Rate (BER) performance. In contrast previous method Echo State Network (ESN) predict one bit, proposed in this paper predicts directly ESN. ESN-based prediction simplifies symbol operation avoiding iterative output waveform points ESN accumulated error operation. With approach, calculation complexity reduced compared approach. achieves better BER performance method. reason superior result twofold. First, more symbols conveyed input obtain accurate rather than which was available. Second, here does not need estimate channel Least Squared (LS) method, avoids extra inaccurate estimation. Simulation results experiment based on a wireless open-access research platform under practical show effectiveness superiority