Sequential training-based receiver for On-OFF-keying diffusion-based molecular communications

In this paper, we purpose and investigate an integrated artificial neural networks (ANNs) to estimate and predict the adaptive threshold detection scheme in diffusion-based molecular communication (DMC). The proposed scheme forwards an estimate of the received bits based on the current cumulative concentrations, which is derived using sequential training principle with weight and bias. The ANN model is implemented on MATLAB software. This method is suitable for binary On-OFF-Keying (Book signaling), where the receiver biological nanomachine (Rx Bio-NM) adapts the 1/0-bit detection threshold based on all previous received bits to alleviate the inter-symbol interference (ISI) problem and reception noise. Numerical and simulation results show the improvement in molecular throughput and the optimal number of molecules in transmission. Furthermore, the performance evaluation in various noisy channel sources shows a promising improvement in the un-coded bit error rate (BER) compared with other threshold detection schemes in the literature.