Abstract
This paper proposes an artificial neural network (ANN) technique, using an adaptive batch training rule-based principle with weight and bias, to estimate and predict the adaptive threshold-based detection of the received molecular signals based on the current cumulative concentrations. The proposed technique considers a single transmitter (Tx Bio-NM) and a single receiver (Rx Bio-NM) biological nanomachines to exchange the molecular packets through diffusion-based molecular communication (DMC) channel. This scheme is suitable for binary On-OFF keying (BOOK signaling); where the 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. We evaluate the proposed technique in term of the bit error rate (BER), complexity cost and molecular throughput. The performance evaluation in various noisy channel sources shows a promising improvement in the un-coded BER and a slight number of molecules in transmission can be achieved extremely massive throughput of the DMC system compared to the other threshold detection schemes in the literature.