Optimal detection for diffusion-based communications in the presence of ISI

Communications based on diffusion refers to the transfer of information using molecules as message carriers whose propagation is governed by the laws of Brownian motion. Molecular communication is considered to be one of the most promising approaches for the end-to-end communication between nanoscale devices. In this paper, both an optimal and a suboptimal receiver detection scheme are proposed for a diffusionbased binary digital communication system in the presence of ISI. The transmission of binary information is accomplished by using On-Off Keying (OOK) with only one molecule. The proposed system can serve as the theoretical basis for end-to-end communication in molecular nanonetworks where molecules of different types are used by different nanoscale devices. The effect of channel memory resulting from the residual molecule diffusion from previous transmissions is treated analytically in the formulation of the detection schemes. Numerical results show that the proposed detection schemes can maximize the mutual information over a practical range of the parameter of signaling interval without a priori information. A channel capacity of 1 bit per channel utilization during a signaling interval can be ultimately achieved by extending the duration of the signaling interval, even with infinite channel memory.