Initial Distance Estimation and Signal Detection for Diffusive Mobile Molecular Communication

Mobile molecular communication (MC) attracts much attention in recent years where mobile nanomachines exchange information using molecules. In this paper, we consider a diffusion-based mobile MC system consisting of a pair of diffusive nanomachines. Due to the Brownian motion of nanomachines, the communication distance between the transmitter and the receiver is dynamic. Thus, the channel impulse response (CIR) is a stochastic process. The stochastic CIR brings the difficulty in the detection process. Contrast to the common static MC characterized by the deterministic CIR, in a mobile MC system, the receiver needs to estimate the dynamic distance for CIR reconstruction and detection threshold setting at each bit interval, which achieve high computational complexity. To tackle these difficulties, a new detection technique for mobile MC is proposed in this paper. It is unnecessary to estimate the dynamic communication distance at each bit interval. Instead, the receiver estimates the initial distance between it and the transmitter. The estimated initial distance can be used to reconstruct the statistical characteristics of CIR, setting detection threshold for all the bit intervals in advance. To achieve this goal, a novel two-step scheme based on maximum likelihood (ML) is proposed to estimate the initial distance. In the first step, the transmitter releases some molecules as a pilot signal before the information bits transmission. Then the receiver estimates releasing distance by observations of received signal. In the second step, the estimated value obtained in the previous step is used to estimate the initial distance by ML estimation. The performances of proposed two-step scheme and the detection technique are evaluated via particle-based simulation of the Brownian motion.