Optimization of Decision Thresholds in Two-Way Molecular Communication via Diffusion With Network Coding

Abstract

In this paper, we investigate a two-way molecular communication via diffusion (MCvD) system, which consists of two source nanomachines and one relay nanomachine in three-dimensional channel. The combinations of decode-and-forward (DF) relaying protocol and network coding (NC) scheme are implemented by using same number of molecules (SNM) and different number of molecules (DNM) which are released by source nanomachines. First, both the effects of actual and estimated inter-symbol interference (ISI) sequences are taken into account under SNM and DNM. Then the mathematical expression of the average probability of error for this two-way MCvD system are derived by using the Genie-aided detector and the detector which estimates the ISI sequence. Furthermore, we use particle swarm optimization (PSO) algorithm to solve the optimization problem whose objective is to minimize the average probability of error to search decision threshold at each nanomachine simultaneously. Finally, the numerical results show the DF and NC scheme outperforms the DF-No-NC scheme which is without NC. Moreover, PSO algorithm has good convergence behaviors and it is the most efficient to find optimal thresholds with fewest iterations compared with gradient descent algorithm and the iterative algorithm based on block coordinate descent algorithm. The obtained results are expected to provide guidance significance in designing two-way MCvD system with optimal decision thresholds and lower average probability of error.