Analysis of Error-Detection Schemes in Diffusion-Based Molecular Communication

Abstract

Despite recent advances in molecular communication among bio agents, the design of reliable communication schemes remains an open problem. One of the requirements is to develop suitable coding schemes, which meet the molecular communication specific constraints in terms of reliability and complexity, and take into account the communication channel properties. In this paper, we consider diffusion-based molecular communication in which the information is encoded into the concentration (e.g., on/off keying). Such a communication system is modeled as operating over a completely asymmetric channel where one of the bits can be transmitted without any error while the other can undergo a random error by the channel. Because of the limitations of bio-agents, we focus on error-detection schemes, which require far less complexity at the receiver relative to error-correction codes. To obtain an optimal detection scheme, we model the detection problem via an erasure channel and propose algorithms to obtain the optimal codewords efficiently for two different optimality measures. Then, we consider an error-free subfamily of such codes, namely constant weight codes, and propose an implementation specific to the molecular communication. We analyze the rate of the constant-weight coding scheme, compare it to the theoretical limits and specify the optimal weights and lengths of such codes.