ART-Rx: A Proportional-Integral-Derivative (PID) Controlled Adaptive Real-Time Threshold Receiver for Molecular Communication

Abstract

Signal detection in diffusion-based molecular communication (MC) is challenged by stochastic propagation, inter-symbol interference (ISI), and rapidly varying microfluidic channels. This paper presents ART-Rx, an adaptive real-time threshold receiver that embeds a proportional–integral–derivative (PID) controller in a conceptual system-on-chip with the detection threshold updated once per symbol interval. Extensive Smoldyn and MATLAB simulations sweep the interferer molecule count, concentration-shift keying (CSK) levels, flow velocity, transmitter–receiver (Tx–Rx) distance, diffusion coefficient, and receptor binding rate. Averaged over the interferer molecule sweep, ART-Rx achieves a mean bit-error ratio (BER) of $1.8\times 10^{-2}$ . Across −4 dB ≤ SNR ≤ 19 dB the BER remains below $6.0\times 10^{-2}$ , and never exceeds $7.4\times 10^{-2}$ for Tx–Rx distances up to $1\times 10^{-2}\,\mathrm {m}$ . The closed-loop strategy outperforms a statistical fixed-threshold detector and achieves a $2.6\times $ lower BER than a prior non-machine learning (ML) baseline while retaining $\mathcal {O}(1)$ operations per symbol. Gain scheduling, coupled with Ziegler—Nichols (Z–N) tuned PID gains and an integral windup clamp, preserves stability across strongly non-linear parameter regimes. These results position ART-Rx as a practical Rx front-end for small, resource-constrained Internet of Bio-Nano Things (IoBNT) nodes and implantable biosensors.