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

Abstract

Molecular communication (MC) in microfluidic channels faces significant challenges in signal detection due to the stochastic nature of molecule propagation and dynamic, noisy environments. Conventional detection methods often struggle under varying channel conditions, leading to high bit error rates (BER) and reduced communication efficiency. This paper introduces ART-Rx, a novel Adaptive Real-Time Threshold Receiver for MC that addresses these challenges. Implemented within a conceptual system-on-chip (SoC), ART-Rx employs a Proportional-Integral-Derivative (PID) controller to dynamically adjust the detection threshold based on observed errors in real time. Comprehensive simulations using MATLAB and Smoldyn compare ART-Rx's performance against a statistically optimal detection threshold across various scenarios, including different levels of interference, concentration shift keying (CSK) levels, flow velocities, transmitter-receiver distances, diffusion coefficients, and binding rates. The results demonstrate that ART-Rx significantly outperforms conventional methods, maintaining consistently low BER and bit error probabilities (BEP) even in high-noise conditions and extreme channel environments. The system exhibits exceptional robustness to interference and shows the potential to enable higher data rates in CSK modulation. Furthermore, because ART-Rx is effectively adaptable to varying environmental conditions in microfluidic channels, it offers a computationally efficient and straightforward approach to enhance signal detection in nanoscale communication systems. This approach presents a promising control theory-based solution to improve the reliability of data transmission in practical MC systems, with potential applications in healthcare, brain-machine interfaces (BMI), and the Internet of Bio-Nano Things (IoBNT).

Figures (7)

• Figure 1: Tx and Rx nanomachines performing MC-based information transfer Kuscu2016OnBiosensors.
• Figure 2: (a) 3-Dimensional and (b) 2-Dimensional view of a rectangular microfluidic propagation channel. The transmitter and receiver locations, together with the dispersion of ligands as they propagate across the channel, are illustrated Kuscu2016ModelingReceiver.
• Figure 3: System-on-Chip (SoC) operations algorithm flow diagram with ART-Rx implemented. The diagram illustrates the sequential processes within the SoC, starting with system initialization and proceeding through molecular detection, signal conditioning, analog-to-digital conversion, and digital signal processing.
• Figure 4: The figure depicts the integrated components within the SoC boundary: the BioFET Sensor Module with its Biorecognition Unit and Transducer Kuscu2016ModelingReceiver; the Analog Front-End (AFE) comprising the Low Noise Amplifier (LNA) and Filters; and the Analog-to-Digital Converter (ADC). The Digital Signal Processing Unit (DSPU) is detailed with its Processor Core, Program Memory (ROM/Flash) storing the code, and Data Memory (RAM) storing parameters such as the setpoint $r(t)$, PID gains $K_P$, $K_I$, $K_D$, and previous values. Data transfer and control signals are managed via the Data Bus, Address Bus, and Control Bus, connecting the Processor Core with the Memory components. The DSPU computes the error signal $e(t)$ and processes it through the PID controller to generate the control signal $u(t)$, which is sent to the Threshold Control Interface (TCI). This interface adjusts the gate voltage $V_{\text{GS}}$ of the BioFET sensor, forming a feedback loop that enhances system responsiveness and accuracy. The Communication Interface enables data exchange between the SoC and external systems.
• Figure 5: BEP performance comparison between PID-based ART-Rx and the optimal method for varying (a) interferer molecule count, (b) $N_2$ CSK concentration levels, (c) average flow velocity, (d) transmitter-receiver distance, (e) intrinsic diffusion coefficient, (f) binding rate. In Figures (b)-(d), the concentration levels of CSK modulation were set at $N_1$ = 1000 for bit '1' and $N_2$ = 600 for bit '0' with the interferer molecule count numI set at 700. For Figure (a), the same concentration level settings were used but with varying interferer molecule count numI from 100 to 1600.