Molecular communication (MC) reimagines information theory for biological environments. Instead of modulating electromagnetic waves, an MC transmitter releases physical particles—such as ions, proteins, or synthetic microparticles—into a fluid medium, where they diffuse to a receiver.
Early research in this field borrowed heavily from traditional wireless communication abstractions. However, treating physical matter in a fluid channel exactly like photons in a vacuum creates a profound theory-practice gap. Traditional mathematical models often assume noise-free diffusion fields, infinitely small point sources, and perfectly absorbing receivers. As noted by Arash Einolghozati et al. in Molecular Communication Between Two Populations of Bacteria, isolating biological noise to the nodes while assuming an idealized diffusion channel overestimates the reliability of real-world signal propagation.
To bridge this gap, researchers are shifting from abstract mathematical conveniences to physically grounded models that account for geometry, chemical diversity, and transceiver hardware constraints.
The Geometry Problem: Moving Beyond Point Sources
In textbook physics, modeling a transmitter as an infinitely small "point source" is highly convenient. It preserves circular symmetry and yields elegant, closed-form differential equations for diffusion. However, real-world micro-scale transmitters possess physical volume that fundamentally alters how particles disperse.
H. Birkan Yilmaz et al. address this discrepancy in A Machine Learning Approach to Model the Received Signal in Molecular Communications, demonstrating that ignoring the physical volume of a spherical transmitter significantly distorts predicted signal characteristics. A real transmitter acts as a physical boundary in the fluid, blocking or altering the local diffusion paths of the very molecules it releases.
Idealized Model (Point Source) Physically Grounded Model
.
* .
* * .
*
* .
* * /---------\ .
. (Point) .
* | Spherical | *
* .
* . |Transmitter| .
.
* .
* \---------/
* | *
(Molecules cannot diffuse
through the transmitter)
A similar abstraction problem occurs at the receiver. Early models assumed "passive" receivers that merely count molecules passing through a space, or "perfectly absorbing" receivers that instantly capture every molecule that touches them. In reality, a receiver's surface relies on chemical receptors with specific binding kinetics.
To resolve this, O. Tansel Baydas et al. introduced a more realistic framework in Received Signal and Channel Parameter Estimation in Molecular Communications. By modeling receivers as partially absorbing spheres, the authors demonstrated that a significant portion of molecules bounce off the receiver before eventually binding or diffusing away. Accounting for this physical reality drastically improves the accuracy of channel parameter estimation, particularly at short communication distances where idealized models fail entirely.
Chemical Diversity vs. Electromagnetic Uniformity
In conventional wireless systems, radio waves are uniform; a photon is a photon. In molecular communication, the information carriers themselves are highly diverse.
Hanlin Xiao et al. argue in What Really is "Molecule" in Molecular Communications? The Quest for Physics of Particle-Based Information Carriers that treating molecules as abstract entities ignores critical physical properties like size distribution (e.g., in superparamagnetic iron oxide nanoparticles) or structural topology (e.g., in DNA strands). These physical traits directly dictate the molecule's diffusivity and introduce unique noise profiles into the channel.
Furthermore, physical molecules do not exist indefinitely in a fluid environment; they interact with it. As explained by Malcolm Egan et al. in Stochastic reaction and diffusion systems in molecular communications: Recent results and open problems, real-world channels are governed by stochastic reaction-diffusion systems where molecules degrade, encounter external fluid forces, or undergo chemical transformations.
Because molecules linger in the environment and slowly degrade rather than vanishing at the end of a time slot, the channel retains a "memory" of past transmissions. This creates severe inter-symbol interference (ISI). Hamidreza Arjmandi et al. developed frameworks to calculate capacity bounds that explicitly account for these memory effects in Capacity of Diffusion-Based Molecular Communication Networks Over LTI-Poisson Channels, proving that molecule degradation dynamics must be integrated directly into the communication architecture to prevent signal saturation.
The Transceiver Hardware Bottleneck
Even if the diffusion channel is modeled perfectly, a communication system cannot function without a physical interface to translate messages between domains. For applications like the Internet of Bio-NanoThings, chemical signals inside the body must be converted into electrical signals for digital networks (Sidra Zafar et al. in A Systematic Review of Bio-Cyber Interface Technologies and Security Issues for Internet of Bio-Nano Things).
A primary candidate for this bio-cyber gateway is the Silicon Nanowire Field-Effect Transistor (SiNW BioFET), which uses a grid of receptors to capture molecules and alter an electrical current accordingly (Murat Kuscu et al. in Modeling and Analysis of SiNW BioFET as molecular antenna for Bio-cyber interfaces towards the Internet of Bio-NanoThings).
[Chemical Domain] ======> [Electrical Domain]
Target Molecules ---> Binds to BioFET ---> Alters Nanowire Conductance ---> Digital Signal
In a comprehensive survey, Murat Kuscu et al. pointed out that traditional information theory often isolates the communication channel from the physical realities of transmitter and receiver operations (Transmitter and Receiver Architectures for Molecular Communications: A Survey on Physical Design With Modulation, Coding, and Detection Techniques). When the physical constraints of a BioFET receiver are modeled—including its finite surface area, receptor saturation, and internal electronic noise—the theoretical capacity of the system drops significantly.
By analyzing these hardware constraints in Modeling and Analysis of SiNW FET-Based Molecular Communication Receiver and On the capacity of diffusion-based molecular communications with SiNW FET-based receiver, the authors demonstrated that receiver processing capacity, rather than channel diffusion, often acts as the true bottleneck in practical molecular communication networks.
Bypassing Mathematical Intractability with Machine Learning
As models incorporate more physical realism—such as fluid boundaries, molecule degradation, and receptor binding kinetics—the underlying differential equations quickly become mathematically intractable. Solving Navier-Stokes and reaction-diffusion equations simultaneously in real-time is impossible for resource-constrained nanodevices.
To bypass this complexity, researchers are turning to data-driven modeling and physical testbeds. In A Novel Experimental Platform for In-Vessel Multi-Chemical Molecular Communications, Nariman Farsad constructed a physical macroscale testbed that used acids and bases to transmit data through fluid-filled tubes. Instead of trying to derive an impossibly complex analytical model for the nonlinearities and memory effects of the tube environment, the author trained a Recurrent Neural Network (RNN) directly on the empirical data. The RNN successfully learned the channel dynamics, enabling a tenfold increase in data rate over traditional modeling approaches.
This shift toward empirical validation and hardware-aware modeling represents the critical next step for the field. As emphasized by Azim Eskandarian et al. in Scanning the Issue, translating mature mathematical communication theory into functional medical and technological tools requires moving past abstract simulations and designing around the messy, stochastic realities of physical matter.
Key Foundational Work
Capacity of Diffusion-Based Molecular Communication Networks Over LTI-Poisson Channels (2015) — contributed 2 concept elements to this direction.
Modeling and Analysis of SiNW BioFET as molecular antenna for Bio-cyber interfaces towards the Internet of Bio-NanoThings (2015) — contributed 2 concept elements to this direction.
Molecular Communication Between Two Populations of Bacteria (2012) — contributed 1 concept elements to this direction.