Some Research Challenges

Communication-Theoretical Modeling

Developing accurate models for the transmitter, channel, and receiver is a crucial step for the design of man-made molecular communication systems.  However, the modeling of these components in MC systems is vastly different from the modeling of traditional communication systems as the size of the nodes of MC systems is on the order of tens of nanometer to tens of micrometer. At nano- and micro-scale, materials show different physical, chemical, electrical, and magnetic behaviour than at macro-scale which has to be carefully accounted for in the modeling, fabrication, and development of new devices, materials, and systems.

Receiver Modeling

A receiver that is designed to respond to the arrival of molecules requires the implementation of a corresponding biochemical process. For example, the surface of a receiver may be covered with receptors that are able to bind the information molecules. Sufficient binding could trigger a chemical reaction that corresponds to the detection of the information molecules at the receiver. Therefore, the design of a reception mechanism has to be implementable via a corresponding reaction mechanism.

Receiver covered with receptors that bind with the information molecules and trigger a biochemical process


The receiver must be able to convert the molecules that it observes into useful information, which may be represented as a sequence of bits

Channel Modeling

A diffusive molecular communication channel is somewhat analogous to a conventional wireless communication channel, but there are significant differences. The properties of the physical layer are distinct from those of electromagnetic-based communication and lead to a series of open problems. They include:

  • Random Movement: The diffusion of any individual molecule can be modeled as a random walk. While the average motion of the molecules is deterministic, how many molecules are at the receiver at a given time is not known a priori.

Diffusion of any individual molecule is modeled as a random walk caused by thermal vibrations and collisions

  • Time of propagation: Diffusion is not an instantaneous process, and the average displacement from the source is proportional to the square root of the time that a molecule takes to diffuse. Thus, it is very difficult for a transmitter to rely solely on diffusion to communicate efficiently with a receiver that is far away.

Information molecules can take a long time to diffuse from the transmitter to the receiver, and may not arrive at the reciever at all

  • Inter-symbol interference: As a side effect of the first two properties, molecules that are emitted can remain in the neighborhood of the receiver for a long time. However, the receiver cannot differentiate between the arrivals of an identical type of molecule emitted at different times. This causes inter-symbol interference.
  • Molecular noise: Molecules in biological systems fulfill multiple roles. The selection of a molecule to act as an information carrier based on biocompatibility will likely mean that noise molecules from unintended sources will arrive at the receiver.
  • Fluid dynamics: Aqueous environments that are candidates for molecular communications will likely be undergoing flow and other dynamics. These phenomena drastically affect the motion of freely-diffusing molecules.

The environment for molecular communications undergoes fluid dynamics

Molecular Communication System Design and Optimization

After developing theoretical models for the transmitter, channel, and receiver, the next step is to design and optimize the building blocks of the communication system such as modulation, channel estimation, and signal detection.

Modulation Design

The goal of the modulation subsystem is to represent the information bits by  physical properties of the signalling molecules released by the transmitter, e.g., the number, type, and time of the release of the molecules. For instance, a simple and efficient modulation scheme is the so-called ON-OFF keying, where to convey a binary „one“, the transmitter releases N of the signalling molecules whereas to convey a binary „zero“, the transmitter remains silent and does not release any molecules.

Channel Estimation

Knowledge of channel state information (CSI) is needed for the implementation of most detection and equalization schemes at the receiver. In MCs, CSI mainly refers to the expected number of molecules observed at the receiver over time after the instantaneous release of molecules by the transmitter which is also known as the channel impulse response (CIR) of the channel. For instance, assuming ON-OFF keying modulation, knowledge of the CIR is required at the receiver to determine the detection threshold.

Detection Design

The final step is to recover the information bits at the receiver via the detection and equalization subsystems. In particular, the design of any detection and equalization scheme depends on the adopted type of modulation as well as the properties of the MC channel. For instance, for ON-OFF keying modulation and a non-dispersive MC channel, the optimal detector is a threshold detector, i.e., if the number of molecules observed in a given symbol interval exceeds a threshold, the detector detects a binary „one“, otherwise, a binary „zero“ is detected. However, in general, the MC channel is dispersive due to the diffusive propagation of the molecules which induces inter-symbol interference. To cope with the  inter-symbol interference, the receiver has to further equalize the received signal before detection.


Implementation of MC systems at microscale is necessary to improve our knowledge and broaden our understanding of nano-networks. However, diffusion-based MC has not been experimentally verified at micro-scale yet.