Page 19 - ITU Journal Future and evolving technologies Volume 2 (2021), Issue 3 – Internet of Bio-Nano Things for health applications

ITU Journal on Future and Evolving Technologies, Volume 2 (2021), Issue 3

based nanobiosensors have also a biorecognition layer, rics. Because of this, there has been tremendous inte-
which replaces the gate electrode in conventional FETs, rest in modeling the MC channels to ind the ultimate
and consists of receptor molecules selectively binding tar‑ performance limits in terms of information theoretical
get molecules via af inity‑based ligand‑receptor interac‑ capacity and data rate. In majority of studies, MC
tions. Depending on the transducer channel material, the channel is usually assumed to be unbounded where
biorecongition layer can host a wide range of receptor information‑carrying molecules propagate through free
molecules, ranging from proteins to DNAs. diffusion with the underlying phenomenon of
Among other options, graphene FET (GFET) biosensors Brownian motion [60, 61]. In a few studies,
provide unique advantages for the practical design of MC diffusion is accompanied by a low which directs the
receiver. The main advantage of graphene is its high sen‑ propagating molecules to a distant receiver [62, 63, 64],
sitivity to the charged analytes, e.g., proteins and DNAs, whereas some studies also consider the existence of
due to its extremely high carrier mobility and one‑atom reactive molecules within the channel which can
thickness, exposing all its atoms to the environment. The chemically degrade the information carrier molecules
advent of new types of receptors, e.g., aptamers, has and reduce the intersymbol interference. A few studies
broadened the target range of nanoscale FET biosensors consider bounded MC channels, for example mi‑
from ions to proteins, peptides, and even whole cells. cro luidic channels where molecules propagate through
Aptamers are short functional oligonucleotides (typically convection‑diffusion. In majority of these studies, it is
20‑60 nucleotides). Their base sequences for speci ic tar‑ assumed that the molecules are transmitted from a hy‑
gets are ied from an oligonucleotide library with pothetical point source, which is capable of releasing a
an in vitro process called systematic evolution of ligands known number of molecules to the channel in the form
by exponential enrichment (SELEX). Their application of an impulse signal at a given time instant. On the other
in biosensors has gained momentum due to their wide hand, the receiver is typically assumed to be a transpa-
target range, chemical stability, and ease of production. rent instrument, which is capable of counting every
Combined with the exceptional properties of graphene single molecule in a hypothetically de ined space [63],
and aptamers, the ability of nanoscale FET biosensors or an ideal absorbing instrument capable of counting
to provide selective, label‑free and continuous detection each molecule that is absorbed [65]. Common to these
makes GFET aptamer‑based biosensors, i.e., GFET ap‑ studies is the ignorance of the impact of the physical
tasensors, very promising candidates for the design of MC architectures of the transmitter and receiver on the
receiver. communication channel. As such, researchers have been
Biological MC receiver designs are based on the enhance‑ able to adopt the EM‑inspired simpli ications in
ment of biosensing and biochemical signal processing modeling, such as linear and time‑invariant (LTI)
functionalities of livings cells with synthetic biology tools channel characteristics with additive white Gaussian
for the receiver operation. This approach consists in the noise, neglecting the effects of interactions and
design of new synthetic receptors that can provide more correlations resulting from transmitter and receiver
sensitivity and selectivity in physiological environments, architectures and channel geometry. This leads to a
for example, through kinetic proof reading mechanisms serious discrepancy between theory and practice, as
[57], and the implementation of new chemical reaction revealed by the initial MC experiments performed with
networks within the cell that can realize the required ‘macroscale’ sensors and dispensers utilized as MC
computations for decoding the received MC signals. Syn‑ transmitter and receiver, respectively, showing that the
thetic biology is already mature enough to allow perfor- nonlinearity and time‑variance caused by the operation of
ming complex digital computations, e.g., with networks transmitter and receiver invalidate the models built upon
of genetic NAND and NOR gates, as well as analog these assumptions [66, 67].
computations, such as logarithmically linear addition, On the other hand, some research groups have studied
ratiometric and power‑law computations, in synthetic MC receivers that rely on ligand‑receptor binding reac‑
cells [58]. Synthetic gene networks integrating tions, the common molecular sensing method in natural
computation and memory is also proven feasible [59]. MC [68, 69]. Deterministic models, assuming free diffu‑
More importantly, the technology enables sion and point transmitter, have been developed for a vir‑
implementing BNTs capable of ob‑ serving individual tual MC receiver with ligand receptors. Although the con‑
receptors, as naturally done by living cells. Hence, it sideration of ligand receptors has advanced the accuracy
stands as a suitable domain for practi‑ cally of the models one step further, the employed assump‑
implementing more information‑ef icient MC detec‑ tors tions about the transmitter and channel strictly limit the
based on the binding state history of individual re‑ applicability of these models. Additionally, stochastic
ceptors. receiver models are developed for FET nanobiosensor‑
based MC receivers [69]. In [62], a model for MC with 2D
b) MC Channel Modeling: To design effective and ef i‑ biosensor‑based receivers in micro luidic channels is
cient MC systems addressing the needs of the envisioned provided. However, these initial models also rely on
IoBNT applications, it is important to have a theoretical unrealistic assumptions, e.g., equilibrium conditions in
framework which can be used to optimize the physical ligand‑receptor binding reaction, and ignore the
components of the system with ICT performance met‑ implications of the receiver geometry.

© International Telecommunication Union, 2021 7