Page 67 - 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

LRT
Conditional distribution of the received signal l(s) = f(y|b =1) > 1 0 P(b =0) j j In ML detection
f(y|b =1)
f(y|b =0)
H
j
j
<
P(b =1)
f(y|b =0) H
j
j
In MAP detection

P(b =0) P(b =1)
j
j
j
j
j
j0
Received signal, y j1 P(b =0)=P(b =1)
Fig. 17 – Conditional distribution of the received signal.

3.1 Mobile nano‑machines in pure diffusive
Blood Vessel
Active
Molecular Diffusion Binding channel
with Drift
3.1.1 Single transmitter and receiver‑based mo‑
bile MC systems

A single sample detector at a ixed sampling time in each
bit‑interval was proposed in [117] for diffusion‑based
Tx/Rx Movement MMC, where the OOK modulation scheme was considered
with Drift at the transmitter. In this work, dynamic CIR was mod‑
Advection Flow
9
eled by using the effective diffusion coef icient value in
the expression of static CIR. No noise and ISI were con‑
Fig. 18 – Mobile nano‑machines under drift and diffusion inside the
blood vessel. sidered while analyzing the system performance in terms
of error rate. The error probability of the order of 10 −4
in consecutive bit intervals. Further, ANN‑based detec‑ was obtained in the case of the low diffusion coef icient of
tors can also be very robust [106] under unknown chan‑ the transmitter (i.e., 10 −12 m /s). However, an increase in
2
nel conditions. These detectors can be trained on the the diffusion coef icient by 100 times led to degradation in
transmitted bit sequence and the corresponding received BER performance by approximately 100 times indicating
signal or features extracted from the received signal. Also, that the ixed sampling time does not work well for MMC.
unsupervised clustering [111] based on a fuzzy clustering
algorithm can be useful in unknown channel conditions. A non‑coherent detection scheme based on the local con‑
vexity of the received signal has been proposed in [135].
In this work, the received signal was iltered using a mov‑
ing average ilter. Three convexity metrics have been
3. TRANSMISSION AND DETECTION WITH found and added together to yield the inal decision met‑
MOBILE NANO‑MACHINES ric. This decision metric was compared with a threshold
to decide in favor of bit‑1 or bit‑0. Note that the convex‑
This section presents various transmission and detection ity metric was higher in the case of bit‑1. Also, the upper
schemes for mobile nodes. Performance and complexity and lower bounds on the detection threshold were eval‑
comparison of the detection schemes are also discussed. uated. Finally, the authors showed that the complexity of
An illustration of mobile nano‑machines inside the blood the convexity‑based detector was lower than the MAP and
vessel is shown in Fig. 18. MMSE detection schemes, however, the proposed detec‑
tion scheme achieved the identical BER values as the MAP
and MMSE detectors at a higher SNR regime.
9 Effective diffusion coef icient is calculated as the sum of diffusion coef‑
icients of signaling molecules and the receptors.

© International Telecommunication Union, 2021 55