ON THE DETECTION OF BINARY

CONCENTRATION-ENCODED UNICAST

MOLECULAR COMMUNICATION IN NANONETWORKS

Mohammad Upal Mahfuz, Dimitrios Makrakis and Hussein Mouftah

School of Information Technology and Engineering, University of Ottawa, K1N6N5 Ontario, Canada

Keywords: Molecular communication, Concentration encoding, Sampling-based detection, Energy-based detection,

Nanonetworks.

Abstract: Molecular communication is a new communication technique where transmitter and receiver communicate

by transmitting molecules and correspondingly modulating their specific characteristics. Molecular

communication is being considered as a new physical layer (PHY) option for a vast number of

communicating nanomachines that form “nanonetworks.” Thus it has become a promising option for a large

number of new applications, offering several benefits over conventional electromagnetic communications

based on radio waves or optics at nanoscale dimension. Concentration-encoding is a simple and good

technique to encode information with molecules. Incorrect detection of concentration-encoded signals

makes molecular communication a real challenge. This paper has addressed sampling-based and energy-

based detection approaches in detail for binary concentration-encoded molecular communication signals

based on diffusion in fluidic media.

1 INTRODUCTION

Molecular communication is a new physical layer

option that is being considered for communication

and networking among a huge number of natural and

man-made nanomachines (Akyildiz et al. 2008,

Mahfuz et al. 2010a, Mahfuz et al. 2010b). A

nanomachine is a tiny machine capable of

performing simple tasks e.g. sensing and actuation.

As shown in Fig.1 a transmitting nanomachine (TN)

transmits molecules in a fluidic medium, the

molecules propagate in the medium following the

diffusion process, and then finally the molecules are

received by the receiving nanomachine (RN).

Detection of concentration-encoded molecular

signals is very crucial in the sense that incorrect

detection of concentration-encoded signals would

result in wrong decoded bit, thus producing bit

errors, and finally the effect / reaction performed by

the RN would be erroneous. In this paper we have

explained two detection approaches, named

“sampling-based detection (SD)” and “energy-based

detection (ED),” for binary concentration-encoded

molecular communication in order to detect the

information bits transmitted by TN and that are

available at the location of RN in the form of

received throughput of the molecular propagation

channel. We have also shown threshold

characteristics for “known-reference” and “blind-

reference” cases for bit detection. In the former case,

RN knows the average transmission rate used by TN

(in molecules/second) beforehand, and thus RN can

compute the threshold concentration as a function of

known transmission rate for a given distance

between TN and RN. On the other hand, in blind-

reference case RN does not know the transmission

rate of molecules beforehand, and so RN computes

the threshold concentration as a function of available

throughput at its own location. Although threshold

computation is an important aspect for molecular

communication, it has not been addressed in depth

so far in open literature. The paper is organized as

follows: section-2 briefly discusses the throughput

of the channel, followed by section-3 discussing the

fundamental concepts of sampling-based (SD) and

energy-based (ED) detection methods for the

amplitude modulated concentration-encoded signal.

Finally, section-4 concludes the paper.

446

Upal Mahfuz M., Makrakis D. and Mouftah H..

ON THE DETECTION OF BINARY CONCENTRATION-ENCODED UNICAST MOLECULAR COMMUNICATION IN NANONETWORKS.

DOI: 10.5220/0003158504460449

In Proceedings of the International Conference on Bio-inspired Systems and Signal Processing (BIOSIGNALS-2011), pages 446-449

ISBN: 978-989-8425-35-5

 2011 SCITEPRESS (Science and Technology Publications, Lda.)

Figure 1: A generic molecular communication channel

between a transmitting nanomachine (TN) and a receiving

nanomachine (RN).

2 PROPAGATION OF

MOLECULES

We have assumed binary amplitude modulation for

the concentration-encoded signal. We assume that

TN is a nanomachine or a biological entity that can

emit only one kind of molecule. We consider

hypothesis H

as the case when bit ‘1’ is to be

transmitted, and hypothesis H

as the case when bit

‘0’ is to be transmitted. Correspondingly, TN

transmits

average

Q molecules per second on an average

for the entire bit duration of

seconds for H

, while

TN does not transmit any molecules at all for

. As

shown in Fig.2, transmission stimulation protocol

can be expressed as

;

()

average

⎧

⎨

⎩

(1)

Assuming a point source type TN the concentration

of molecules

(,)Urtin molecules per unit volume

at a three-dimensional space

ˆˆ

rixjykz=⋅+⋅ +⋅

and

at time

t changes with time and space as the well-

known Roberts equation (Bossert and

Wilson 1963)

as shown below

()

{}

()

222

()

,,,

4( )

xyz

Uxyzt e d

πτ

−++

−

=⋅

−

∫

(2)

where

is the distance vector between TN and RN,

is a dummy variable of integration, and D is the

diffusion constant of information molecules in the

medium in cm

/second unit. Considering spherical

symmetry Eq.(2) can be written as a function of

r (in

cm)

as (Berg 1993).

()

{}

()

Urt e d

πτ

−

=⋅

−

∫

(3)

where

when a Cartesian coordinate

system is assumed. In a simple manner

U(r,t) is also

known as the throughput of the molecular

propagation channel.

Figure 2: Binary concentration-encoded molecular

signalling, Q

representing the total number of molecules

transmitted by TN during T

3 DETECTION APPROACHES

3.1 Sampling-based Detection (SD)

As shown in Fig.3 with SD approach the propagation

channel is excited with a random bit sequence of

bits transmitted by the TN, and correspondingly the

throughput

U(r,t) is sampled at any suitable time

instant during the bit duration

. However, with the

assumption that TN is in time-synchronization with

RN we assume that the sampling instants are at

exactly the middle of any bit duration and thus can be

expressed as

Samples

35 (21)

, , ,......., ,...

22 2 2

(2 1)

...,

bbb b

TTT n T

−

(4)

t (seconds)

Q(t)

(0,0)

average

(b) Binary modulation of transmission rate

Bit ‘1’

average

molecules/second)

Bit ‘0’

(No transmission of molecules)

(a) Input stimulation protocol

Information

molecules

Molecular propagation channel

Information

Source

Modulator

Reaction

or Effect

Demodulator Channel

U( r

,t)

r=0,

t=0

ON THE DETECTION OF BINARY CONCENTRATION-ENCODED UNICAST MOLECULAR COMMUNICATION

IN NANONETWORKS

447

where N is the total number of bits in the bit

sequence,

is the duration of each bit, and

1, 2,3,...,nN=

is the index of bits. As a result, the

detection variable or test statistic (Kay 1998)

for

detecting the

bit is denoted as

(2 1)

ZUr

−

⎡⎤

⎢⎥

⎣⎦

(5)

For a given data rate f, bit duration is fixed at T

=1/f.

However, since throughput U(r,t) received at the

location of RN largely depends on the distance r

(Mahfuz et al. 2010b),

is a function of both

distance r and diffusion constant D. For all practical

reasons assuming D to remain constant over the

entire observation period T, we find that

becomes significantly affected by r. For known-

reference SD approach, RN has prior knowledge of

the transmission rate of input (i.e. Q(t)=Q

average

), and

thus the most common threshold can be given as

Known

average

K =

(6)

Figure 3: Sampling based (SD) detection for transmitted

bits {101100}. Sampling instants are shown in red arrows,

denoting the level gap between maximum and

minimum points of U(r,t) when a bit toggles.

Since

is significantly affected by r, known-

reference SD approach should be used in short-range

molecular communication (Mahfuz et al. 2010b).

Threshold for blind reference case is given below as

the average throughput

U(r,t) over the interval from 0

to 2

seconds as shown in transmission protocol in

Fig.2,

{}

Blind

over time duration 0 to 2

Average ( , )

KUrt=

(7)

However,

Blind

K is highly influenced by r as shown in

Fig.4 below.

Figure 4: Distance dependence of throughput, with

average

=10,000 molecules/sec., T

=10 seconds, random bit

sequence {10100} in air medium.

3.2 Energy-based Detection (ED)

As shown in Fig.5 with ED approach available

throughput U(r,t) in molecules per unit volume is

integrated over any entire bit duration T

seconds and

the resulting accumulated amount of molecules is

termed as the detection variable

for the

corresponding bit and expressed as

() ( )

(1)

nUrtdt

−

∫

(8)

where

1, 2,3,...,nN

is the index of the bit and N is

the total number of bits in the random bit sequence.

Referring to Fig.2, in known-reference ED approach

RN detects a bit when the accumulated molecules

during T

is greater than or equal to the threshold

Known

K that is most commonly the half of the total

transmitted energy during 0 to 2T

seconds as shown

below

Known

()

average b

Qtdt

K ==

∫

(9)

The threshold given by Eq.(9) is independent of r,

however, as shown in Fig.4 the throughput U(r,t) is

highly influenced by the variation of distance r. As a

result, this distance-dependence of throughput makes

the threshold selection inappropriate for known-

reference ED approach. On the other hand, in blind-

reference ED approach RN does not know the

transmission rate of TN. As a result, RN can compute

the threshold as the average integral of throughput

for the duration from 0 to 2T

as shown in Fig.2.

t (sec.)

U(r,t)

Q(t)=Q

average

Samplin

Bit

‘1’

Bit

‘0’

Bit

‘1’

Bit

‘1’

Bit

‘0’

Bit

‘0’

… nT

Threshol

BIOSIGNALS 2011 - International Conference on Bio-inspired Systems and Signal Processing

448

Thus the threshold for blind-reference case can be

expressed as

Blind

(,)

Urtdt

K =

∫

(10)

as shown in the transmission protocol in Fig.2. As

shown in Fig.4 since throughput U(r,t) varies with

the varying distance, threshold in blind-reference

case should vary as per the variation of the distance

between TN and RN.

Figure 5: Concept of energy-based detection with a

random bit sequence {101100}.

4 CONCLUSIONS

In this paper detection methods for binary

concentration-encoded molecular communication

channel has been addressed. Sampling-based and

energy-based detection methods have been proposed

as possible detection schemes for concentration-

encoded signals. Three important factors that affect

the performance of detection methods are noise

immunity, distance-dependence of throughput, and

timing synchronization of TN and RN. The SD

approach detects the bit based on only one sample

value of the throughput taken at the sampling

instant. Thus SD approach is applicable for ideal

environment and for short ranges, and so it is not

recommended for most real cases where the

communication is impaired with noise and/or for

medium-to-long range communications. Since

throughput varies with varying distance a fixed

threshold for known-reference case tends to be a

strict selection, whereas for blind-reference case RN

has to compute the threshold from the throughput

only, making the threshold highly dependent on the

distance between TN and RN. Finally, timing

synchronization can be achieved by correctly

characterizing the propagation delay between TN

and RN. While a synchronizing clock for molecular

communication can be difficult, asynchronous

signalling (Moore et al. 2007) can also be used for

the purpose of detection of molecular signals.

ACKNOWLEDGEMENTS

The first author would like to thank the Natural

Sciences and Engineering Research Council of

Canada (NSERC) for the financial support in the

form of doctoral scholarship to carry out this

research work.

REFERENCES

Akyildiz, I. F., F. Brunetti, and C. Blazquez (2008)

“NanoNetworking: A New Communication

Paradigm,” Computer Networks, Vol. 52, pp. 2260-

2279, Elsevier, August.

Berg, H. C. (1993) Random Walks in Biology, Princeton

University Press.

Bossert, W. H., and E. O. Wilson (1963) “Analysis of

Olfactory Communication among Animals,” Journal

of Theoretical Biology, pp.443-469.

Kay, S. M. (1998) Fundamentals of Statistical Signal

Processing, Vol. 2: Detection Theory, Prentice Hall.

Mahfuz, M. U., D. Makrakis, and H. Mouftah (2010a)

“Characterization of Molecular Communication

Channel for Nanoscale Networks,” in Proc. 3

International Joint Conference on Biomedical

Engineering Systems and Technologies (BIOSTEC

2010), Valencia, Spain, 20-23 January.

Mahfuz, M. U., D. Makrakis, and H. Mouftah (2010b)

“Spatiotemporal Distribution and Modulation

Schemes For Concentration-Encoded Medium-to-

Long Range Molecular Communication,” in Proc. 25

Biennial Symposium on Communications, Kingston,

ON, Canada, 10-12 May.

Moore, M., A. Enomoto, T. Suda, T. Nakano, and Y.

Okaie (2007) “Molecular Communication: New

Paradigm for Communication Among Nano-Scale

Biological Machines,” The Handbook of Computer

Networks, John Wiley & Sons Inc., November.

t (sec.)

Q(t)=Q

average

Bit

‘1’

Bit

‘0’

Bit

‘1’

Bit

‘1’

Bit

‘0’

Bit

‘0’

… nT

Desired energy for the first Bit ‘1’

Interference energy from the first Bit ‘1’ to

the first Bit ‘0’

U(r,t)

ON THE DETECTION OF BINARY CONCENTRATION-ENCODED UNICAST MOLECULAR COMMUNICATION

IN NANONETWORKS

449