Internal Soliton

8/11/2019 Internal Soliton

1/26

Journal Code: Article ID Dispatch: 23.07.14 CE: Maria Nicasana Buenavista

D A 2 8 4 3 No. of Pages: 16 ME:

INTERNATIONAL JOURNAL OF COMMUNICATION SYSTEMSInt. J. Commun. Syst. (2014)

Published online in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002./dac.2843

01

02

03

RESEARCH ARTICLE0405

Performance analysis of distributed underwater wireless acoustic0607

sensor networks systems in the presence of internal solitons08

09

10 Amit Kumar Mandal,

, Sudip Misra, ,

, Mihir K. Dash and Tamoghna Ojha 11

12 School of Information Technology, Indian Institute of Technology, Kharagpur -721302, India

132Center for Oceans, Rivers, Atmospheric and Land Sciences, Indian Institute of Technology, Kharagpur -721302, India

14

15

16

17SUMMA

RY

18

In this paper, we have analyzed the performance of distributed Underwater Wireless Acoustic Sensor Net-

Q

119 works (UWASNs) in the presence of internal solitons in the ocean. Internal waves commonly occur in a

20 layered oceanic environment having differential medium density. So, in a layered shallow oceanic region,21 the inclusion of the effect of internal solitons on the performance of the network is important. Based on var-

22 ious observations, it is proved that nonlinear internal waves, that is, solitons are one of the major scatterersof underwater sound. If sensor nodes are deployed in such type of environment, internode communication is

23

affected because of the interaction of wireless acoustic signal with these solitons, as a result of which network24 performance is greatly affected. We have evaluated the performance of UWASNs in the 3-D deployment25 scenario of nodes, in which source nodes are deployed in the ocean floor. In this paper, four performance26 metrics, namely, Signal-to-interference-plus-noise-ratio (SINR), bit error rate (BER), Delay (DELAY), and

27 energy consumption are introduced to assess the performance of UWASNs. Simulation studies show thatin the presence of internal solitons, SINR decreases by approximately 10%, BER increases by 17%,delay

increases by 0.24%, and energy consumption per node increases by 53.05%, approximately. Copyright

2014 John Wiley & Sons, Ltd.

Received 25 April 2014; Revised 7 June 2014; Accepted 4 July 2014

KEY WORDS: distributed underwater sensor networks; acoustic communication; internal soliton; acoustics

signal; scattering

1. INTRODUCTION

38 This paper presents the performance analysis of distributed UWASN systems in the presence of39

internal solitons, which commonly occur in shallow coastal oceanic environments having differen-40

tial medium density. UWASNs can be used to collect oceanographic data, monitor pollution, prevent41

disaster, and for tactical surveillance applications [14].42

Unlike the communication channel in terrestrial wireless sensor networks, which is mostly static43

in nature, underwater communication channel is much more dynamic [5]. This dynamism arises44

because of the nonuniform distribution of different physical parameters. There are various exist-45

ing literature (e.g., [612]) on UWASNs. However, they have ignored the effect of solitons on the46

network performance.47

A UWASN is formed of sensor nodes deployed in some region of the ocean, and they commu-48

nicate with one another through wireless mode using acoustic signals [13, 14]. The characteristics

49 of the deployment region varies both spatially and temporally. Some regions are prone to more nat-


2/26


3/26

2 A. K. MANDALET AL.

01 urally occurring more dynamic oceanic phenomena than others. In those regions, communication

02 performance is affected because of the internode link quality degradation caused by the interaction

03 of acoustic signal with the dynamic oceanic physical phenomenon. One such phenomenon having

04 strong influence on acoustic signal propagation is internal soliton, which arises in a stratified oceanic

05 environment, typically in shallow water region of depth limited by 200 m. A strong internal wave,

06 which is nonlinear in nature, is known as an internal soliton, and it advances in packet form through

07

the ocean column. Solitons are well known in acoustics as sound scatterers [15]. They affect the08 acoustic signal of frequency, f, in the range 1kH 6f 650 kH , which is the optimum frequency

9 range of internode communication in UWASNs.10 In the soliton induced region, the wireless acoustic link quality is perturbed by the existence of11 soliton wave packets. The creation of solitons depends on both intrinsic dispersion and nonlinearity12

of the medium. In shallow water regions, internal solitons are often strongly nonlinear in nature13

[15]. They mainly affect the upper or shallow coastal regions of an ocean.14

15

2. RELATED WORK AND CONTRIBUTION16

17

18 There exist works (e.g., [1618]) on the study of performance of terrestrial wireless sensor network

19 under different scenarios. Unlike terrestrial wireless sensor network, UWASNs are subjected to

20 increased challenges because of adverse conditions of underwater channel. Again, compared to

21 deep water, shallow water is prone to increased dynamism. Therefore, it is important to assess the

22 performance of UWASNs in shallow underwater regions.

23 There are works (e.g., [19, 20]) specifically on shallow underwater acoustic networks, and gen-24 erally, on UWASNs, such as [21, 22]. Particularly, from the perspective of UWASNs, performance25 evaluation in the physical layer was undertaken on various aspects. Most of these pieces of work are26 focused on the physical characteristics of acoustic signal. There is lack of work on the performance27 evaluation of UWASNs in the presence of waves existing inside the water body of the ocean.28 Ismail et al. [23] assessed the performance of UWASNs on the basis of signal attenuation through

29 oceanic under water columns only. In [24], Zorzi et al. have taken linear topologies of sensor nodes30 and considered noise, propagation delay, and their impacts on the transmission power and band-31 width. They have not considered any realistic dynamic phenomena existing in the channel. Stefanov32 and Stojanovic [25] analyzed interference-induced performance analysis of wireless acoustic net-33 works. They modeled frequency dependent path loss of acoustic signals connecting nodes using34 the wireless mode of communication. The research work does not consider any realistic physically

35 occurring oceanic phenomenon in the channel. In [26], Babu et al. have considered the frequency

36 dependent fading and time variation characteristics of underwater channel. However, they have not

37 considered any underwater dynamic phenomenon like internal solitons. Xu et al. [27] evaluated the

38 performance of UWASNs by considering different metrics, such as packet delivery ratio, network

39 throughput, energy consumption, and end-to-end delay. However, performance evaluation was exe-

40 cuted in the presence of node mobility and other common aspects relevant to underwater channel.

41 They have not considered any dynamic underwater phenomena occurring in the ocean. In addition

42 to considering the commonly reported problems in underwater such as delay, path loss, and Doppler

43 spread [28], Ancy et al. have also shown the technique of data transmission in the presence of

44 shadow zone. However, they did not consider any dynamical phenomenon in the channel through45 which data transmission takes place. Llor et al. [29] analyzed the transmission loss of signals for46 UWASNs by considering the environmental factors, such as surface waves. However, they have not

47 considered any phenomenon governing the volume of water. Xie et al. [30] statistically modeled48 the path loss between two sensor nodes at a particular frequency and time. In predicting the path49 loss of an acoustic signal in underwater, they have only considered the surface wave activity on the

50 movement of sensor nodes.51 From the review of existing literature, it can be inferred that the effect of internal solitons on52 the communication performance has not been studied so far. In our work, we have considered the

53 existence of internal solitons in shallow oceanic coastal region and have studied their effect on the

54 performance of UWASNs.

Copyright 2014 John Wiley & Sons, Ltd. Int. J. Commun. Syst. (2014)

DOI: 10.1002/dac


4/26

301

02

03 Internal solitons

are ubiquitous in coastal oceanic water having stratified layered column. In this04 section, we briefly

introduce their analytical characteristics and dynamics.05

06 3.1. Characteristics of internal soliton

07

08

091011

1213

141516

17

18

1920

21

22 _ The solitons in shallow water regions are usually observed in packet form. They are also known

23 as solitary waves. In particular, they are observed during summer when they are trapped in24 strong seasonal thermocline [33].

25 _ The solitons in shallow water are described by the Korteweg-de Vries equation [15], which is

26 explained in this Section briefly. These solitons exist in the rank order.

27 _ The wave packets are highly correlated. The maximum number of packets occurs during

the28 spring tide, and the minimum number of packets occurs during the neap tide [33]. After for-29 mation, they propagate shoreward, and their variation is also observed because of the change

30 in bottom topography.

31 _ Their surface signature is mainly observed in summer season because of the shift of

thermo-32 cline toward surface.33

34 3.2. Governing equations35

For an incompressible stratified fluid in a gravity field, the hydrodynamical equations are given36

as [34]:37

38

d U139 E

C _r

p f k U g

40 dt D _

_O _ E

__E (1)

41

42

@_

:U

_

D

0

(2)

Internal solitons propagate along the pycnocline of a ocean. However, their generation can be achieved in

different ways such as direct displacement of pycnocline, and conversion of complex tidal energy into the

pycnocline motion. The interaction of internal tide with the irregular bottom topography leads to the

formation of internal solitons. As their names imply, they propagate through the interior of the ocean.

Solitary waves are a class of non-sinusoidal, nonlinear, mostly isolated waves of complex shape occurring

frequently in nature [31]. Internal solitons consist of several oscillations confined in a particular region or

space. They always remain in packet form and can carry considerable shear velocity leading to turbulence

and mixing. This mixing often introduces bottom nutrients of ocean to the upper part in it. They play a

significant role in the dynamics of shal-low oceanic coastal zone as well as deep water zone [32]. They are

mostly found in the shelf region of an ocean and have significant impact on the acoustical properties in

shallow coastal area.

In our work, we consider UWASNs deployed in the shallow coastal zone of an ocean. Therefore, it is

important to understand the characteristics of internal solitons affecting the communication performance ofthe whole network. The following are the characteristics of internal solitons:

.

UNDERWATER WIRELESS ACOUSTIC SENSOR NETWORKS


5/26

_

rEE _44

45 E:U 0 (3)E

D46 r

47 In Equation (1), .u x; u ; u/ is the velocity vector of a fluid parcel,p is the fluid pressure,gE

D48

k E49 is the gravitational acceleration, is the unit vector acting vertically upward along the increasing +z

direction, and f D 2_si n_is the Coriolis parameter, where_and_are, respectively, the Earths50 angular velocity of rotation and geographical latitude of a particular place on the Earth.51

Because of the presence of internal waves, density perturbation occurs. Using the Boussineq

52 approximation, we can write the density field in the presence of internal waves as

53

54_D _0./ C _per.x; y; ; t /; _per __0 (4)


DOI: 10.1002/dac


6/26


01 In Equation (4), _0 is the equilibrium or unperturbed depth dependent density, and _per is the

02

perturbed density. As per Boussineqs approximation, vertical variation of _0./is significant for

03 only the buoyancy term, which is proportional tod_0 [15]. Let the shallow water region be bounded

04 by two surfaces, z = 0 and z = -D, where D is the depth in meters. Zero vertical displacement is05

applied at these two surfaces. Using Boussineq approximation, the hydrodynamical equations in the

06

presence of internal waves can be written as [35]:

07

@u

08 E :U 0 (5)

09Eh

C @ Dr10

11 @U 1 @U12 Eh Eh

rE Uh !D 0@t C _0r

E

p

0C

_

0

f k U

Cu

@C

:U

(6)13

_

_ h Eh

14 _ _ _

15@_

0 d_0

@_0

_D 0

16 :U

C

u d

C_

u@C_

rE_

_017 @t Eh (7)

18

19 @p0

@u :U u_C _0

@u20 @t

_

_ @ r @t (8)Eh

21 _ _22 In Equations (5)(8), D.ux; uy is the velocity vector existing in the x-y plane, and the23

Eh

operator is a two dimensional one, which is as

24 r25

E @ @ (9)26 i jO

@x CO

27

r D @y

28

29

3.3. Theoretical model for shallow water internal solitons

30

Tidal interaction with the ocean bottom topography appears to be the dominant mechanism for31

the generation of coherent internal waves near the continents [15]. It is well known that acous-32

tic transmission loss depends strongly on frequency. Again, the results of experiments reported in 33

the literature [36][33] show that wireless acoustic communication is affected because of acous-34

tic signal scattering over various frequency ranges by solitons arising due to vertical (upward and 35

downward) displacement of the pycnocline layer under the combined effect of buoyancy force and36

gravitational force. Therefore, while sensor network is deployed in such internal solitons induced 37

region, internode wireless acoustic communication is affected. 38 Uu39 We expand the vertical velocity, , and the horizontal velocity, h, in terms of eigenmodes. The

40 eigenmode representation of these two terms are as follows:

41142

u D

.U/n.u/n.x; y; t /

(10)

43

44

n

X45

146 U D _ d.U/n.u / .x; y; t / (11)47 dEh n h n48 nD1

X49

In Equations (10) and (11), nis the mode number, .U/n denotes the orthogonal eigen functions,50

and _n is a constant. The vertical displacement, _, of an isopycnal surface is represented as5152

153 _ r;E

t D _n.x; y; t /.U/n (12)54nD1


7/26

XCopyright 2014 John Wiley & Sons, Ltd. Int. J. Commun. Syst. (2014)

DOI: 10.1002/dac


8/26

UNDERWATER WIRELESS ACOUSTIC SENSOR NETWORKS 5

01

02

03

04

05

06

0708

09

10

11

12

13

Figure 1. Schematic view of generation and propagation of internal solitons in shallow coastal ocean.

14

15 Under the nondispersive and linear approximation, the function .U/nsatisfies the following second16 order differential equation as

17d

2U N218

(13)

d

2

c2

19

20 In Equation (13), the quantity Nis known as the BruntVisl frequency, which is in the order21

of 10 cycles/hr. It is expressed in terms of rad/s. When N2 > 0, the water column is stably stratified.22 It is expressed as2324

g d_025

N D __0 d (14)2627

With this frequency, N, a stably stratified column of water oscillates under the combined influence28

of gravity and buoyancy forces. Figure 1 shows a two-layered internal wave model in shallow water, 29

which shows the generation and advancement of internal solitons in the ocean. From the figure, we 30

see that the upper layer of density _1and depth d1is separated by the lower layer of density _2 >

_

1

31

and depth d2. Typically, in the mid-latitude continental shelf water [15], L varies in the range 132

5 km, the amplitude or the maximum elevation of pycnocline, _0 ranges from 0 to 30 m. From the33

horizontal view, we see that the characteristic soliton packet width, w , has value 100 m, the packet34

wavelength ranges from 50 to 500 m, the packet crest height, H , changes in the range 030 km,35

and two successive inter-packet distance varies in the range 1525 km. The velocity of the internal 36

wave, _i n, is expressed as3738

g._2__1/39 _ (15)40

i n 2._2

C_1 /.d1

Cd2/

41

42 The situation becomes different when solitons propagate through a medium having arbitrary stratifi-

43 cation. In this case, Korteweg-de Vries introduced a new analytical solution [15] for internal soliton.

44 Let us consider the situation where internal solitons propagate through arbitrarily stratified medium

45 with no rotation, that is, the Coriolis parameter, fc, is zero.46 Under small dispersion and small nonlinearity, the quadratic Korteweg-de Vries equation for

47 internal waves propagating along the +x direction is expressed as [3739]:48

@

_ @_ @_

@3

_49

(16)

_i

n C _1_ C _250 @t @x @x @x3

51 In Equation (16), the constant quantities, _1 and _2 are expressed as

52

53

_1 D

3_i n (17)

54


9/26

ColorOnline,B&WinPrint

F1


DOI: 10.1002/dac


10/26


01

_2 D

_i n

ep

2

(18)02

03

04 In Equations (17) and (18), the parameters %and e are nondimensional environmental parameters.05

The solution of Equation (16) is as follows [40]:

06

07

sech2

x _V t08 _.x; t /

D

_ (19)

09 0 _ _10

11 This solution is alternatively known as the SECH solution. In Equation (19), Vis the nonlinear12 speed of soliton, and is the characteristic width. The parameters V and can, respectively, be

13 expressed in terms of _1 and _2 as14

_1_015

V D_i nC (20)31617

18

12_2

19

2

(21)D20

_1

_021

22 In Equation (23), _2 represents the positive oceanic gravity waves. However, in the shallow sea,

23 with strong mixing, _1 and _2 are positive. From Figure 1, we write the expressions for _1 and _2

24 as follows:

25

26

27

28

29

30

31

32

33

34

35

_

D

3_ d1_d2 ; f or 0 > > d (22)

1 2i n

d1d2 _ 1

_2 D

_

i n

d1

d2

; f or _d1 > >_D (23)6

4. NETWORK ARCHITECTURE

36 In this paper, we have considered the shallow water oceanic coastal region. In this region, the nodes

37 are deployed in such a manner that the sink nodes float on the water surface, and the other nodes

38 stay at the sea-bed. We have considered single hop communication, that is, the source nodes at the39 sea-bed directly communicate with the surface sink. The source nodes are equipped with long-range40 acoustic transceiver, whereas the surface sinks are equipped with both acoustic and RF transceivers.

41 The communication architecture is shown in Figure 1.42 The following assumptions are made in this work.43 _ The deployed nodes are homogeneous in nature.44 _ The nodes in the sea-bed span an area covering the maximum transmission range of a node.45 _ The nodes have the capability of directional transmission.46 _ The sink nodes can be stationary as well as mobile.47

_ Mutual acoustical interference among the source nodes is negligible.48

49 However, regarding communication, two issues may arise. One issue is the bandwidth decrement

50 of acoustic signal during propagation through the channel. Another is the nodes mobility for which

51 Doppler effect results. Bandwidth problem is minimized to some extent at the receiving end. As a

52 sensor node has the capability of amplification, the destination node will amplify the received signal.

53 As the source and destination nodes are not affected by solitons, there is a less probability of arising

54 Doppler effect.


11/26


DOI: 10.1002/dac


12/26

01

02

03

04

05

06

0708

09

10


Table I. Acoustic frequency ranges throughinternal solitons.

Frequency band Frequency range

High frequency f >50 kHzMid frequency 1 kHz 6f 650 kHz

Low frequency

f 61 kHz

5. PROPAGATION OF WIRELESS ACOUSTIC SIGNAL THROUGH INTERNAL SOLITON

11 The internal solitons strongly scatter the underwater sound signal, and this scattering has strong12

dependency on the frequency, f, of the signal used. Internal solitons mainly affect three frequency

T113

bands [15], which are given in Table I.14 It is yet an open research problem to the researchers particularly which of the tracers in the15

internal wave scatters acoustic signal. In this paper, we have assumed that sediment carried out by16

internal solitons is the tracer responsible for scattering of acoustic signal. In shallow water, signal17 of frequency of the order of few hundred Hertz (Hz) faces less attenuation [15]. However, the opti-18 mum operating frequency of UWASNs lies in the range of few kH , thereby increasing the signal

19 attenuation. Interaction of acoustic signal with solitons leads to fluctuation in intensity, I, and pulse20 travel time, _. Along with, Iand _, the signal attenuation affects amplitude and phase.

21

22

23

24 5.1. Problem formulation for signal field calculation25 An internal soliton makes changes to the medium properties along its direction. Therefore, the range26 dependent Helmholtz equation is written as [41]:27

28

29

30

31

32

33

34

35

36

37

38

39

r2.r; / C _

2.r; / D 0 (24)

In Equation (24), ris the range, .r; /is the range and depth (z) dependent scalar function, and_.r; / D

! is the wavenumber of the propagating acoustic signal, which also signifies theC .r;/

range dependent properties of a wave propagating through the medium. Using the method ofpartial separation of variables, the scalar function .r; /is expressed as [41]:

X

.r; / D


13/26


01 In Equation (27), the parameters,CmlandDml, are the mode-coupling coefficients, and are02 functions of the density,_./, and the

eigen function.

03 We obtain the adiabatic mode solution corresponding to the single mode from coupled mode 04 equations by

setting the coefficients,CmlandDml, to zero, for acoustic field propagating along the 05 internal wavefront as

06

07

p.r; / D_

_

_

12 _

08

2

ei4A.r; /.r/.r/ (28)r

09

10 In Equation (28), Ais the amplitude of signal, is the phase, and the signal attenuation is accounted

11 by the parameter, . The aforementioned parameters are presented as follows:

1213

.s/.r/

14

A.r; / (29)

r

15

D s0 k.r/ dr

16

R

17

18r

i k.r / d r

19

0(30).r/ De R

2021

_

r

.r / d r22

.r/ 0

De ( 1

23

24 where ris the range of a sensor node over which communication takes place.

25

26 5.2. Modeling amplitude and attenuation coefficient of wireless signal27 In general, any wave is associated with two fundamental properties: amplitudeandphase. During28

propagation through internal soliton, the amplitude and phase of an acoustic signal get modified. In29

this section, we have modeled the amplitude and phase of an acoustic signal propagating through a30

medium consisting of internal solitons.31

Amplitude:32

The analysis in Section 5.1 for an acoustic field p.r; /is valid when signal propagation takes place33 across the soliton wavefront. However, as evident from the Figure 1, the propagating acoustic signal

34 makes angles with the propagation direction of an internal soliton. The schematic representation of

F2 35 the scenario is shown in Figure 2. From the figure, we observe that the wavenumber makes an angle 36 with the propagation direction of the soliton. So, the component of the wavenumber along the

37direction of propagation of solitons is

k cos . Therefore, the integral,38 E

39 r40

dDZ

0

k

(32)41 drr4243

44

45

46

47

48

49

50

51

52 (a) Signal propagation through internal solitons (b) Network architecture of UWASNs53

Figure 2. Signal propagation through internal solitons.54


DOI: 10.1002/dac

ColorOnline,B&W

inPrint


14/26


01

02

03

04

05

06

0708

09

10

11

12

13

14

15

16

17

18

19

20

21

22

23

t an .D=r /

_

_f

_cot :

d

D Z0

2

:t an (33)C

0 D

t an .D=r /

D .f =C0/ Z0 d (34)

We calculate the eigen functions, .s/and .r/. For source depth, s, of 200 m and receiver

at the surface, that is, r D0m, the eigen functions are expressed as

. /

D

cos __D 1 (35)

Ds _

_D _

. /

D

cos __0

_ D

1 (36)

r _ DTherefore, the numerical value of phase, A, is expressed as

AD

1 (37)

Dan r

fdC0

R

0

24 Attenuation coefficient:25

During the propagation of acoustic signal through internal solitons, scattering of signals occurs.26

This scattering occurs due to the interaction of signal with suspended sediments carried by waves.27

We have considered sand as the major contributor to sediments. We have assumed that there is a28

homogeneous suspension of sphere of radius Rs, and we have adopted the sphere scattering model.29

The attenuation coefficient due to scattering with the internal wave is expressed as [42]:30

31

D

3M

.k/

(38)

32

4Rs_s

33

34 In Equation (38), Mis mass concentration of suspended sediment, _s is the density of suspended

35 particles, whose value is taken as 2650kg_ms [43], Rs is the dimension of the suspended particles,

36 and is the normalized total scattering. We have considered four sets of frequencies, viz., 20, 30,

37 40, and 50 kHz, and four sets of radii for suspended sediment, which are100, 150, and 200_m.

38 The calculation of wavelength shows that the wavelength of the propagating acoustic wave is sub-

39 stantially greater than the circumferential length of the suspended particles. Therefore, scattering40

falls in the Rayleigh regime. For Rayleigh scattering, the function, , can be expressed as [44]:

41

_

e _1 g _142

D

2x4

C

(39)

3e 2g C14344

In Equation (39), xDkRs; eD39is the ratio of elasticity of sand grains to water, gis the ratio45 of density of the sand grains to water. Substituting the values of eand g, we obtain the final equation46 of as47

48D 0:26x

4

49

50

D

0:26_16_4 .f R

s

/4 (40)

c451

52

53 The variation of attenuation coefficient for different frequencies under different particle sizes is

54 shown in the Figure 3. F3


DOI: 10.1002/dac


15/26

10

01

Print

02

03in

04

B&W

05

06

Onli

ne,

0708

09

Color

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

A. K. MANDALET AL.

0.6

]

Rs= 100 [m]

Rs= 150 [m]

0.5 Rs= 200 [m]

()

0.4

Co-efficient

0.3

0.2

Attenuation

0.1

025 30 35 40 45 5020

Frequency [kHz]

Figure 3. Variation of attenuation coefficient.

Table II. Scattering parameter and attenuationcoefficient of signal for different particles size.

Rs._m) f (kHz)

100 20 0:0013 _10_6 0:0018 _10

_6

30

0:0065_10_6

0:0092_10_6

40 0:0021_10_6 0:0290_10_6

50 0:0500 _10_6

0:0708 _10_6

150 20 0:0065_10_6 0:0061_10_6

30 0:0328_10_6 0:0310 _10_6

40 0:1037_10_6 0:0980_10_6

50 0:2533 _10_6

0:2390 _10_6

200 20 0:0205_10_6 0:0145_10_6

30 0:0137_10_6 0:0730_10_6

40 0:3279 _10_6 0:2320 _10_6

50 0:8004_10_6

0:5670_10_6

33 As the intensity of scattering mainly depends on the particle size, and the frequency of acoustic

34 signal, we have observed attenuation coefficient for different particle sizes and variable frequencies.

35 Therefore, the intensity of the acoustic signal is expressed as

36

In Dp:p_37 (41)

38

39 C0 1 : exp : 100 (42)

40

D f t an 1

D

__ si n2

_f r41

d42 C0 0

43 R

44 C si n 100

_

45

D

0

: exp__:

(43)

f

D

si n2

46

47

Finally, the transmission loss of a propagating signal is given as48

T L D_10 logjInj49

(44)50

51

6. PERFORMANCE EVALUATION52

53 In this section, we have analyzed the performance of UWASNs in the presence of internal solitons

54 with respect to different metrics.


DOI: 10.1002/dac


16/26


17/26


01 ideal channel scenario, and ours consider the case of channel induced by internal soliton, we have02 taken

the Thorp model as the benchmark for performance comparison. In this work, we measure the 03

performance variation when the channel switches from ideal channel in Thorp to the internal soliton 04

induced one.0506

6.4. Result and discussion

07

08

The results obtained with respect to the three metrics are given in the succeeding text. The results

9 were compared with the results obtained using Thorp. To denote the Thorp scenario, we have used

10 the notation p0, and to signify the internal wave induced scenario, we have used the notation p1.11

F4 12 1) SINR: Figure 4 shows the variation of SINR with frequency under variable data rates. We

13 observe from the figures that irrespective of whether the nodes are static or dynamic, the SINR14 values for Thorp are always more than the internal wave induced one. A close observation reveals

15 that with the increase in data rate, SINR decreases. This is because as the data rate increases,16 the probability of occurrence of interference increases, and at the same time, more data will be17 associated with noise. It is seen from the figure that SINR in the static state is higher than in the18 dynamic state. With the increase in the mobility of nodes, SINR decreases. Additionally, we see19 from the figure that with the increase in frequency, SINR increases gradually at first and then20 decreases monotonically. This is because with the increase in frequency, the attenuation of sound

21 also increases. So, even an addition of a noise of low intensity with low intensity sound decreases

22 the SINR. We see from the figure that, as compared to the dynamic node, the rate of decrease of23 SINR with the increase in frequency is lower in case of static nodes under the Thorp scenario. For

24 the static scenario, the value SINR using the Thorp model increases by 9.35% for static situation of

25 the nodes, and by 9.79%, when the nodes are mobile.26

F5 27 2) BER: Figure 5 shows the plot of BER for different frequencies for different data rates. From

28 the figure, we observe that BER for the Thorp model is always higher than the internal soliton

29 induced channel model. We infer that as compared to the data rate of 1000 and 2000 bps, BER30

31

3233

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

Figure 4. Variation of SINR in presence of internal soliton. 54


DOI: 10.1002/dac

ColorOnline,B&W

inPrint


18/26


01 decreases rapidly with the increase in frequency for data rate of 500 bps. This is because for low

02

data rate, the volume of data sent per unit time is less, and therefore, less data is available for getting

03 corrupted. On the other hand, for the internal soliton induced case, there is no significant decrease04 in BER. The high BER is due to the interaction of acoustic signal with the internal soliton. We see05 from the figure that BER gradually decreases up to 30 kHz and then again increases up to 50 kHz.

06 This is because with the same data rate, when frequency increases, the volume of data passing

07

between the intermediate nodes becomes faster, as a result of which the interaction of data with08 channel increases. Again, we see that when the mobility of nodes increases, the BER increases.

09

10 0.044 0.048

11 0.042 0.04612

0.04

BER

0.044

0.042

13

0.0380.04

0.036

14

0.

038

0.034

0.

036

0.032

0.034

16

10

20

30

40

50

10

20

30

40

50Frequency [kHz] Frequency [kHz]

17

18

19 (a) BER for data rate 500 bps (b) BER for data rate 1000 bps20

0.054

21

0.052

22

0.05

BE

R

0.048

23 0.046

24

0.044

0.042

25

0.04

26 0.038 10 20 30 40 50

27Frequency [kHz]

28

29 (c) BER for data rate 2000 bps30

Figure 5. Variation of BER in the presence of internal solitons.31

32

33 0.243 0.195

34

0.

242 0.194

0.241

[sec]

0.193

35 0.24 0.192

36 0.239

Delay

0.191

37 0.238 0.190.237 0.189

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

Int. J.Co

mmun.Syst.(2014)

DOI

:10.

10

02/da

c

Copyright 2014 John Wiley & Sons, Ltd.

e ay or ata rate psc

gure . ar at on o e ay n t e presence o nterna so tons.


19/26

ColorOnline,B&W

inPrint

ColorOnline,B&W

inPrint


20/26

F6

Q3

F7


01

0.3 0.30203 0.25 0.25

04 0.2 0.2

05

0.15 0.15

06

0.1 0.1

07

0.05

0.0508

0 009 10 20 30 40 50 20 30 40 50

Avg.e

nergyc

onsum

ption/n

ode[J]

Avg.energyconsumption/node

[J]

10

Prin

tin

B & W O

nlin

e , C ol or

10 Frequency [kHz] Frequency [kHz]

11 (a) Avg. energy consumption/node for data rate 500 bps (b) Avg. energy consumption/node for data rate 1000 bps

12 [J]

0.3

consumption/node

13

0.2514

0.215

16 0.15

17 0.1

energy

18 0.05

19

Avg

.

0

20

10

20

30

40

50

Frequency [kHz]

21 (c)Averageg energy consumption/node for data rate 2000 bps

22

23 Figure 7. Variation of average energy consumption/node in the presence of internal solitons.

24

25

26

27

28

2930 When a moving node transmits data to its neighbor, the probability of corruption of data increases.31 Analytically, it can be shown that in the presence of internal solitons, BER increases by 15.76% for32

33 static nodes and 17.21% when the nodes are.

3435 3) Delay: Figure 6 shows the variation in delay with respect to frequency under variable data36 rates. From the figure, it is obvious that the time taken by a stat ionary source node to transmit data37

38 to sink nodes is much less than the time taken by the mobile source nodes. However, in the presence

39 of solitons, delay is more than in the absence of it. However, the delay profiles under the static40 situation of the nodes are almost similar for all the data rates. It can be inferred that the travel time41

does not depend on data rate. It depends on the frequency of the signal used and the kinematics of42

43 the nodes. On an average, the delay for the moving nodes increases by 0.24% in the presence of

44 internal solitons.45

464) Energy: Figure 7 shows the average energy consumed by each of the nodes due to commu-

47

48 nication. When the nodes are in the static mode, the energy consumption is more in the presence

49 of internal soliton than using the Thorp model. When the nodes move with the meandering cur-50 rent, they consume more energy. Again, we observe from the figure that compared to the case using51

52Thorp, the nodes consume more energy than in the internal soliton induced region. Numerically,

53 we can infer that in the presence of internal solitons, the energy consumption per node under static

54 situation increases by 53.31% and 52.78% for the moving nodes.


DOI: 10.1002/dac


21/26


01 7. CONCLUSION02

03 Our study focuses on the performance analysis of distributed UWASNs in the presence of inter-

04 nal solitons. The study was performed in the shallow coastal region of the ocean. To simulate the

05

environment, we used NS-3 simulator. We took 200 m depth of the simulation region. We have ana-

06

lyzed the performance of UWASNs in terms of some performance metrics, viz., SINR, BER, delay,

07

and energy consumption per node. These metrics are observed to be affected by the presence of08 internal solitons. To calculate transmission loss, we have calculated the intensity of acoustic signal

9 propagating through internal wave.10

Based on the analysis, it is observed that in the presence of internal solitons, SINR decreases11

by 9.57 %, and BER increases by 16.49 %, delay increases by 0.24 %, and energy consump-12

tion per node increases by 53.05 %. In the future, we plan to perform similar studies with other13

oceanic phenomena, such as the effects of raindrops on near-surface oceanic environments. Follow-14

ing the present analytical and simulation based studies, we also plan to perform real-life field test15

experiments in the future.16

17

18

ACKNOWLEDGEMENT

19

This work is partially supported by a grant from the Department of Electronics and Informa-

20 tion Technology, Government of India, grant no. 13(10)/2009-CC-BT, which the authors gratefully21

acknowledge.22

23

24 REFERENCES

25 1. Watfa MK, Selman S, Denkilkian H. Uw-mac: an underwater sensor network mac protocol.International Journal of26 Communication Systems 12

t h March 2010;23:485506.

27 2. Wahid A, Lee S, Kim D. A reliable and energy-efficient routing protocol for underwater wireless sensor networks.

28 Internation Journal of Communication Sys tems 11t h

October 2012.

29 3. Obaidat MS, Misra S.Principles of Wireless Sensor Networks. Cambridge University Press, 2014.4. Dhurandher SK, Obaidat MS, Gupta M. An efficient technique for geocast region holes in underwater sensor

30 networks and its performance evaluation. Simulation: Modeling Practice and TheorySeptember 2011; 19(9):31 21022116.

32 5. Akyildiz IF, Pompili D. Underwater acoustic sensor network: research challenges.Ad Hoc NetworksFebruary 2005;

33 3(3):257279.34 6. Heidemann J, Stojanovic M. Underwater sensor networks: applications, advances and challenges.Philosophical

Transactions of the Royal Society August 2012;370:158175.35

7. Heidemann J, Mitra U, Preisig J, Stojanovic M, Zorzi M. Underwater wireless communication networks. IEEE36 Journal of Selected Areas in Communication December 2008;26(9):16171619.37 8. Erol-Kantarci M, Mouftah HT, Oktug S. A survey of architectures and localization techniques for underwater38 acoustic sensor networks.IEEE Communications Surveys and TutorialsMarch 2011; 13(3):487502.

39 9. Nicopolitidis P, Christidis K, Papadimitriou G, Sarigiannidis PG, Pomportsis AS. Performance evaluation of acousticunderwater data broadcasting exploiting the bandwidth-distance relationship.Jounal of Mobile Information Systems

40 7t h November 2011; 7(4):285298.41 10. Erol-Kantarci M, Oktug SF, Filipe L, Vieira M, Gerla M. Performance evaluation of distributed localization42 techniques for mobile underwater acoustic sensor networks.Ad Hoc Networks2011; 9(1):6172.43 11. Erol-Kantarci M, Mouftah HT, Oktug SF. Localization techniques for underwater acoustic sensor networks.IEEE

44 Communications Magazine 2010;48(12):152158.45 12. Misra S, Dash S, Khatua M, Vasilakos AV, Obaidat MS. Jamming in underwater sensor networks: detection and

mitigation.IET Communications2012; 6(14):21782188.46

13. Dhurandher SK, Obaidat MS, Gupta M. A novel geocast technique with hole detection in underwater sensor net-47 works.Proceedings of ACS/IEEE International Conference on Computers and Applications, Hamammet, Tunisia,48 May 2010.49 14. Dhurandher S, Obaidat MS, Goel S, Gupta A. Optimizing energy through parabola based routing in underwater

50 sensor networks.Proceedings of the IEEE GLOBECOM 2011 - Communications QoS, Reliability, and Modeling51 Symposium, 2011.

15.Apel JR, Ostrovsky LA, Stephanyants YA, Lynch JF. Internal solitons in the ocean and their effect on underwater52 sound.Journal of Acoustical Society of AmericaFebruary 2007; 121(2):695722.53 16. Kostin A, Oz G, Haci H. Performance study of a wireless mobile ad-hoc network with orientation dependent internode

54 communication scheme.International Journal of Communication Systems17t h

January 2014; 27:322340.

Q4

Q5

Q6

Q7

Q8


DOI: 10.1002/dac


22/26


23/26

DOI: 10.1002/dac


24/26


25/26

Highligh a word or sen ence.

Click on the Replace (Ins) icon in the

Annotations section.

Type the replacement text into the blue box

that appears.

Highligh a word or sen ence.

Click on the Striketh

(DeAnnotations section

3. Add note to textTool for highlighting a section

to be changed to bold or italic.

Highlights text in yellow and opens up a text

box where comments can be entered.

How to use it

Highlight the relevant section of text.

Click on the Add note to text icon in

the Annotations section.Type instruction on what should be

changed regarding the text into the yellow

box that appears.

4. Add sticky noteToolfo

specific points in the tex

Marks a point in the

w needs to be highlig

How to use it

Click on theAdd sticky

Annotations section.

Click at the point in thew should be inserted.

Type the comment in

yel appears.


26/26

7. Drawing MarkupsTools for drawing shapes, lines a

annotations on proofs and commenting on these m

Allows shapes, lines and freeform annotations to be drawn

pr comment to be made on these marks..

How to use it

Click on one of the shapes in the

DrawingMarkups section.

Click on the proof at the relevant point and

draw the selected shape with the cursor.To add a comment to the drawn shape,

move the cursor over the shape until

an arrowhead appears.

Double click on the shape and type

any text in the red box that appears.

For further information on how to annotate proofs, click on the Helpmenu to reveal a list o

Documents

Internal Soliton