8/3/2019 05076827

1/5

The Evaluation of Modulation Techniques forUnderwater Wireless Optical Communications

Meihong Sui

Key Lab of Submarine Geosciences and ProspectingTechniques

Ocean University of ChinaQingdao China

sui.meihong@yahoo.com.cn

Xinsheng Yu, Fengli Zhang

Key Lab of Submarine Geosciences and ProspectingTechniques

Ocean University of ChinaQingdao China

Abstract Currently, there are lots of efforts at using

underwater vehicles, gliders and moorings for the spatial

and temporal measurements in oceanography research.

Sensor data collected by these platforms is usuallyinternally recorded and then transmitted via a cable or

wireless communication. Traditional acoustic links are

fundamentally bandwidth limited to low rates of bit per

second (bps). Optical methods are well posed to provide

an alternative solution for high bandwidth

communications in undersea. In this paper, we examined

several modulation technologies for undersea

environment application. Through modeling and

simulation, the advantages and limitations of these

modulating configurations are discussed. It is

demonstrated that the Pulse Position Modulation (PPM)

is better suited for low powered undersea systems and the

phase shift keying (PSK) yields the best performance in

term of bandwidth and error performance with poor

power efficiency.

Keywords-underwater wireless optical link, optical

channel, modulation techniques; performance evaluation

I. INTRODUCTIONOceanography is entering a new stage in which the

research and development activity has been extendedfrom coast to deep-ocean. Therefore, it is desirable todeploy long term underwater observatory with multisensors to monitor the physical, chemical, geologicaland biological progress with many spatial and temporaltime scales. To gather data from these facilities is acrucial task. Although radio frequencies have enjoyedlarge success in free space, they experience highattenuation in water and typically not used forunderwater communication. Acoustic technology hasadvantage to transmit data over a long distance in water.However, the attenuation of the acoustic carrier and theeffects of multi-path reflection will ultimately limit thedata rate and bandwidth for a large amount datacommunication and even at the short range the bandwidth is limited to sub-Mbps. This brings a"bottleneck" problem for a large amount of data

collection (such as multi-sensor data, image information,etc.). Wireless optical communication have shown promise of supporting large bandwidths, high data

transfer rate, small in size, low power consumption,immune to electromagnetic interference. Thus,underwater wireless optical communication can be analternative method for fast data transmission. Byincorporating optical wireless system into autonomousunderwater vehicle (AUV), we can make use of AUVto approach to the location of seabed observatories ormooring systems to gather logged data and then totransport the data package to research ships within shortrange. It provides an alternative solution for real timedata passing and monitoring mission coordinationwirelessly with free swimming underwater vehicle in aexpand distance.

The most common optical communications

modulation scheme used for free space optical link ison-off key (OOK) and this modulation method has beenusually adopted in underwater wireless optical link due

to its simplicity in system implementation[1

. Whilethe other modulation techniques have been studied forfree space optical communication extensively recently,the feasibility studies of these modulation techniquesfor underwater optical channel have been rarelyreported. Compared to atmospheric propagation, oceanwater is more complex medium for light propagationand communication underwater is very challenging. As practical underwater systems are constrained in size,weight and power, some tradeoffs must be consideredin the choice of modulation format for underwater

optical communication system in terms of overall performance, communication distance and powerconsumption. The work reported in this paper attemptsto fill this gap.

,2,3]

In this paper, we evaluated the LED(light EmittingDiode) based transmitter with direct detection forrealizing modulated optical signal for underwatercommunication based on inherent optical properties ofocean water. The performances of different modulationschemes for underwater optical channel are investigated.

2009 International Conference on Communication Software and Networks

978-0-7695-3522-7/09 $25.00 2009 IEEE

DOI 10.1109/ICCSN.2009.97

138

2009 International Conference on Communication Software and Networks

978-0-7695-3522-7/09 $25.00 2009 IEEE

DOI 10.1109/ICCSN.2009.97

138

2009 International Conference on Communication Software and Networks

978-0-7695-3522-7/09 $25.00 2009 IEEE

DOI 10.1109/ICCSN.2009.97

138

2009 International Conference on Communication Software and Networks

978-0-7695-3522-7/09 $25.00 2009 IEEE

DOI 10.1109/ICCSN.2009.97

138

2009 International Conference on Communication Software and Networks

978-0-7695-3522-7/09 $25.00 2009 IEEE

DOI 10.1109/ICCSN.2009.97

138

2009 International Conference on Communication Software and Networks

978-0-7695-3522-7/09 $25.00 2009 IEEE

DOI 10.1109/ICCSN.2009.97

138

8/3/2019 05076827

2/5

The simulated results of bit error rate, power efficiency,data rate and implementation complexity is discussedand compared. It is shown that the pulse-positionmodulation (PPM) and phase shift keying (PSK) aresuitable candidates for underwater wireless optical linkapplications.

II. CHARACTERISTICS OF NDERWATERPTICALHANNEL

The optical properties of sea water are function ofwater salinity, water temperature, and concentration ofdissolved organic and inorganic matter, suspended particles and organisms. The attenuation of the light beam in sea water is much more serious than in theatmosphere. Lighting power attenuated in water ismainly dominated by wavelength dependent processes:absorption and scattering. The main cause of lightabsorption in water is excitation of vibration state ofthe water molecule by photons and other dissolved particles and detritus. Scattering of light refers to

processes in which the direction of the photon ischanged and it can take place either on molecules or ondissolved particulate. In the literature, absorption andscattering coefficients are extensively used tocharacterize the light transmission and the attenuation

coefficient ( )TK is defined as :]4[

( ) ( ) ( )T A S K K K (1)

Where ( )AK is the total absorption coefficient

and ( )Ks

is total scattering coefficient.

The light power of the receiver end after transmittedat a distance rcan be expressed as follows:

( )

2 2T

R T

A K r P e P

r

where A is the area of the reception, is beam

divergence angle, is transmitting power,TP ( )TK is

the total diffuse attenuation coefficient for spectralirradiance. In this work, the total diffuse attenuation

coefficient ( )TK were assumed to be influenced by

the chlorophyll so that the attenuation was re-written asfollowing:

( ) ( ) ( ) ( ) ( )T w CDOM chl d K K K K K ()

Where, ( )WK ( ) is the spectral irradiance

attenuation coefficient for clear ocean waters,

1m

( )CDOMK ( ) is the spectral irradiance

attenuation coefficient for colored dissolved organicmatter, which usually named yellow substance.

1m

( )ch lK ( (mg pigment ) is the specific

spectral irradiance attenuation coefficient due to

chlorophyll-like pigments,

1m 3m

( )dK is the spectral

irradiance attenuation coefficient for drossy.According to [12], and combined with [5], the total

attenuation coefficient can be defined as :

1

4.322

1.7

0.3

( ) ( ) ( ) 62.6039 exp(0.12327 0.0189 )3.6402 exp(0.12343 0.01105 ) 0.005826(400/ )

0.01739 exp(0.11631 ) 1.151302(400/ )

0.76284 exp(0.03092 ) 0.341074(400/ )

T w c c c

c c

c c

c c

K a a C C C C C

C C

C C

1cC (4)

2

4.322

1.7

0.3

( ) ( ) ( ) 62.6039 exp(0.12327 0.0189 )

3.6402 exp(0.12343 0.01105 ) 0.005826(400/ )

0.01739 exp(0.11631 ) 1.151302(400/ )

0.76284 exp(0.03092 ) 0.341074(400/ )

T w c c c

c c

c c

c c

K a a C C C

C C

C C

C C

1cC (5)

cC (mg pigment3m ) is the average concentration

of Chl a and bio-pigments in the ocean water to a depth

of 1 attenuation length.

The assumption was made that the chlorophyll

was distributed homogeneously throughout the water

column, and the transmission distance r is 10m.

According to equations (4) and (5), we analyzed the

affects of attenuation coefficient at differentchlorophyll concentration on visible light theoretically

and the results are shown in figure 1and 2. It is

indicated that with different wavelengths of light, the

transmission characteristic is diverse when transmitting

in different chlorophyll concentration.

Figure 1. Chlorophyll concentration of less than 1, the opticalpower of different wavelengths

139139139139139139

8/3/2019 05076827

3/5

Figure 2. Chlorophyll concentration of more than 1, the optical power of

different wavelengths

From the figures above we can see when chlorophyllconcentration is lower than 1, the blue, green and cyanspectrum have a better communication window, but whenin high concentration chlorophyll, the attenuation effect on blue, green and cyan light becomes serious. And the red

spectrum has better characteristic in high chlorophyllconcentration.

III. MODULATION CHEME FORWIRELESS PTICALOMMUNICATIONS

The common modulation schemes used for underwaterenvironment are intensity modulation and direct detection.The optical power is controlled to change the pulse rate,

width, frequency and location. The basic modulationtechniques for optical wireless system uses non-coherentdetection methods that only the presence or absence ofpower is ascertained and no phase information is recovered,such as amplitude shift keying (ASK), and 2ASK alsonamed On-Off Keying (OOK), pulse position modulation(PPM), the differentially coherent or coherent techniques,for example frequency shift keying (FSK) and phase shiftkeying (PSK), needs further phase symbols comparison.

ASK, OOK and PPM modulation formats are typicallyused in a simple direct detection scheme and can beimplemented at lower complexity. For the FSK and PSKmodulation schemes, they can be implemented by usinginterferometers in conjunction with balanced optical

receivers. Table compares the selected modulationschemes described above.

TABLE I. DIGITAL OPTICAL COMMUNICATIONS MODULATION FORMAT

IV. EVALUATION OF ODULATION ECHNIQUE FORNDERWATEROPTICAL OMMUNICATIONS

For the underwater optical channel, the quality ofreceived signals is affected by two factors. One is caused bywater medium during light transmission; the other is

additive white Gaussian noise . The choice of modulationtechnique for real time applications needs to tradeoff thesignal to noise ratio (SNR), power performance andbandwidth efficiency.

]6[

A. Signal to noise ratio for underwater channelIn this study, compared some different wavelength

(660nm, 530nm, 505nm, 470nm), and taking the

transmitted output power 136 mW with the divergence

angle of 0.6 mrad. Using the S5493-01 photodiode produced by Hamamatsu Company as the receiver.Therefore, the signal to noise ratio at the receiver is givenby the following formulae [7].

( )

28

GP A K rr t r TS e

r hc Rw

(),

Wherer

is the detector quantum efficiency, G is the

photo multiplication gain of the detector, is the

transmitted power of the LED, A is the area of the

tP

r

Modulation format Threshold Complexity of implementation

OOK 1( )

0 0

isp t

Half of the amplitude low

FSK 1( ) ( ) c o s ( )

0

M p t A g t t

j jj

No need to set high

PSK 1( ) ( ) cos( )

0

Mp t A g t t

j j

j

The phase in one symbol compared to the phase inthe previous symbol

middle

L-PPM

[ 1] / , /

0

i t m T L mT L s f f p

manother

{1, 2 , ... }m L

k( 0 1k times of the amplitude, selectedbased on specific value of L

lower

140140140140140140

8/3/2019 05076827

4/5

receiving aperture, R is the data rate, r is the transmission

distance, h is Planck constant , ( )TK is the total diffuse

attenuation coefficient for spectral irradiance at 0.01cC .

The results of the relationship between transmissiondistance and signal to noise ratio as the prerequisite atdifferent chlorophyll concentration is showed in Figure 3.Because we just study the optical transmission with bluelight for more than 10m, we only consider the SNR between-10dB and 30dB in the rest of study.

Figure 3. For different attenuation coefficient, the relationship betweentransmission distance and the channel SNR

B. Error Probability performance EvaluationMost of optical wireless communications to date have

made use of a simple OOK in direct detection scheme. ForOOK demodulation format, it is used to compare the

received voltage with the threshold to decide"1" or "0" .In an AWGN channel model, the received voltage

is , is Gaussian process. For the

data bit "1", the probability density of

is

]8[

( ) "1"

( ) "0"( )

i n ts c

p t nc t

)(tc

n

)(tx

1 2 2( ) exp[ ( ) / 2 ]1 2

p x x is

, and for the data

bit "0", the probability density of

is)(tx 1 2 2( ) exp[ /2 ]0 2

p x x

. If the judgment threshold

is set assi

2

1 , the bit error rate is defined as following:

1 1( ) (

( ) 2 22 2 2 2

i Ss P erfc erfce ook

) ()

According to [8], in additive Gaussian white noisechannel, the bit error rate of 2FSK coherent modulation and2DPSK coherent modulation are given by followingequations:

1

( ) 2 4

S P erf

e FSK

1(1 )

( ) 2 2 2

S S P erfc erfc

e D P S K ()

For the L-PPM modulation in the Gaussian white noisechannel, there are many performance evaluation methods

for the bit error rate . In this paper, the evaluation of

BER is defined according to :

]11,10,9[

[9 ]

1 1 1 1[ ( ) (

( ) 2 22 2 2 2

k L k P er )] fc LS erfc L

e L PPM LS

()

As the performance of BER is depended on the

threshold value , we evaluated the different k values and

the result for 4 PPM and 8 PPM is showed in Figure 3. Itcan be seen from Figure 4 that when the k value is from0.45 to 0.65, the error rate research low. Thus, the k= 0.5 isused in the rest of study.

ski

Figure 4. The relationship between BER and different k values

We simulate the bit error rate with different modulationformat at the different SNR condition and 4 PPM and 8PPM is employed for PPM scheme. The numericalsimulations of the performance of these modulationschemes for undersea communication are depicted in Figure

5. It is shown that the 2DPSK and 8PPM offer the betterperformance than other modulation techniques.

Figure 5. The BER performance of different modulation schemes withSNR

C. Data rate and powerFor underwater optical communications applications,

the system is powered by its own batteries. In order toimprove underwater system performance, it is desirable tomake the system as small and light as possible. Much likeradio frequency communication systems, the power

c ()

141141141141141141

8/3/2019 05076827

5/5

performance and data transfer efficiency is also mainconsideration of the choice of modulation techniques forunderwater optical communication systems.

A frequency shift key (FSK) modulator generates aspecific frequency carrier wave for digital "1", and adifferent frequency carrier wave for digital "0", Thedeficiency of FSK for underwater application is the opticaltransmit power required as the transmitting duration isalways on.

Although PSK modulation is good at bit error rate andbandwidth, but PSK modulator generate an in phase signalfor digital "1", and an out of phase signal for a digital "0".The difference coherent demodulations are needed tocompare the current phase with the previous phase. Thus,PSK demodulators are complex and power inefficient forembedded system of underwater applications.

Although there is no full use of the frequency or phaseinformation in OOK and PPM modulation, the design ofreceiver and transmitter circuit is simple and it is suitablefor the underwater equipment integration. The datathroughput of PPM modulation is smaller than OOKmodulation, but the required receive power is just

1

log22

LL

of OOK modulation at the same error rate

performance. It means that PPM could transmit longerdistance than OOK at the same transmitting powercondition. Let P represent the smallest pulse width, thecomparison of different modulation techniques is shown in

table .

TABLE II. THE SIMPLE CONTRAST OF TRANSFER RATE ANDIMPLEMENTATION COMPLEXITY

OOK FSK DPSK 4-

PPM

8-PPM

Maximumrate

1/(2P) 1/(2P) 1/(2P) (1/2P) (3/8P)

Transmitpower

Middle Higher Highest Low Lowest

Complexityof

modulation

Low Higher Highest Lower Lower

V. CONCLUSIONSIn this paper we evaluated the performance of different

modulation schemes for underwater optical wirelesscommunication with marine bio-optical model. Thesimulation results show that the red light has bettertransmission characteristics in waters with higherchlorophyll concentration. It is suggested that for

underwater application, it is required to trade off the water property and wavelength to improve the communicationefficiency.

Although most reported underwater communicationworks use OOK modulation technique because of itssimplicity for implementation, our simulated results arguethat OOK has the disadvantages in power efficiency andcontrol capacity of the error rate for underwater optical

channel. DPSK has good error control capability and highbandwidth, but due to the using of optical interferometer inreceiving system, it consumes large power and morecomplex to implement in embedded device. As thetransmitter in a FSK system is always on, the powerconsumption is relatively higher than other techniques and itis undesirable option for undersea optical system. In

consideration of bit error rate performance, bandwidthrequired and the optical power in underwater opticalwireless applications, and also implementation complexity,we have shown that the PPM modulation technique is agood option for designing underwater optical wirelesssystem. In order to achieve lower error rate performance,one can increase the value of L. For high bandwidthapplication, the improve PPM modulation such as DPPMcan be used to increase its bandwidth performance.

ACKNOWLEDGMENT

We are grateful for the support of National "863" Hi-Tech program under grant number 2006AA09Z177.

REFERENCES

[1] M.Wang, W.Liu, Study on Optical Characteristics ofSeawater in Blue-green Laser Uplink Propagation Model.Journal of East China Shipbuilding Institute (Natural ScienceEdition), Vo1.19, No.1, Feb, 2005

[2] Y.Zhou, Y.Liu, Y.Zhao, The Current Situation andDevelopment Trend of VisibleLight WirelessCommunication by LED, Journal of Huaiyin Institute ofTechnology. Vol.15 No.3 Jun, 2006.

[3] MKahn ,Joseph, Modulation and Detection Techniques forOptical Communication Systems

[4] B.Cochenour, L.Mullen, A.Laux, Effects of MultipleScattering on the Implementation of an Underwater WirelessOptical Communication Link, Naval Air SystemsCommand, Patuxent River,MD 206-70-1161.

[5] C.S.Raymond and S.B.Karen, Optical classification ofnatural water. U.S. Department of Commerce, NOAA, National Environmental Satellite Service grant 04-6-158-44033 and contract UCSD-5-35406

[6] F.Schill , R. Z.Uwe , J.Trumpf, Visible Spectrum OpticalCommunication and Distance Sensing for UnderwaterApplications, Proc. of the Australasian Conference onRobotics and Automation, December 6-8, Canberra,Australia2004.

[7] Z.Xu. Approximate Performance Analysis of WirelessUltraviolet Links.

[8] C.X.Fan, F.Zhang, B.Xu, and C.Wu. Communications theory[M], Defense Industry Press, 2001.

[9] H.Ma. Research on Optical Modulation and DemodulationTechniques in Mobile Atmospheric Laser Communication, National University of Defense Technology, November2003.

[10] D.Audeh Malik, M.Kahn Joseph,and R.Barry John.,Performance of Pulse-Position on Measured Non-DirectedIndoor Infrared Channels, IEEE Transactions onCommunications, Vol. 44, No. 6, June, 1996.

[11] C.M.David Lee, and M.K.Joseph, Coding and Equalizationfor PPM on Wireless Infrared Channels, IEEE Transactionson Communications, Vol. 47, No. 2, February, 1999

[12] V.l.Haltrin Chlorophyll-based model of seawater opticalproperties Applied Optics Vol.38,No.33 20 November 1999

142142142142142142