Jingwei Yin,*,Pengyu Du,Guang Yang,and Huanling Zhou

1.Acoustic Science and Technology Laboratory,Harbin Engineering University,Harbin 150001,China;2.College of Underwater Acoustic Engineering,Harbin Engineering University,Harbin 150001,China

Space-division multiple access for CDMA multiuser underwater acoustic communications

Jingwei Yin1,2,*,Pengyu Du1,2,Guang Yang1,2,and Huanling Zhou1,2

1.Acoustic Science and Technology Laboratory,Harbin Engineering University,Harbin 150001,China;
2.College of Underwater Acoustic Engineering,Harbin Engineering University,Harbin 150001,China

Time reversal mirror(TRM)can use the physical characteristics of the underwater acoustic(UWA)channel to focus on the desired user in multi-user UWA communication.The active average sound intensity(AASI)detector can estimate all azimuths of users with the same frequency band at the same time in order to achieve directional communication by vector combination. Space-division multiple access(SDMA)based on TRM combined with the AASI detector is proposed in this paper,which can make the capacity of the code division multiple access(CDMA)UWA system significantlincrease.The simulation and lake test results show that the 7-user UWA multi-user system can achieve low bit error communication.

underwater acoustic(UWA)communication,spacedivision multiple access(SDMA),time reversal mirror(TRM),active average sound intensity(AASI)detector.

1.Introduction

Code division multiple access(CDMA)technology can achieve underwater acoustic(UWA)multi-user communication,however,the complexity of the UWA channel[1–4]seriously affects the system performance.In the wireless multi-user communication system,multiple access interference(MAI)will be more and more serious with the increasing number of the users and thus become a major interference[5],while in the UWA multi-user communication system,multipath delay spread in shallow waters which is usually on the order of 10 ms to 30 ms[6] makesMAIbecomemoreserious:i)fortheundesireduser, its multipath interference can be regarded as a new MAI; ii)for the desired user,multipath interference causes the received symbols to suffer from inter-symbol interference (ISI).Over the last decades,much effort has been directed at developing adaptive channel equalizers to remove the ISI and compensate for the channel variations[7–10]. These techniques,however,are quite demanding in terms of computational complexity,algorithm stability,and selection of channel parameters[11–13].And these methods are commonly done in the baseband which means that when one focuses on the desire user in channel equalization,other users with their UWA channels can only be treated as interference.

Passive time reversal mirror(PTRM)[14–18]usually uses a large-aperture vertical array to receive signals and it is based on the theory of signal propagation in a waveguide,which says that the Greens function convolved with its time-reversed conjugate,summed over a vertical array of receivers(denoted as the Q-function)is approximately a delta function in space and time[11].This method is relatively simple to achieve in an algorithm and easy to be realized.However,it is toocomplicatedto be usedin UWA communication,which pursues simple nodes and lowpowerconsumptionespeciallyinmulti-userUWA communication.The single-element time reversal mirror(STRM) and focusing gain were analyzed in[19].Although STRM cannot get the spatial gain processed by the TRM array processing bringing higher sidelobe,it can still superimpose multi-path signals generated by the acoustic channel.

The system function of the UWA channel is very sensitive to environmental parameters and the orders of sensitivityare:changesinverticalposition,thicknessofthewater layer,horizontal position,and sound speed in water layer. In the multi-user UWA system,the factors such as the horizontal distance and the vertical depth of each node,the sea and seabed ups and downs are different,which makes the channel impulse response function of each user different considerably.Hence,the cross-correlation of the users channelimpulseresponsefunctionis weak.TheSTRMcan use the physical characteristics of the UWA channel to fo-cus the desired user and shield undesired users at the same time and the MAIs of the multi-user system from undesired users and UWA channels will be effectively reduced [20,21].

Acoustic vector sensors[22–24]are capable of measuring three orthogonal particle velocity components of the acoustic field in addition to the scalar acoustic pressure,at a collocated point in space,which increases the number and variety of information and expands the postposition signal processing space[25,26].A single vector sensor can measure the target azimuth while only pressure hydrophone array can do the same thing.It has good spatial directivity which can suppress isotropic noise and the differentorientationsoftheMAI.Theactiveaveragesound intensity(AASI)detector[27]can simultaneouslymeasure the azimuths of multiple users with the same frequency band.Using the estimated azimuth to complete vector sensor electronic rotation can realize directional communication[28,29],which can reduce the MAI and improve the communicationquality.Space-divisionmultiple access (SDMA)based on STRM combined with the AASI detector is proposed in this paper,and the advantage of it is that eachpartofthealgorithmis verysimpleto berealized.The STRMcancompletethematchingUWA channelbysimple channel estimation without any channel information.

It is noted that the TRM normally requires an array of receivers.TheadvantageofCDMA is that a single receiver is often sufficien[30,31].In this context,the STRM is basically a matched filte,or a correlator.Since the filte uses the channel impulse-response,the method is still referred to as PTRM.Obviously,the method can be applied to an array of receivers with the added benefi of minimal signal fading and reduced phase variance[32,33].

In this paper,the basic principle of the AASI detector is firs introduced in Section 2.The SDMA system is discussed in Section 3.Simulation and lake test results are provided in Section 4 and concluding remarks are given in Section 5.

2.Active average sound intensity detector

The AASI detector uses the excellent autocorrelation and weak cross-correlation of pseudo-random sequence to simultaneously estimate azimuth for multiple users.Its direction of arrival(DOA)estimation is shown in Fig.1, wherevxandvyare output signals of the vector sensor vibration speed channel.

Fig.1 AASI detector diagram

Given that there areNusers working simultaneously and a pseudo-random codePNi(t)is assigned to each user,wherePNi(t)is probe signal for each user,the receivedsignal of vectorsensorcan be shown[24]as follows (assuming no UWA channel):

wherePNi(t)is the target signal,np(t),nx(t)andny(t) are isotropic additive incoherent interferences,andθiis each user’s azimuth.np(t),nx(t),ny(t)andPNi(t)are independent from each other.It can be seen that sound pressureis non-directionaland it is a scalar while vibration velocity is a vector and each componenthas dipole directivity.Velocity directional characteristics have nothingto do with frequency as well as velocity sensor directivity,as a result,a small-sizedvectorsensor canmeasuretheazimuth of the sound source and it is the advantage of the vector sensor.

When estimating the azimuth of thekth user,AASI usesvxandlocalreferencedsignalPNk(t)todothecorrelation operation.After peak selection,one can get

whereAkis the correlation peak of thekth user.It can be learnt thatIx,kconsists of two parts.The firs part is a desired item,which is composed by the correlation peak and the information of a user’s azimuth,and it will be selected from the peak selector.The second term is an interference componentthat comprises an MAI componentanda noise component.As the pseudo-random sequence has weak cross correlation,multiple access interference and noise interference components are both small values.Similarly,one can get(3)by correlatingvywithPNk(t).

Divide(3)by(2)to get thekth user’s azimuth estimation:

The AASI detector can detect the matching signal by changing the local reference signal,thereby estimating the azimuth of the matching signal.

3.Spatial-division multiple access criterion

3.1 STRM used in UWA multi-user communication

The receiving end decoding terminal uses STRM technology and its principle is shown in Fig.2.The process is given as follows:

wheresi(t)is the sending signal of each user andrTRMk(t)is the STRM output of thekth user.It is composed of three parts.The firs part is the expected time reversal focusing signal.The second part is the MAI component.Spatial distribution makes different user’s UWA channels have weak cross-correlation.That is,the second part can be regardedas small values.The third part is the noise interference component.Hence,STRM can use physical properties of the UWA channel to realize SDMA UWA communication.

Fig.2 SDMA based on STRM diagram

3.2 SDMA UWA communication based on the AASI detector and STRM

TheSDMAtechnologyisshowninFig.3.Inthecommunication system,each user will be distributed a pseudo random code as the synchronization code.The synchronous codeplaysan importantrolein the communicationsystem: it is not only the basis of signal detection and synchronous but also the input of the AASI detector to realize the azimuth estimation of each user.At the same time,it also serves as the probe signal for channel estimation of each user in STRM.The channel estimation is not discussed in this paper.

Fig.3SDMA UWA system

The SDMA-CDMA system is shown in Fig.4.Received signals will firs be processedbythe SDMA system,whose output is changed by using different pseudo-random sequencesforprocessing.Hence,theSDMAsystemwilloutputN-channel signals,whereNis the number of users in the CDMA system and each signal corresponds to a user. As discussed above,the MAI will be greatly reduced and the outputs of the CDMA system will get a much betterperformancecomparedwiththetraditionalCDMA system.

Fig.4SDMA-CDMA system

4.Simulation and lake test

There must be subsequent signal processing after SDMA processing because STRM cannot eliminate the ISI of the desired user.Therefore,the SDMA and CDMA systems will be put together to analyze the performance of SDMA using the spread spectrum to overcome the residual ISI in the desired user.Here are nine UWA channels created by channel simulation software shown in Fig.5.

Fig.5 Simulation channel structure of 9 users

The 9-user UWA CDMA system with each user channel impulse response(CIR)shown in Fig.5 is simulated by MATLAB.The 5th-order balance gold codes are used for the address code and the modulation is quadrature phase shift keying(QPSK).The orientation distribution of the respective users is uniformly distributed.The 10th-order balance gold codes are chosen as probe signals.Assume user 1 is the desired user and the decoding constellation is shown in Fig.6(a)in the 5 dB band-limitedsignal-to-noise ratio(SNR).

Fig.6 Decoding constellation of desired user

The decoding constellation of CDMA is bad in Fig.6(a) leadingto a very highbit error rate(BER)of desired user 1 eventhoughthe band-limitSNR is 5 dB which is relatively high in spread spectrum communication.It shows that the MAI is the major interference in multiple access communication.Actually,it is easy to achieve low BER communication in radio communication with the same CDMA system,but in UWA communication,the UWA channel makes the MAI of CDMA more complicated leading to decline significan of the system capacity.Fig.6(b)gives a much better decoding constellation,which owes to SDMA processing.The SDMA processing reduces the impact of the UWA channel(by matching with the UWA channel) and reduces the MAI twice(weak cross-correlation of UWA channels and directional communication).The performance of CDMA,STRM-CDMA and SDMA-CDMA is shown in Fig.7.

Fig.7 Performance of three multiuser systems

An experimental investigation was done in the Lotus Lake water of Heilongjiang province,China in August, 2012.As shown in Fig.8,the experiment site is an open water area and the average water depth is 40 m.The waves were about 0.1 m,and the velocity gradient distribution map can be seen in Fig.9.The sound velocity showd negative gradient of the speed of sound distribution.The lakebed of the Lotus Lake was originally a village,and then turned to be a reservoir lake after building a dam.For this reason,the interface conditions of the lakebed were complicated.Sending and receiving nodes were located on two anchored boats,but due to large waves at that time, relative movement still existed,which lead to the Doppler Effect.The dipping depths of the transducer and the receiver were both 6 m.There was relative motion between the transducer and receiver,which was about 0.5 m/s to 1 m/s.

Fig.8 Lotus lake

Fig.9 Velocity gradient distribution map

The 7th-order gold sequence is chosen as the address code.There are seven users in the multi-user system and they are distributed in the waters of different levels of distance and depth from the sending end to the receiving end at the same time.The receiving end estimates each user’s azimuth by the AASI detector and the result is shown in Table 1.

Table 1 DOA estimation results of seven users

Fig.10(a)istheresultofthematchedfilte oftheCDMA system.Fig.10(b)shows the matched filte output results of the CDMA system combined with the AASI detector. Fig.10(c)givestheresultoftheSTRM-CDMA systemand Fig.10(d)is the result of the SDMA-CDMA system.

Fig.10 Matching output comparison chart

Fig.10(a)shows that under the influenc of channel multipath extension,there is almost no correlation peak in the CDMA system matched filte output,which means the system will generate errors.It can be found that there are correlation peaks in Fig.10(b),which illustrates that vector combinationsdo reduce some undesiredusers interference.However,the output result is multimodal because the process is just to reduce the MAI in the CDMA system and does not deal with the multipath interference of the desired user and undesired users.The STRM-CDMA system uses weak correlation function of different users’channel to reduce the system MAI and realizes the desired user’s channel equalization at the same time.Thus the result shown in Fig.10(c)is relatively ideal,which furtherillustrates the necessity of jointly consideringMAI and multipath interference in UWA communication of multi-user. The SDMA-CDMA system proposed in this paper takes theadvantageoftheabovetwosystems(AASI-CDMAand STRM-CDMA),whose outputs are the most ideal.Therefore,the system capacity is the largest.

5.Conclusions

The weak mutual correlation of different users’channel is the basis of realizing SDMA technology based on STRM. The AASI detector does correlation processing of the received signals and the local reference signal to estimate each users azimuth.As long as the pattern for each user can be distinguished(i.e.mutual correlation is weak),each usersazimuthcanbedistinguishedtheoretically.Thisisthe advantage of the AASI detector.SDMA based on STRM combined with the AASI detector is easy to achieve and it can meet the requirements of pursuing simple nodes and low-power consumption especially in CDMA UWA communication.

[1]J.A.Catipovic.Performance limitations in underwater acoustic telemetry.IEEE Journal of Oceanic Engineering,1990, 15(3):205–216.

[2]D.B.Kilfoyle,A.B.Baggeroer.The state of the art in underwater acoustic telemetry.IEEE Journal of Oceanic Engineering,2000,25(1):1–25.

[3]T.C.Yang.Properties of underwater acoustic communication channels in shallow water.The Journal of the Acoustical Society of America,2012,131(1):129–145.

[4]P.A.Walree.Propagation and scattering effects in underwater acoustic communication channels.IEEE Journal of Oceanic Engineering,2013,38(4):614–631.

[5]L.Fan,G.Li,J.Chen.Performance of MC-CDMA systems with orthogonal transmit diversity over Rayleigh fading channel.Journal of Systems Engineering and Electronics,2003, 14(3):33–39.

[6]Q.Parastoo,M.Stojanovic.Statistical characterization and computationally efficien modeling of a class of underwater acoustic communication channels.IEEE Journal of Oceanic Engineering,2013,38(4):701–717.

[7]M.Stojanovic,J.Captipovic,J.G.Proakis.Adaptive multichannel combining and equalization for underwater acoustic communications.The Journal of the Acoustical Society of America,1993,94(3):1621–1631.

[8]M.Stojanovic,J.A.Catipovic,J.G.Proaki.Phase-coherent digital communications for underwater acoustic channels.IEEE Journal ofOceanic Engineering,1994,19(1):100–111.

[9]M.Stojanovic,L.Freitag.Multichannel detection for wideband underwater acoustic CDMA communications.IEEE Journal of Oceanic Engineering,2006,31(3):685–695.

[10]Y.Luo,Z.Liu,S.Hu,et al.Self-adjusting decision feedback equalizer for variational underwater acoustic channel environments.Journal of Systems Engineering and Electronics,2014, 25(1):26–33.

[11]T.C.Yang.Differences between passive-phase conjugation and decision-feedback equalizer for underwater acoustic communications.IEEE Journal of Oceanic Engineering,2004, 29(2):472–487.

[12]T.C.Yang.Correlation-based decision-feedback equalizer for underwater acoustic communications.IEEE Journal of Oceanic Engineering,2005,30(4):865–880.

[13]Y.Zhang,J.Zhao,Y.Cai,et al.Blind adaptive MMSE equalization of underwater acoustic channels base on the linear prediction method.Journal of Marine Science and Application, 2011,10(1):113–120.

[14]H.C.Song,W.A.Kuperman,W.S.Hodgkiss.A time-reversal mirror with variable range focusing.The Journal of the Acoustical Society of America,1998,103(6):3234–3240.

[15]W.J.Higley,P.Roux,W.A.Kuperman,et al.Synthetic aper-ture time-reversal communications in shallow water experimental demonstration at sea.The Journal of the Acoustical Society of America,2005,118(4):2365–2372.

[16]A.Song,M.Badiey,A.E.Newhall,et al.Passive time reversal acoustic communications through shallow-water internal waves.IEEE Journal of Oceanic Engineering,2010,35(4): 756–765.

[17]J.A.Flynn,J.A.Ritcey,D.Rouseff,et al.Multichannel equalization by decision-directed passive phase conjugation: experimental results.IEEE Journal of Oceanic Engineering, 2004,29(3):824–836.

[18]A.Song,M.Badiey,V.K.McDonald,et al.Time reversal receivers for high data rate acoustic multiple-input-multipleoutput communication.IEEE Journal ofOceanic Engineering, 2011,36(4):525–538.

[19]J.W.Yin,Y.L.Wang,L.Wang,et al.Multiuser underwater acoustic communication using single-element virtual time reversal mirror.Chinese Science Bulletin,2009,54(8):1302–1310.

[20]H.C.Song,W.S.Hodgkiss,W.A.Kuperman,et al.Multiuser communications using passive time reversal.IEEE Journal of Oceanic Engineering,2007,32(4):915–926.

[21]S.E.Cho,H.C.Song,W.S.Hodgkiss.Successive interference cancellation for underwater acoustic communications.IEEE Journal of Oceanic Engineering,2011,36(4):490–501.

[22]A.Song,A.M.Badiey,P.Hursky.Experimental demonstration of underwater acoustic communication by vector sensors.IEEE Journal ofOceanic Engineering,2011,36(3):915–926.

[23]A.Nehorai,E.Paldi.Acoustic vector-sensor array processing.IEEE Trans.on Signal Processing,1998,46(9):2481–2491.

[24]J.Y.Hui,J.Hui.The basis of vector acoustic signal.Beijing: National Defense Industry Press,2009.(in Chinese)

[25]Y.Yu,L.Qing,X.Jiang.Pressure and velocity cross-spectrum of normal modes in low-frequency acoustic vector fielof shallow water and its application.Journal of Systems Engineering and Electronics,2015,26(2):241–249.

[26]J.Liu,Z.Liu,Q.Liu.Direction and polarization estimation for coherent sources using vector sensors.Journal of Systems Engineering and Electronics,2013,24(4):600–605.

[27]J.W.Yin,S.Yang,P.Y.Du,et al.Code divided multiple access underwater acoustic communication based on active acoustic intensity average.Acta Physica Sinica,2012,61(6): 064302-1-7.(in Chinese)

[28]H.Guo,A.Abdi.Multiuser underwater communication with space-time block codes and acoustic vector sensors.Proc.of IEEE Oceans Conference,2008:1–6.

[29]A.Abdi,H.Guo.Signal correlation modeling in acoustic vector sensor arrays.IEEE Trans.on Signal Processing,2009, 57(3):892–903.

[30]T.C.Yang,W.B.Yang.Performance analysis of directsequence spread-spectrum underwater acoustic communications with low signal-to-noise-ratio input signals.The Journal of the Acoustical Society of America,2008,123(2):842–855.

[31]X.Xu,S.Zhou,K.M.Andrey,et al.Per-survivor processing for underwater acoustic communications with direct-sequence spread spectrum.The Journal of the Acoustical Society of America,2013,133(5):2746–2754.

[32]T.C.Yang,W.B.Yang.Low probability of detection underwater acoustic communications using direct-sequence spread spectrum.The Journal of the Acoustical Society of America, 2008,124(6):3632–3647.

[33]H.C.Song,W.S.Hodgkiss,P.A.Walree.Phase-coherent communications without explicit phase tracking.The Journal of the Acoustical Society of America,2010,128(3):969–972.

Biographies

Jingwei Yinwas born in 1980.He received his M.S.degree in signal and information processing and Ph.D.degree in underwater acoustic engineering from Harbin Engineering University,China, in 2006 and 2007,respectively.He is a professor of College of Underwater Acoustic Engineering, Harbin Engineering University,China.He has issued more than 80 papers,including 40 indexed by SCI and EI.His current research interests are underwater acoustic engineering,underwater acoustic communication,arctic acoustic.

E-mail:yinjingwei@hrbeu.edu.cn

Pengyu Duwas born in 1988.He is a Ph.D.student of College of Underwater Acoustic Engineering, Harbin Engineering University,China.He received his M.S.degree in signal and information processing from Harbin Engineering University,China,in 2014.He has issued more than 13 papers,including nine indexed by SCI and EI.His current research interest are underwater acoustic communication and acoustic signal processing.

E-mail:heaven663@163.com

Guang Yangwas born in 1981.He is a Ph.D.student of College of Underwater Acoustic Engineering,Harbin Engineering University,China.His interests focus on underwater acoustic communication and underwater acoustic physics.

E-mail:edit231@163.com

Huanling Zhouwas born in 1990.She is a master student of College of Underwater Acoustic Engineering,Harbin Engineering University,China.Her current research interest is underwater acoustic communication.

E-mail:zhou huanling@163.com

10.1109/JSEE.2015.00129

Manuscript received November 13,2014.

*Corresponding author.

This work was supported by the National Natural Science Foundation of China(61471137;51179034),the Ships Pre-research Support Technology Fund(13J3.1.5),the Natural Science Foundation of Heilongjiang Province(F201109),the Innovation Talents of Science and the Technology Research Projects of Harbin(2013RFQXJ101)and the National Defense Basic Technology Research(JSJC2013604C012).