Int. J. Communications, Network and System Sciences, 2013, 6, 361376 http://dx.doi.org/10.4236/ijcns.2013.68039 Published Online August 2013 (http://www.scirp.org/journal/ijcns) Review Article: Multicarrier Communication for Underwater Acoustic Channel Hamada Esmaiel, Danchi Jiang School of Engineering, University of Tasmania, Hobart, Australia Email: hamada.esmaiel@utas.edu.au, Danchi.Jiang@utas.edu.au Received June 24, 2013; revised July 22, 2013; accepted July 26, 2013 Copyright © 2013 Hamada Esmaiel, Danchi Jiang. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. ABSTRACT In past decades, there has been a growing interest in the discussion and study of using underwater acoustic channel as the physical layer for communication systems, ranging from pointtopoint communications to underwater multicarrier modulation networks. A series of review papers were already available to provide a history of the development of the field until the end of the last decade. In this paper, we attempt to provide an overview of the key developments, both theoretical and applied, in the particular topics regarding multicarrier communication for underwater acoustic communi cation such as the channel and Doppler shift estimation, video and image transmission throw multicarrier techniques, etc. This paper also includes acoustic propagation properties in seawater and underwater acoustic channel representation. Keywords: Underwater Communication; Multicarrier Communication; Channel Coding; Orthogonal Frequency Division Multiplexing (OFDM); Filterbank Multicarrier (FBMC) 1. Introduction Underwater acoustic channels are considered to be “quite possibly nature’s most unforgiving wireless medium” [1]. The complexity of underwater acoustic channels is domi nated by the ocean environment characteristics which in clude significant delay, Doublesidespreading, Doppler spreads, frequencyselective fading, and limited bandwidth [2]. However, efficient underwater communications are critical to many types of scientific and civil missions in the ocean, such as ocean monitoring, ocean exploration, undersea rescue, and undersea disaster response. Human knowledge and understanding of the oceans, rests on our ability to collect information from remote undersea loca tions. Together with sensor technology and vehicular technology, wireless underwater communications are de sirable to enable new applications ranging from envi ronmental monitoring to gathering of oceanographic data, marine archaeology, and search and rescue missions. New technologies of high speed communication for image and video transmission are also desirable to facili tate the next generation of efficient undersea expeditions. However, current acoustic communication technologies can only provide limited data rates due to the particular physical features of channel [3]. The corresponding wire less technology for undersea communications still needs significant further development. Research has been ac tive for over a decade on designing the methods for wire less information transmission underwater. Due to electro magnetic waves in underwater channel propagate only over extremely short distances acoustic wave used. In contrast, acoustic waves can propagate over much longer distances. However, an underwater acoustic channel pre sents a communication system designer with many diffi culties. Figure 1 [4] illustrates the scenario of shallow water multipath propagation. In such situation, in addition to the direct path, the signal also propagates via reflections from the surface and bottom, resulting in a multipath effect with much larger time dispersion that of wireless propagation in air. Figure 1. Shallow water multipath propagation: in addition to the direct path, the signal propagates via reflections from the surface and bottom. C opyright © 2013 SciRes. IJCNS
H. ESMAIEL, D. C. JIANG 362 There are many issues that have to be carefully exam ined when designing an acoustic based transmission sys tem for underwater channels. Some of these are [5]: 1) Attenuation due to the absorption of the acoustic waves in water, which limits the distance the sound, can travel; 2) Low propagation speed of the sound, roughly around 1500 m/s; 3) Multipath due to the reflection from the bottom and surface of sea, causing echoes and interfer ence; 4) The transmitted signal suffers from the hetero geneous characteristics of the underwater channel as well as Doppler’s effect caused by the movement of transmi ter and receiver; 5) Noise in the ocean. The noise level can mask the portion of the signal and block the corre sponding carried data. The properties of the underwater medium are also ex tremely varied, and change in both space and time. Fluc tuations due to environmental characteristics include sea sonal changes, geographical variations both in tempera ture and salinity, seabed relief, currents, tides, internal waves, movement of the acoustic systems and their tar gets, etc. All this makes the underwater acoustic signal fluctuating randomly. As such, the selection of modula tion and error correction techniques is very challenging. To mitigate the bandwidth limitation, multicarrier modu lation is used in the underwater acoustic channel. It is an alternative to overcome the longtime delays in under water acoustic channels. It increases the symbol interval and thereby decreases the intersymbol interference (ISI) [6]. In this paper, we aim to provide a brief overview of the key developments, both theoretical and applied, in the field in the subsequent period. We also hope to pro vide an insight into some of the open problems and chal lenges facing researchers in this field in the near future. Rather than attempting to provide an exhaustive survey of all research in the field, we will concentrate on multi carrier modulation for underwater acoustic channel simu lation, challenges, ideas and developments that are likely to be the keystone of future digital signal processing for underwater acoustic communication systems. This paper summarizes several aspects of underwater acoustic communication. It is organized into the follow ing sections. Section 2 briefly summarizes acoustic propa gation properties in seawater. Section 3 analyses channel coding performance for multicarrier modulation in un derwater acoustic communication systems. In Section 4, multicarrier modulations for underwater acoustic com munications are presented. The underwater acoustic chan nel estimation for underwater communication is included in Section 5. In Section 6, Doppler shift estimation for underwater acoustic communications is studied. Image & video transmission over the underwater acoustic channel (UWAC) with multicarrier modulation is discussed in Section 7. A summary of this paper is included in Section 8. 2. Acoustic Propagation Properties in Seawater A good understanding and reasonably accurate modelling of the underwater acoustic channel is the required basis upon which all other works for underwater networks can be carried out. Several models are already available for calculating and predicting the attenuation [7,8], which can also help to model other aspects of the underwater acoustic channel. Furthermore, parameters of frequency, distance, depth, acidity to salinity, and temperature of the underwater environment can be used to characterize how the channel acts and how network can possibly perform. This section summarizes underwater channel models for that purpose. 2.1. Propagation Loss The transmitted acoustic signal in underwater acoustic communication reduces strength with increasing distance due to many factors such as absorption caused by mag nesium sulphate and boric acid, particle motion and geo metrical spreading, etc. Propagation loss is composed mainly of three aspects, namely, geometrical spreading, attenuation and the anomaly of propagation. The latter is nearly impossible to model. However it is known that the signal attenuation, in dB, that occurs over a transmission distance for a signal frequency l can be approxi mated as [9]: 10 log1,10 log10 log,Af kll (1) where is the absorption coefficient in dB/km, which can be obtained from the particular models characterizing it, and k represents the geometrical spreading factor with its value between 1  2. 2.2. Absorption Coefficient Attenuation by absorption occurs due to the conversion of acoustic energy into heat in seawater. This process is frequency dependent since at higher frequencies more energy is absorbed. The attenuation by absorption mod els considered for inclusion into the Thorp model [9]. Equation (2) provides the absorption coefficient in dB/km as a function in carrier frequency c : 22 42 22 0.1 40 10log2.75 100.003. 1 4100 cc c cc ff f ff (2) 2.3. Ambient Noise Model Ambient noise in the ocean can be described as Gaussian and having a continuous power spectral density (p.s.d.) [9]. The four most prominent sources for ambient noise are the turbulence, shipping, wind driven waves and thermal noise. Their p.s.d. in dB re µPa per Hz are given Copyright © 2013 SciRes. IJCNS
H. ESMAIEL, D. C. JIANG 363 by the formulae Equations (3)(6) shown below, respec tively [9]: 10 log1730 log, t Nf f (3) 10log40 200.526log 60 log0.03, s Nf sf f (4) 10log507.520log40log0.4, w Nff f (5) 10log1520log . th Nf f (6) The ambient noise in the ocean is affected by different factors in specific frequency ranges. In the noise models given in Equations (3) to (6), the effect of colored noise denoted by represents the turbulence noise at frequency the shipping noise (with as the shipping factor which lies between 0 and 1), t Nf ,f s N fS f w N the wind driven wave noise (with as the wind speed in m/s), and the thermal noise. The composite noise p.s.d. can obtained in µPa from [ ω th N f 9]: . tswth NfNf Nf NNf (7) 2.4. SignaltoNoise Ratio In UWAC, signaltonoise ratio can be calculated [10] based on signal attenuation and the noise p.s.d., Specifi cally, the SNR observed at the receiver can be calculated in µPa re dB per Hz using the following equation: SNR ,, , P lf lf N ff (8) where is the SNR over a distance and a transmission center frequency f, P is the signal transmis sion power and SNR ,lf l represents the receiver noise band width. Equation (8) clearly shows that the underwater acoustic channel SNR is a function of transmission fre quency. As such we can find the optimal frequency for UWAC to maximize SNR. The attenuationnoise (AN) factor, given by , lf can be used to reflect the fre quency dependent part of the SNR. By close analysis of this relationship, it can also be used to determine the op timal frequency at which the maximal narrowband SNR is achieved for each transmission distance l. Since the SNR is inversely proportional to the attenua tionnoise factor, the optimal frequency is that for which the value of 1/AN (represented in dB re µPa per Hz) is the highest over the combination of a certain distance, o l. In Figure 2, [10] frequencydependent part of the narrowband SNR, 11, , fNf is shown. In this plot the attenuation and noise parameters are selected as and 1.5,k0.5,S0, to reflect the UWAC with moderate shipping activity and no wave noise. It can be seen that there is a frequency for which the Figure 2. Narrow band SNR, 1/A(1, f)N(f) ; k = 1.5; s = 0; ω = 0. narrowband SNR is maximized for a particular distance, for the given attenuation and noise constants. This optimal frequency, let it denoted , o l can be selected as the carrier frequency c for that particular transmission distance [10]. 2.5. Channel Capacity As per the Shannon theorem the channel capacity i.e. the theoretical upper bound on data that can be communi cated through a undistorted channel subject to additive white Gaussian noise is given by the following formula [9]: 2 log 1, S CB N (9) where is the channel bandwidth in Hz and BSN represents the channel SNR. The basic Shannon rela tionship shown in Equation (9) can be extended to be applicable in cases where the noise is dependent on fre quency to take the form of [9]: 2 log1d. B Sf Cf Nf (10) For timeinvariant channel, if a certain time interval with Gaussian noise, we can obtain the total capacity by dividing the total bandwidth into multiple narrow sub bands and summing their individual capacities collec tively. Then each subband has a small width which is centered on the transmission frequency, i.e. the band width. In the case where the transmission bandwidth, ,Bl over a distance is known along with the trans mission power l Pl , we can extend Equation (10) to obtain the channel capacity over a distance [ l9]: 2 logd . , B Pl Cf Al fNfBl (11) The choice of the underlying absorption coefficient model imposes a dependence of the capacity on depth, temperature, salinity and acidity as well. Copyright © 2013 SciRes. IJCNS
H. ESMAIEL, D. C. JIANG 364 2.6. Doppler Shift and UWAC Multipath In underwater communication, the relative movement be tween the transmitter and the receiver due to the constant motion of nodes results in Doppler shifts, which signifi cantly distort received signals. It is required to estimate the Doppler shift and compensate it for all UWAC ap plications. Different from the case of terrestrial commu nication where the Doppler effect is modeled by a fre quency shift, due to the slow sound speed in water, the effect of transceiver motion on the duration of the sym bol cannot be neglected [11]. Doppler phase , d de pending on the relative velocity and the ratio between the carrier frequencyc v and the symbol rate 1RT [12] caused in the received signal as: 2π2π 1 c dc fv T. Rc v (12) From [13,14], UWAC multipath representation for mul tipath arrival p is characterized by its mean magnitude gains and delay t. , These quantities are dependent on the path length l which in turn is a function of the given range R. The path magnitude gain is given by , pp pc , lf where 12 r p is the amount of loss due to reflection at the bottom and surface, and r is the number of reflections for path From Equation (1), the acoustic propagation loss, represented by resulting in the following equation: p. ,, pc lf ,. l k pc pc Al flf (13) The delay for path given by ,ppp tlc (c = 1500 m/s is the speed of sound in water) and l is the path length for path p [14]. for each path re spectively, the path lengths can be calculated using pla nar geometry. 0,1,3,5, 7 p r 2.7. Underwater Acoustic Noise Model An empirical model for the noise of the acoustic under water channel in shallow water from the analysis of field data measurements has been presented in [15]. In that paper, a probability density function for the noise ampli tude distribution is proposed and the associated likeli hood functions are derived. As a result, an expression to the probability of symbol error for binary signaling is presented for the channel. In Addition, the results of simulations conducted using the field collected noise samples are presented, in order to verify the noise effect on the performance of underwater acoustic communica tion binary signaling systems. The analysis of field data measurements has shown that the noise amplitude distri bution presents good fitting with the Student’sdistribu tion. From [ t 15], the symbol error probability of the bi nary UWAN channel can be expressed as: 2.75 2 0 24 15 ebobo d. kENENxx (14) where is a vector of discrete amplitude levels of noise, integer 1, 2,,,kM and bo EN is the energy per bit to noise power spectral density ratio. 2.8. Underwater Acoustic Channel Simulator The signal transmission in underwater acoustic commu nications can be modeled as a timevarying channel. In particular, the noisefree signal at a receiver is described as a convolution [16]: ,d,0,ytht sttTs , (15) between the channel impulse response ,ht and the source signal t, where Ts is the signal duration. This description of linear timevariant systems is quite generic. In underwater acoustic communications, other variants of the description may also be used. For example, for a channel with discrete multipath components, the model [1719] used, and UWA channel represented as: ,. pp p ct Att (16) Within a data block of interest, each path delay can be associated with one Doppler scale factor as: , pp ta t (17) and the path amplitudes are assumed constant within one data block . p tA Furthermore we assume that the UWA channel can be well approximated by N domi nant discrete paths. Hence, the channel model can be simplified to be: 1 ,. p N ppp p ct Aat (18) 3. Channel Coding Performance for Multicarrier Modulation in Underwater Acoustic Communication It is required that digital communication systems, par ticularly for underwater use, to perform accurately and reliably in the presence of noise and interference. Among many possible ways to achieve this goal, forward error correction coding is the most effective and economical. The fast temporal variations, long multipath delay spreads, and severe frequencydependent attenuations of under water acoustic communication channels are extremely complex that impedes underwater acoustic data transmis sion. To alleviate this problem, channel coding is indis pensable in UWA communication system to increase the Copyright © 2013 SciRes. IJCNS
H. ESMAIEL, D. C. JIANG 365 reliability [20]. Also for underwater communication com monly used symbol demodulation schemes do not de pend on the noise power, so bit error performance with out error correcting coding will not be improved [21]. Here, Forward Error Correction (FEC) is a type of error correction which improves simple error detection schemes by enabling the receiver to correct errors once they are detected. This reduces the need for retransmissions and energy consumption [22]. The power of error correcting codes increases with the channel coding length constraint and approaching the Shannon limit with a large number of length constraints. But in return the complexity of the decoder also increases with length constraint [23]. For these reasons, it is desir able to construct long codes and minimize the complex ity of the decoder. The concatenation of codes is a cheap solution for that [24] for multicarrier modulation in un derwater acoustic communication. The major drawback of the concatenation code is that the decoder is unable to decode correctly in the presence of a burst of erroneous bits. Hence, an interleaver can be designed to introduce a dependency between the bits input. In order to minimize the error rate two evolved coding scheme [22], which are summarized in the following two subsections. 3.1. Reed Solomon Coder Reed Solomon codes ,,nkt are cyclic codes, built from symbols with a maximum of where is the number of elements in the Galois Field n1,nq q q GF and t is the power correcting code. So the number of control symbols is 2n q 2.t It has been approached in [25] to develop a sufficiently robust acoustic link allowing the transmission of differ ent information using Reed Solomon coder and conven tional coding to protect data transmission over underwa ter acoustic channel. Where underwater acoustic link is designed to transmit different kinds of data as text, im ages and speech signal, blind spatialtemporal equalizer is used to reduce different underwater acoustic perturba tions. To improve the underwater acoustic link perform ance and obtain a higher code rate, Reed Solomon Block Turbo Codes (RS BTC) has been introduced and tested in real conditions, with the aim to decrease the BER. A dif ferential coding has been used to solve the phase ambi guities. Channels block coders algorithm applies to trans mission of constant data. Where Block codes are FEC codes that enable a limited number of errors to be de tected and corrected without retransmission. Block codes has been used to improve the performance of a commu nications system when other means of improvement (such increasing transmitter power or using a stronger modulator) are impractical [26]. 3.2. Low Density Parity Check Code (LDPC) LDPC codes are a special type of linear block coder. The paritycheck matrix H of LDPC codes are very sparse, i.e. they can be specified by a matrix containing mostly 0’s [27]. It is used to reduce error codes and achieve credible transmit performance of underwater digital signal. Turbo code has also been recommended to apply to underwater digital speech communication system and simulation re sults in underwater digital speech system discussed in [27] have shown that LDPC has a better performance than the turbo coder. Modification in the LDPC has been proposed to more matched multicarrier underwater communication systems [28], by focusing on the case of matching the coding symbols with the modulation symbols. Experimental re sults show, with real data that whenever the uncoded BER is below 0.1, normally no decoding errors will oc cur for the rate 1/2 of nonbinary LDPC codes used in [28] and it consistent with the simulation results. The unen coded BER can serve as a quick performance indicator to assess how likely the decoding will succeed. The re searcher results show that LDPC code system has a better error correct performance and can achieve a better BER under the relative lower SNR [20]. Performances of LDPC codes with different parame ters over different underwater acoustic communication channels are studied [20], by adjusting the encoding and decoding parameters according to different underwater acoustic channels. When LDPC coder used as channel coder for zero padding orthogonal frequency division multiplexing (ZP OFDM) multicarrier modulation the spectral efficiency and the data rate are [29,30]: 2 log bitssHz, D g S Tr TT KM (19) kb s.RB (20) where is the ZPOFDM symbol duration, T T is the guard interval, is number of all subcarrier, S is data subcarrier in total, is the rate of nonbinary LDPC code [ r 28], is the quadrature amplitude modulation symbol, and is the channel bandwidth. LDPC pro duces high blockerrorrate (BLER) performance in ZP OFDM multicarrier system. B 4. MultiCarrier Modulation for Underwater Acoustic Communications Multicarrier modulation Systems are well known to be attractive for communications through multipath com munications channels. The traditional approach expects the symbol duration of the transmitted signal to be larger than channel delay spread. This results in a low rate Copyright © 2013 SciRes. IJCNS
H. ESMAIEL, D. C. JIANG 366 “sampling” of the channel impulse response (once per symbol in each subchannel) and a subchannel band width that is less than the coherence bandwidth of the physical propagation channel [31]. MultiCarrier Modu lation (MCM) have become popular in UWA channel for two reasons. First, a signal can be processed in a receiver without the increase of noise or interference caused by linear equalization of a single carrier signal; second, the long symbol period used in MCM ensures greater immu nity to impulse noise and fast fades [32]. 4.1. Orthogonal Frequency Division Multiplexing Multicarrier modulation in the form of orthogonal fre quency division multiplexing (OFDM) has been studied and implemented for broadband wired and wireless com munications for the past two decades. OFDM is widely adopted because of a number of its advantages as [3335]: 1) Orthogonality of subcarrier signals that allows: a) Easy generation of transmit signal through an inverse fast Fourier transform (IFFT) block; b) Easy separation of the transmitted data symbols at the receiver through a fast block; c) Easy equalization through a scalar gain per subcarrier; d) Easy adoption to multipleinput multiple output (MIMO) channels; 2) Closely spaced orthogonal subcarriers partition the available bandwidth into a maxi mum collection of narrow subbands; 3) Adaptive modu lation schemes can be applied to subcarrier bands to maximize bandwidth efficiency/transmission rate; 4) The very special structure of OFDM symbols simplifies the tasks of carrier and symbol synchronizations. OFDM as an MCM is particularly efficient when noise is spread over a large portion of the available bandwidth. It transmits signals over multiple orthogonal subcarriers simultaneously and performs robustly in severe multi path environments achieving high spectral efficiency. OFDM used in underwater communications as a superior alternative to single carrier broadband modulation to achieve high data rate transmission [4,17,36,37]. It has been proved to be an effective technique for combating the multipath delay spread without the need for complex timedomain equalizers, due to its robustness against fre quency selective fading and narrowband interference. In fact, if the number of subcarriers is large enough, each subcarrier only deals with flat fading rather than with frequency selective fading as a wideband carrier does. The narrowband interference will affect only one or two subcarriers of the whole bunch of subcarriers. The major issue in applying OFDM to underwater channels is the motion induced Doppler distortion which creates non uniform frequency offset in a wideband acoustic signal [4,17,36,37]. 4.1.1. ZPOFDM for Underwater Acoustic Communication Zeropadded (ZP) [29,30,38] orthogonal frequency divi sion multiplexing (OFDM) has been extensively investi gated for high data rate underwater acoustic communica tion. Frequencydomain oversampling method [30] is used to avoid information loss incurred by the overlapadd operation. A larger FFT size used to improve system performance over underwater acoustic channels with sig nificant Doppler spread. The system is validated using real data collected from field experiments. Zero padding is used instead of the conventional cyclic prefix in order to save transmission power during the (long) guard in terval. ZPOFDM transmitted and received signals in the time domain during underwater channel paths signals after implemented to make zeropad guard interval accept chan nel path delay shown in Figure 3 [30]. For ZPOFDM With symbol duration the subcar rier spacing is ,T 1, T and the subcarriers are located at frequencies ,,, 22 kc kK ff k T , 1 K (21) where c is the center frequency, and is the total number of subcarriers, leading to the bandwidth B = K/T. k denote the information symbol on the subcar rier. The transmitted passband signal is [ th k 30]: 2 2Ree,0 , k A jft kS tskgtt ,T (22) where T is the ZPOFDM block duration as shown in Figure 3. The baseband signal tz is obtained with the pass band to baseband downshifting and the lowpass filtering, leading to [30]: 2 1 21 e e1 p cp p pp A N jfbt p p jmk bt pp kS zt A skgb t , (23) Figure 3. Illustration of the transmitted and received signals in the time domain. (a) One transmitted ZPOFDM block; (b) One received ZPOFDM block. Copyright © 2013 SciRes. IJCNS
H. ESMAIEL, D. C. JIANG 367 where b represents the residual Doppler rate satisfy ing: 1 1. ˆ 1 p a ba (24) ZPOFDM saves Power transmitted and increases channel capacity. In [39] authors derived bounds to the channel capacity of OFDM systems over the underwater (UW) acoustic fading channel as a function of the dis tance between the transmitter and the receiver. 4.1.2. Pilot Sig nal Desi g n for OFDM Mul ti carri er Transmission The challenges in pilot design for multicarrier transmis sion over underwater acoustic timevarying channels are twofold [40,41]: 1) Sets of adjacent observations are needed to estimate the intercarrier interference (ICI) coefficients; 2) Keeping pilot and data symbols orthogo nal at the receiver is challenging due to the ICI. Gener ally there are three types of pilot insertion methods for OFDM system, combtype; blocktype; hexagonal grid type, etc. [42]. Random selection of pilot subcarriers is motivated by the compressive sensing [43] with the sys tematic use of pilot blocks in regular intervals as seen in nonsparse channel estimation of timevarying channels [29]. In [40], authors are interested in how to address the ICI between data and pilot subcarriers without guard zeros. As it decreases spectral efficiency, focus has been put on whether data symbol carrying subcarriers should be used as observations in channel estimation, which also contain ICI originating from the pilot symbols. The per formance under varying amounts of pilot overhead has also been studied. Specifically, authors are looking for an optimum tradeoff between using more pilots or a more robust modulation scheme, to achieve the highest spectral efficiency. 4.1.3. OFDM Based on Discrete Cosine Transfo rm for Underwater Acoustic Communication The orthogonal feature of conventional OFDM can also be achieved by inversing DCT (IDCT)DCT structure for underwater communication, which reduces implementa tion area and increases computational speed, as only real calculations, is required. This system provides higher peak toaverage power ratio (PAPR) reduction and achieves better noise immunity and hence a better bit error rate (BER) performance than standard OFDM, while main taining a low implementation cost [32,44]. DCT based OFDM is a better technology for under water acoustic communication, because the bandwidth required for DCT is half of that required for DFT when both systems have same number of subcarriers which will be matched with underwater channel limited band width. It has also been shown that the speed of calcul tion of orthogonal components is increased three folds while the implementation size reduces to half as com pared to fast FFT based design [44]. Furthermore, it is known that the DCT basis have excellent spectral com paction and energy concentration properties which in turn lead to improved performance with suitable channel estimation [44]. As DCT is widely adopted in image/ video coding standards, by using it for modulation/de modulation on frequency selective channels it will result in a better integrated system design and a reduced overall implementation cost [44]. 4.1.4. Orthogonal SignalDivision Multiplexing for Underwater Acoustic Communication Orthogonal signaldivision multiplexing (OSDM) is pro posed as a UWA communication system scheme that measures the multipath profile without an adaptation or interpolation process, to achieve stable communication in doubly spread channels [45]. The performance compare son of the OSDM scheme and existing schemes in dou bly spread channels has been done. The illconditioned problem exists for conventional OSDM, which employs a single transducer in the receiver. The introduction of a multichannel receiver has been found to be effective against the illconditioned problem. Evaluation of OSDM communication done by comparing it to existing schemes with singlecarrier Recursive Least Square Differential Feedback Equalizer (RLSDFE) [46] and OFDM. OSDM with a multichannel receiver is attractive in terms of communication quality. It achieves much better BER per formance comparing to the other schemes in both static and dynamic channels, although its complexity is less than that of RLSDFE. OSDM can become a viable al ternative offering a highly reliable communication envi ronment for UWA communication with multipath and Doppler spread (tested only for shallow water) with prac tical complexity [45] and it can be a very important of point for future research. 4.1.5. Time Domain Synchronous Orthogonal Frequenc y D ivisio n Multip lexing A time domain synchronous orthogonal frequency divi sion multiplexing (TDSOFDM) scheme with dual Pseudo noise (PN) sequence [47] is proposed in [48] for UWA communication. TDSOFDM system is used for under water acoustic channel shown in Figure 4 [48]. Instead of using cyclic prefix (CP) or zero padding (ZP) as the guard interval, the proposed TDSOFDM scheme uses two identical PN sequences as guard interval, and utilizes them for frame synchronization and channel es timation. TDSOFDM increases the spectrum efficiency over the conventional CP or ZP OFDM systems where additional pilots have to be inserted for channel estima Copyright © 2013 SciRes. IJCNS
H. ESMAIEL, D. C. JIANG 368 Figure 4. TDSOFDM system for underwater acoustic channel. tion. With the dual PN at each frame header, compressed sensing channel estimation is adopted and a rather simple equalization design to reduce the receiver complexity. Conventional performance of ZPOFDM receiver is severely limited by the ICI due to the fast channel varia tions within each OFDM symbol. Furthermore, the UWA channel is wideband in nature due to the small ratio of the carrier frequency to the signal bandwidth [30,49]. In [50,51] cyclic shift keying spread spectrum OFDM method propose as a UWA communication system. In [50,51] cyclic shift keying spread spectrum OFDM method is proposed to use in a UWA communication system. The aim is to solving the problem of low data rate of direct sequence spread spectrum underwater acou stic communication [52] and that of the complexity of the receivers of Mary spread spectrum. The paper also aimed to improve bandwidth efficiency and bit rates. This method has high data rate comparing to conven tional direct sequence spread spectrum for the underwa ter acoustic communication and low bit error rate. 4.1.6. P o wer and Bit Loading for Underw a ter Acoustic OFDM System Adaptive bit and power loading is a constraint optimiza tion problem with generally two cases of practical inter est, where the objectives are the achievable data rate maximization (RM) and system margin maximization (MM) [53]. In [41,54,55] researchers propose a different optimiza tion model for underwater acoustic (UWA) channels, which is achieved by two algorithms: one is the band widthefficient bit loading algorithm; the other is the Lloyd algorithm based limited feedback procedure. It aims at minimizing the power consumption under con straints of the constant symbol data rate and desired bit errorrate (BER). Algorithms are employed to quantize the CSI at the receiver and construct the codebook, which is adopted to achieve the limited feedback process. After selecting an initial bit loading vector upon the current CSI, the receiver will broadcast its index to the transmit ter, then the transmitter will compute the bandwidth efficient bit loading algorithm and allocate the corre sponding power and bits to each OFDM subcarrier. Also algorithms are used for UWA cooperative communica tion system, which involves the DecodeandForward (DF) transmission protocol [41,56]. The ambient noise power i for each OFDM subcarrier can be proposed as [41,56]: 09 10 5 10 , iH iH iL iL ff N iii i ff i Nfdff df (25) where iL and iH are the bound of ith subcarrier frequency. The maximum capacity of a UWA SISO OFDM system with total system transmits power con straint describe as [41,56]: 2 1 1 maxlog 1 s.t.,0,1, , i N ii i N isi i g Cf iN , (26) N is the number of subchannels, is the sub channel transmitted bandwidth, i is the transmitted power in the subchannel, ith the total transmitted power, i is the subchannel power gain and the power loading will be based on this parameter, and 22 ii gH , i where i is the fading amplitude of the subchannel, ith 2 i is the subchannel power, 2 i is the ambient noise power and it’s a constant of subchannel, and ii ith is the or carriertonoise ratio SNR CNR per subchannel. is the “SNR gap” for characterizing the difference between the needed to achieve a certain data rate for a practical system and the theoretical limit. SNR 4.1.7. OFDM Receiver Designs for Underw ater Acoustic Communication OFDM receiver designs for underwater acoustic channels with user and path specific Doppler scaling distortions were proposed in [57]. The method was motivated by the cooperative communications framework [58], where dis tributed transmitter/receiver pairs may experience sig nificantly different Doppler distortions, as well as by the single user scenarios, where distinct Doppler scaling fac tors may exist among different propagation paths. The conventional approach of front end resampling that cor rects common Doppler scaling may not be appropriate in such scenarios, rendering a post FFT signal that is con taminated by user and pathspecific intercarrier inter ference. To counteract this problem, authors propose a family of front end receiver structures that utilize multi ple resampling (MR) [59] branches, with each matched to the Doppler scaling factor of a particular user and path. Following resampling, FFT modules transform the Dop pler shift compensated signals into the frequency domain for further processing through linear or nonlinear detec tion schemes. As part of the overall receiver structure, a Copyright © 2013 SciRes. IJCNS
H. ESMAIEL, D. C. JIANG 369 gradient descent approach is also proposed to refine the channel estimates obtained by standard sparse channel estimators. 4.2. Filterbank Multicarrier for Underwater Communications In order to combat time dispersion of UWA channels, it has been proposed to deploy OFDM technique with a sufficiently long CP. Moreover, to keep the bandwidth efficiency of the transmission high, long OFDM symbols that are at least four times of the length of the CP should be used. Due to small sideband power leakage, filter bank multicarrier techniques are considered as interesting al ternatives to traditional OFDMs for spectrum pooling Cognitive Radio [60]. This leads to an OFDM system in which channel variation over each OFDM symbol may be unacceptably large, thus, results in a significant level of ICI [61]. For that filter bank multicarrier (FBMC) systems can be optimized for robust performance in the doubly dispersive UWA channels [61]. OFDM multicar rier losses bandwidth efficiency of the transmission due to the allocation of 20% of each OFDM symbol to its CP. This is equivalent of saying the CP length is one quarter of the length of each fast FFT block in the OFDM system. Moreover, since the length of CP should be at least equal to the duration of the channel impulse response, and the latter is usually very long in UWA channels, very long symbols is used in the OFDM systems for UWA com munications [61]. On the other hand, to avoid ISI, the prototype filter p(t) designed as Nyquist filter [62,63]. The design method proposed in [64] constructs an iso tropic filter according to the equation. 4 0 , L kk k ptah t (27) where are the set of Hermite functions defined as [ k ht 61]: 2 2 2 1d ee d 2 n t nnn ht t 2 . t (28) In [61], it is noted that the presence of channel will result in a disturbed ambiguity function, ,, d p v in which the null points of , p v are smeared out. Thus, it is argued that to design a robust prototype filter, the constraints on the nulls of the ambiguity function , p v may be relaxed. Each null point is replaced by a region in the plane, wich is termed a null region. It is thus proposed to design ,v ,pt one should choose to mini mize the cost function: 0 2 0 1 ,1dd ,dd k N pkp k AA 2 . vv Avv (29) where k are sets of positive weighting factors, 0 is the region around ,0,v 0 over which the peak of , p v remains approximately equal to one, and k for 1, 2,,,kN are sets of null regions. UWA communication method using a class of FBMC systems was proposed [61]. This class of FBMC systems was designed to be robust against dispersions in time and frequency domain. When the Filterbank multicarrier tech nique is compared with OFDM, it clearly shows that there is a wide gap between the performance of FBMC and OFDM for underwater communication in saving band width [61]. For the Singleuser communications case OFDM offers a lower complexity. FBMC offers higher bandwidth efficiency. For Multiinput multioutput (MIMO) communications case, OFDM provides full flexibility. FBMC can be used in certain MIMO setups. Only FMT can offer the same flexibility as OFDM. But, FMT suf fers from the same bandwidth loss as OFDM [65]. FBMC for underwater communication still have high poor re search paper and need more research interest. 4.3. MIMO for Underwater Acoustic Channel Multiinput multioutput (MIMO) techniques have been extensively discussed in underwater acoustic communi cations to overcome the bandwidth limitation of undersea channel [66]. Combined with OFDM modulation, MIMO techniques provide substantial spectral efficiency and rea sonable robustness against frequency fading while keep ing simple equalizer structure [6770]. Long acoustic mul tipath, however, limits the applicability of MIMO chan nel estimation methods that require inversion of a matrix whose size is proportional to both the number of transmit elements and the multipath spread. To overcome this problem, an adaptive algorithm is used [71] that does not require matrix inversion and operates in a decisiondi rected manner, thus reducing both the computational complexity and the overhead. Reduction in complexity has been sought through selection of significant impulse response coefficients which results in a reducedsize ma trix inversion [7274]. MIMOOFDM design consists of the following key components: 1) Null subcarriers are inserted at the transmitter to facilitate the compensation of Doppler shifts at the receiver; 2) Pilot tones are used for MIMO channel estimation; and 3) An iterative re ceiver structure is adopted that couples MIMO detection with channel decoding [66]. OFDM has a number of desirable features, including low complexity of implementation and mature technolo gies that keep it as the dominant technology for sin gleuser (pointtopoint) underwater communications. Moreover, while OFDM can be easily adopted for MIMO channels, development of MIMOFBMC systems/net works is still nontrivial and may be very limited. Only FMT, the less bandwidthefficient member of the class of FBMC systems, can offer a similar level of flexibility as Copyright © 2013 SciRes. IJCNS
H. ESMAIEL, D. C. JIANG 370 OFDM in MIMO channels. Therefore, the poor fre quency spectra of subcarrier signals in OFDM are the main issue that limits the applicability of OFDM in some present and future development of broadband underwater communication systems. FBMC [75], on the other hand, is an elegant method scope with that by taking a filtering approach to underwater multicarrier communications. 5. Underwater Acoustic Channel Estimation To reduce computation complexity of signal processing and improve the accuracy of symbol detection, receiver structures that are matched to the physicalfeedback equalizer is designed first in [76], which rely on an adap tive channel estimator for its parameters computation. The channel estimation complexity is reduced in size by selecting only the significant components, whose delay span is often much shorter than the multipath spread of the channel. This estimation is used to cancel the post cursor ISI prior to the linear equalizer involved. Optimal coefficient selection is performed by truncation in mag nitude. The advantages of this approach are the number reduction in receiver parameters, optimal implementation of sparse feedback, and efficient parallel implementation of adaptive algorithms for the multichannel precombiner, fractionally spaced channel estimators and the short feed forward equalizer filters [77]. Coherent modulation schemes such as phase shift key ing (PSK) along with adaptive decision feedback equal izers (DFE) for spatial diversity used as an effective way of communication in such channels [78,79]. However, the long delay spread and rapid time variation of the channel often makes this approach computationally too complex for realtime implementations [80]. Although the underwater channel has a long impulse response, the multipath arrivals are often separated. This introduces the possibility of using a sparse equalizer with tap placement based on the actual channel response. This can poten tially dramatically reduce the number of required taps and hence leads to a lower complexity, faster channel tracking and an enhanced performance [81]. In [81], the authors have proposed an algorithm to track the channel explicitly and determine the tap placement for the DFE based on this channel estimate. The equa lizer and the channel estimator are updated individually throughout the packet. The channel estimator can also update either the whole estimation or a set of selected channel coefficients at one time in batch, depending on computational and channel considerations [82]. The channel estimation algorithms can cope with spa tial diversity by multichannel combining before equali zation. Adaptive estimation is performed using minimum mean square error as the overall optimization criterion. The receiver is implemented in a multichannel configu ration, which provides the array processing gain neces sary for many of the underwater acoustic channels. The complexity of the detection algorithm is linear in the number of receive elements and in dependent on the modulation level of the transmitted signals [83]. DFE structures may suffer from error propagation due to the feedback of possible erroneous decisions in the loop. Hence powerful forward error correction (FEC) codes are needed to ensure low bit error rate (BER) communi cation. Turbo codes are a class of powerful codes that utilize iterative information exchange between two de coders to correct errors. Inspired by this idea, researchers have developed turbo equalization techniques where it erative interactions between the equalizer and a decoder result in joint estimation, equalization and decoding [84]. A new channel estimation [85] equalization and phase correction scheme has been developed. The new scheme estimates the acoustic fading channel without separating the phase drift and phase rotation for each symbol and then the SIMO receive signals are equalized and com bined. Finally the phase drift/rotation of symbols is cor rected per group of symbols using estimated average phase drift/rotation. Authors in [29] considered sparse channel estimation using subspace methods and compressed sensing on chan nels subject to moderate Doppler effects, and extended the compressed sensing receivers to handle channels with different Doppler scales on different paths. Channel es timation and efficient symbol detection studied in [86] where the design of stateoftheart training sequences and sparse learning via iterative minimization (SLIM) algorithm was proposed to achieve sparse channel esti mation. The authors have developed a conjugate gradient (CG) based detector, which exploits the diagonalization properties of the circulant channel matrix to significantly improve the performance of multiinput multioutput (MIMO) underwater acoustic communications (UAC). The UWA communication system under consideration employs orthogonal frequency division multiplexing (OFDM) and receiver preprocessing to compensate for the Doppler effects before channel estimation. First ex tend the original homotopic algorithm from realvalued signals to the complex valued ones. Then propose two enhancements to the sparse recoverybased UWA chan nel estimator by exploiting the UWA channel temporal correlations, including the use of a firstorder Gauss Markov model and the recursive leastsquares algorithm for channel tracking. Moreover the authors propose a scheme to optimize the pilot placement over the OFDM subcarriers based on the discrete stochastic approxima tion. Sparse channel estimation for multicarrier underwater acoustic communication were proposed in [29]. Based on the pathbased channel model, subspace methods are Copyright © 2013 SciRes. IJCNS
H. ESMAIEL, D. C. JIANG 371 wellknown techniques from the arrayprocessing litera ture for the channel estimation problem. Also recent com pressed sensing technique is employed to develop some new methods, namely Orthogonal Matching Pursuit (OMP) and Basis Pursuit (BP) [87,88]. Based on the continuous time characterization of the path delays, finer delay has been suggested to be used of resolution over complete dictionaries. Also the compressed sensing receivers can be extended to handle channels with different Doppler scales on different paths, supplying ICI pattern estimates that can be used to equalize the ICI [29]. Using extensive numerical simulation and experimen tal results, researchers find that, in comparison to the LS receiver, the subspace methods show significant perfor mance improvements on channels that are sparse, but perform worse if most received energy comes from dif fuse multipath. To improve LS estimation, simple win dowing and dewindowing technique has been used to improve the accuracy of an existing basis expansion model (BEM) and develop a windowed leastsquares (WLS) estimator for doublyselective fading channels [89]. The compressed sensing algorithms do not suffer this draw back, and benefit significantly from the increased time resolution using sophisticated dictionaries. When account ing for different Doppler scales on different paths, BP can effectively handle channels with very large Doppler spread. 6. Doppler Shift Estimation for Underwater Acoustic Communications In underwater acoustic transmissions, Doppler effects can be caused by propagation medium movements or by the relative velocity between the transmitter and the re ceiver [12]. It often causes additional difficulties for the processing of received signals. For instance, underwater acoustic communications with submarines navigating at speeds up to 25 knots (12.8 m/s) is challenging, due to large Doppler range that the receiver must be able to cope with. Doppler effects result in a compression/ex pansion of the transmitted signal [90]. At the receiver side, it is desirable to remove the effect of these phe nomena before any further processing, such as timing and carrier recovery. A receiver which performs optimal phase synchronization and channel equalization jointly has been suggested for underwater communication sys tem [91]. Based on crosscorrelation among training sequences located at the beginning and at the end of transmitted data frame, an algorithm is developed for Doppler shift estimation [12]. From these training sequences two phase estimates are calculated: a coarse one and a precise one. The precise estimation has much smaller ambiguity range than the coarse one, but it yields higher precision for the estimated Doppler frequency. Authors have shown that a convenient combination of the coarse and precise phases leads to a good Doppler shift estimation within the speed range under consideration in the particular application discussed. Frequencydependent Doppler shifts caused by the low carrier frequency of the underwater acoustic channel communication. Depending on the null subcarrier of the ZPOFDM, multicarrier facilitates Doppler compensation, and the pilot subcarriers used in channel estimation [17] where an application of OFDM is investigated in wideband UWA channels with nonuniform Doppler shifts. To com pensate for the nonuniform Doppler distortion, a two step approach was used by resampling followed by high resolution uniform compensation of the residual Doppler. Also based on that, a blockbyblock receiver is suitable for fastvarying channels. Based on the availability of pilot tones in the OFDM transmission, a method devel oped to estimate the offset parameters, and a hybrid chan nel estimator was then proposed to combine the offset compensated [92]. Compressed Sensing (CS) can be developed as a method to solve the channel estimation problems for an UWA system. Based on the existence of a sparse representation of the treated signal and an overcomplete dictionary with a set of nonorthogonal bases, in [93] a new type of channel estimator using the compressed sensing theory is proposed, leading to a sparse channel estimation from the highly dependence on Doppler compensation. Instead of using various compensation approaches, the proposed index is designed by modeling Doppler shifts as the at oms shifts in overcomplete dictionary. From the results, this method improves the mean square error (MSE) per formance with lower complexity and hardware cost. In addition, the method has the additional benefit being less sensitive to Doppler rate variety. 7. Multicarrier Communication for Image & Video Transmission over Underwater Acoustic Channel Recently, there has been a growing interest in develop ment and deployment of image and video transmission techniques for underwater communication networks for scientific, environmental, commercial, and military pur poses [94,95]. High speed underwater image transmis sion capabilities can enable the next generation of un dersea expeditions. Efficient image transmission over the band limited underwater channels relies on two aspects [96]: 1) Efficient data compression, and 2) Bandwidth efficient modulation. The compressed sensing technique [97] generates mini mum amount of information necessary for transmission which makes it useful in underwater communication. The Copyright © 2013 SciRes. IJCNS
H. ESMAIEL, D. C. JIANG 372 combination of compressed sensing and nonlinear analog processing can also been employed as joint source and channel coding [3]. Underwater multimedia sensor net works (UMSNs) [98] have been proposed and drawn the immediate attention in the research community. However, the practical implementation of these currently designed and envisioned applications directly depends on reliabil ity and qualityaware communication capabilities of the deployed UMSNs. Comprehensive performance evalua tion of error concealment and error correction algorithms for qualityaware image transmission over UMSNs is reported in [99]. For highspeed image transmission using Multicarrier modulation [100,101] MMSE based equalization with the placement of a pilot symbol for very of three subcarrier payload resulted in good performance close to the ideal performance of the equalization. Set Partitioning in Hie rarchical Trees (SPIHT) is an efficient waveletbased progressive imagecompression technique, designed to minimize the meansquared error (MSE) between the original and decoded imagery used for highly compres sion technique [102105]. Since underwater acoustic chan nel suffers from significant bit error rates, some mecha nism to protect the encoded image is required. 16HQAM used as mapper for ZPOFDM, for unequal error protection using HQAM modulation technique it produce highly capability of highly speed image trans mission over UWAC. Multicarrier communication sys tem can transmit high speed image without equalizer and also with small overhead forward error correction bits. System depend on the guard interval, large zeropad guard interval of ZPOFDM system used to avoid infor mation loss incurred by the overlapadd operation to im prove system performance over underwater acoustic channels with large Doppler spread [106,107]. The realtime wireless video transmission from an un derwater vehicle to a surface platform represents one of the last milestones in the development of autonomous systems for ocean exploration and monitoring. The goal of combining video transmission over UWAC approaches is to mitigate the gap between the bit rate needed for video transmission and that supported by the acoustic channel. Konstantinos Pelekanakis [108] developed a vi deo transmission technique over underwater acoustic chan nel using coherent single carrier modulation technique for phase detection. For that a high rate acoustic link for underwater video transmission was implemented, where image encoding was accomplished using the JPEG DCT, scalar quantization and runlength Huffman encoding. Transmitter processing includes signal constellations. A system employing variable rate MQAM techniques was designed and applied to the experimental data trans mitted over a short vertical channel. Excellent results were obtained at bit rates up to 150 kbps, using modula tion methods with bandwidth efficiency as high as 6 bits/sec/Hz. Such rate is sufficient to support realtime transmission of compressed video. The feasibility of real time video transmission over short horizontal acoustic links was addressed in [109] where the standard MPEG 4 video compression technique and a waveletbased me thod were combined with acoustic transmission based on coded OFDM modulation to study the feasibility of video transmission using an acoustic system for deepsea oil field supervisory control and inspection. The wavelet based encoder algorithm includes techniques that deal with spatial and temporal redundancies in video sequences. In contrast to MPEG4 compression, spatial redundancies are exploited by applying the Discrete Wavelet Trans form (DWT) to each of the frames composing the video sequence. A motion compensation algorithm [110] to reduce temporal redundancies has been incorporated, which is a core functionality of the MPEG4 encoder. The codec is based on the popular Differential Pulse Code Modulation (DPCM) model, which is widely used in video compression standards. Its main components are the DWT multilevel decomposition of the input frame using the biorthogonal wavelet, quantization of the DWT coefficients using Vector Quantization (VQ). The images are smoother due to the high quantization ratio for the high subbands of the wavelet decomposition and, as expected, no blocking effect is noticed at all [109]. 8. Summary & Conclusion In this paper we surveyed multicarrier communication techniques used for data transmission over the underwa ter acoustic channel up to as the major aspects can be summarized as: 1) Underwater channel model and its challenging in multicarrier communication modulation like OFDM and FBMC summarized and detail reviewed; 2) The advances of underwater acoustics technology ap plications and areas emerged in; 3) Channel characteris tics and channel models in reducing noise and channel noise estimates; 4) Multicarrier modulation and multi input multioutput techniques detailed description for un derwater communication; 5) Doppler estimation and un derwater channel estimation challenging discuses and tech niques used evaluated and summarized; 6) Performance evaluation for different channel coder used for underwa ter acoustic multicarrier modulation; 7) Changing and used techniques for high speed image and video trans mission over UWAC also summarized and discussed. REFERENCES [1] D. Brady and J. C. Preisig, “Underwater Acoustic Com munications,” In: H. V. Poor and G. W. Wornell, Eds., Wireless Communications: Signal Processing Perspec tives, Vol. 8, PrenticeHall, Inc., Upper Saddle River, 1998, pp. 330379. Copyright © 2013 SciRes. IJCNS
H. ESMAIEL, D. C. JIANG 373 [2] A. C. Singer, J. K. Nelson and S. S. Kozat, “Signal Proc essing for Underwater Acoustic Communications,” IEEE Communications Magazine, Vol. 47, No. 1, 2009, pp. 90 96. doi:10.1109/MCOM.2009.4752683 [3] I. Iglesias, A. Song, J. GarciaFrias, M. Badiey and G. R. Arce, “Image Transmission over the Underwater Acoustic Channel via Compressive Sensing,” 45th Annual Confer ence on Information Sciences and Systems, Baltimore, 2325 March 2011, pp. 16. [4] M. Stojanovic, “Underwater Wireless Communications: Current Achievements and Research Challenges,” IEEE Oceanic Engineering Society Newsletter, Vol. 41, No. 1, 2006, pp. 1013. [5] Y. Labrador, M. Karimi, D. Pan and J. Miller, “Modula tion and Error Correction in the Underwater Acoustic Communication Channel,” International Journal of Com puter Science and Network Security, Vol. 9, No. 7, 2009, pp. 123130. [6] S.J. Hwang and P. Schniter, “Efficient Communication over Highly Spread Underwater Acoustic Channels,” In ternational Conference on Mobile Computing and Net working, Montreal, 914 September 2007, pp. 1118. [7] F. De Rango, F. Veltri and P. Fazio, “A Multipath Fading Channel Model for Underwater Shallow Acoustic Com munications,” IEEE International Conference on Com munications, Ottawa, 1015 June 2012, pp. 38113815. [8] L. M. Wolff, E. Szczepanski and S. BadriHoeher, “Acou stic Underwater Channel and Network Simulator,” OC EANS, Yeosu, 2124 May 2012, pp. 16. [9] A. Sehgal, I. Tumar and J. Schönwälder, “AquaTools: An Underwater Acoustic Networking Simulation Toolkit,” OCEANS, Sydney, 2427 May 2010, pp. 110. [10] M. Stojanovic, “On the Relationship between Capacity and Distance in an Underwater Acoustic Communication Channel,” ACM SIGMOBILE Mobile Computing and Communications Review, Vol. 11, No. 4, 2007, pp. 3443. doi:10.1145/1347364.1347373 [11] A. Fish, S. Gurevich, R. Hadani, A. Sayeed and O. Sch wartz, “DelayDoppler Channel Estimation with Almost Linear Complexity,” 2012, in press. [12] J. Trubuil and T. Chonavel, “Accurate Doppler Estimation for Underwater Acoustic Communications,” OCEANS, Yeosu, 2124 May 2012, pp. 15. [13] M. Vajapeyam, S. Vedantam, U. Mitra, J. C. Preisig and M. Stojanovic, “Time Cooperative Schemes for Under water Acoustic Communications,” IEEE Journal of Oce anic Engineering, Vol. 33, No. 4, 2008, pp. 489501. doi:10.1109/JOE.2008.2005338 [14] Y. Chen, X. Xu, L. Zhang and Z. Zou, “Design and Ap plication of Dynamic Coding in Shallow Water Acoustic Communications,” OCEANS, Yeosu, 2124 May 2012, pp. 16. [15] J. S. Panaro, F. R. Lopes, L. M. Barreira and F. E. Souza, “Underwater Acoustic Noise Model for Shallow Water Communications,” Presented at the Brazilian Telecom munication Symposium, 2012. [16] L. Chunshan, Y. V. Zakharov and C. Teyan, “Doubly Se lective Underwater Acoustic Channel Model for a Mov ing Transmitter/Receiver,” IEEE Transactions on Vehi cular Technology, Vol. 61, No. 3, 2012, pp. 938950. doi:10.1109/TVT.2012.2187226 [17] B. Li, S. Zhou, M. Stojanovic, L. Freitag and P. Willett, “Multicarrier Communication over Underwater Acoustic Channels with Nonuniform Doppler Shifts,” IEEE Jour nal of Oceanic Engineering, Vol. 33, No. 2, 2008, pp. 198209. doi:10.1109/JOE.2008.920471 [18] S. Yerramalli and U. Mitra, “On Optimal Resampling for OFDM Signaling in DoublySelective Underwater Acou stic Channels,” OCEANS, Quebec City, 1115 September 2008, pp. 16. [19] G. L. Stüber, “Principles of Mobile Communication,” Springer, New York, 2011. [20] X. Xiaomei, C. Yougan, Z. Lan and F. Wei, “Comparison of the Performance of LDPC Codes over Different Un derwater Acoustic Channels,” 12th IEEE International Conference on Communication Technology, Nanjing, 11 14 November 2010, pp. 155158. [21] C. R. Berger, W. Chen, S. Zhou and J. Huang, “A Simple and Effective Noise Whitening Method for Underwater Acoustic Orthogonal Frequency Division Multiplexing,” The Journal of the Acoustical Society of America, Vol. 127, No. 4, 2010, pp. 23582367. doi:10.1121/1.3309452 [22] N. Nasri, L. Andrieux, A. Kachouri and M. Samet, “Effi cient Encoding and Decoding Schemes for Wireless Un derwater Communication Systems,” 7th International MultiConference on Systems Signals and Devices, Am man, 2730 June 2010, pp. 16. [23] M. D. Haque, S. E. Ullah and M. R. Ahmed, “Perform ance Evaluation of a Wireless Orthogonal Frequency Di vision Multiplexing System under Various Concatenated FEC ChannelCoding Schemes,” 11th International Con ference on Computer and Information Technology, Khul na, 2427 December 2008, pp. 9497. [24] M. Chitre, S. Shahabudeen, L. Freitag and M. Stojanovic, “Recent Advances in Underwater Acoustic Communica tions & Networking,” OCEANS, Quebec City, 1115 Sep tember 2008, pp. 110. [25] J. Trubuil, A. Goalic and N. Beuzelin, “Synchronization and Channel Coding in Shallow Water Acoustic Commu nication,” OCEANS, Quebec City, 1115 September 2008, pp. 15. [26] J. Trubuil, A. Goalic and N. Beuzelin, “An Overview of channel Coding for Underwater Acoustic Communica tions,” Military Communications Conference, Orlando, 29 October1 November 2012, pp. 17. [27] W. Han, J. Huang and M. Jiang, “Performance Analysis of Underwater Digital Speech Communication System Based on LDPC Codes,” 4th IEEE Conference on Indus trial Electronics and Applications, Xi’an, 2527 May 2009, pp. 567570. [28] J. Huang, S. Zhou and P. Willett, “Nonbinary LDPC Cod ing for Multicarrier Underwater Acoustic Communica tion,” IEEE Journal on Selected Areas in Communica tions, Vol. 26, No. 9, 2008, pp. 16841696. doi:10.1109/JSAC.2008.081208 [29] C. R. Berger, S. Zhou, J. C. Preisig and P. Willett, “Sparse Channel Estimation for Multicarrier Underwater Copyright © 2013 SciRes. IJCNS
H. ESMAIEL, D. C. JIANG 374 Acoustic Communication: From Subspace Methods to Compressed Sensing,” IEEE Transactions on Signal Pro cessing, Vol. 58, No. 3, 2010, pp. 17081721. doi:10.1109/TSP.2009.2038424 [30] Z. Wang, S. Zhou, G. B. Giannakis, C. R. Berger and J. Huang, “FrequencyDomain Oversampling for ZeroPad ded OFDM in Underwater Acoustic Communications,” IEEE Journal of Oceanic Engineering, Vol. 37, No. 1, 2012, pp. 1424. doi:10.1109/JOE.2011.2174070 [31] A. K. Morozov and J. C. Preisig, “Underwater Acoustic Communications with MultiCarrier Modulation,” OCEANS, Boston, 1821 September 2006, pp. 16. [32] P. Kumar and P. Kumar, “Performance Evaluation of DFTSpread OFDM and DCTSpread OFDM for Under water Acoustic Communication,” IEEE Vehicular Tech nology Conference, Quebec City, 36 September 2012, pp. 15. [33] B. FarhangBoroujeny, “OFDM Versus Filter Bank Mul ticarrier,” IEEE Signal Processing Magazine, Vol. 28, No. 3, 2011, pp. 92112. doi:10.1109/MSP.2011.940267 [34] R. van Nee and R. Prasad, “OFDM for Wireless Multime dia Communications,” Artech House, Inc., Norwood, 2000. [35] Y. G. Li, “Orthogonal Frequency Division Multiplexing for Wireless Communications,” SpringerVerlag, New York, 2009. [36] B. Li, S. Zhou, M. Stojanovic, L. Freitag, J. Huang and P. Willett, “MIMOOFDM over an Underwater Acoustic Channel,” OCEANS, Vancouver, 29 September4 October 2007, pp. 16. [37] A. Thottappilly, “OFDM for Underwater Acoustic Com munication,” Master of Science, Virginia Polytechnic In stitute and State University, Blacksburg, 2011. [38] B. Li, S. Zhou, M. Stojanovic, L. Freitag and P. Willet, “Multicarrier Underwater Acoustic Communications over FastVarying Channels,” IEEE Journal of Oceanic Engi neering, Vol. 33, No. 2, 2008, pp. 198209. doi:10.1109/JOE.2008.920471 [39] C. Polprasert, J. A. Ritcey and M. Stojanovic, “Capacity of OFDM Systems over Fading Underwater Acoustic Channels,” IEEE Journal of Oceanic Engineering, Vol. 36, No. 4, 2011, pp. 514524. doi:10.1109/JOE.2011.2167071 [40] C. R. Berger, J. Gomes and J. M. Moura, “Study of Pilot Designs for CyclicPrefix OFDM on TimeVarying and Sparse Underwater Acoustic Channels,” OCEANS, San tander, 69 June 2011, pp. 18. [41] X. Huang, “Capacity CriterionBased Power Loading for Underwater Acoustic OFDM System with Limited Feed back,” IEEE International Conference on Wireless Com munications, Networking and Information Security, Bei jing, 2527 June 2010, pp. 5458. [42] T. Pedersen and A. Bloomberg, “OFDM PilotAided Un derwater Acoustic Channel Estimation,” International Journal of Electrical, Electronics and Computer Systems, Vol. 9, No. 1, 2012, pp. 13. [43] R. G. Baraniuk, “Compressive Sensing,” IEEE Signal Processing Magazine, Vol. 24, No. 4, 2007, pp. 118121. doi:10.1109/MSP.2007.4286571 [44] P. Kumar, “DCT Based OFDM for Underwater Acoustic Communication,” 1st International Conference on Recent Advances in Information Technology, Dhanbad, 1517 March 2012, pp. 170176. [45] T. Ebihara and K. Mizutani, “Underwater Acoustic Com munication With an Orthogonal Signal Division Multi plexing Scheme in Doubly Spread Channels,” IEEE Journal of Oceanic Engineering, Vol. PP, No. 99, 2013, pp. 112. doi:10.1109/JOE.2013.2245273 [46] M. Stojanovic, “Recent Advances in HighSpeed Under water Acoustic Communications,” IEEE Journal of Ocea nic Engineering, Vol. 21, No. 2, 1996, pp. 125136. doi:10.1109/48.486787 [47] R. Mutagi, “Pseudo Noise Sequences for Engineers,” Electronics & Communication Engineering Journal, Vol. 8, No. 2, 1996, pp. 7987. [48] H. Jinxing, Y. R. Zheng, W. Jintao and S. Jian, “Dual PN Padding TDSOFDM for Underwater Acoustic Commu nication,” Oceans, Hampton Road, 1419 October 2012, pp. 14. [49] Z. Wang, S. Zhou, J. Catipovic and P. Willett, “Asyn chronous Multiuser Reception for OFDM in Underwater Acoustic Communications,” IEEE Transactions on Wire less Communications, Vol. 12, No. 3, 2013, pp. 112. [50] L. Y. Jing and J. G. Huang, “Cyclic Shift Keying Spread Spectrum OFDM Method over Underwater Acoustic Channel,” IEEE International Conference on Signal Pro cessing, Communication and Computing, Hong Kong, 1215 August 2012, pp. 798801. [51] J.G. Huang, C.B. He, J. Han and Q.F. Zhang, “Cyclic Shift Keying Spread Spectrum Underwater Acoustic Communication,” Acta Physica Sinica, Vol. 58, No. 12, 2009, pp. 83798385. [52] E. Sozer, J. Proakis, R. Stojanovic, J. Rice, A. Benson and M. Hatch, “Direct Sequence Spread Spectrum Based Modem for Under Water Acoustic Communication and Channel Measurements,” OCEANS, Seattle, 1316 Sep tember 1999, pp. 228233. [53] N. Papandreou and T. Antonakopoulos, “Bit and Power Allocation in Constrained Multicarrier Systems: The Sin gleUser Case,” EURASIP Journal on Advances in Signal Processing, Vol. 2008, No. 1, 2008, p. 11. [54] X. Huang and V. B. Lawrence, “BandwidthEfficient Bit and Power Loading for Underwater Acoustic OFDM Communication System with Limited Feedback,” IEEE 73rd Vehicular Technology Conference, Yokohama, 15 18 May 2011, pp. 15. [55] L. Yang, L. Zhou and M. Yu, “Adaptive Bit Loading Algorithm for OFDM Underwater Acoustic Communica tion System,” International Conference on Electronics and Optoelectronics, Dalian, 2931 July 2011, pp. 350 352. [56] H. Xiaopeng and L. Yong, “Capacity CriterionBased Power Allocation for OFDM Cooperative Underwater Acoustic Communications with Limited Feedback,” OCEANS, Hampton Road, 1419 October 2012, pp. 14. [57] K. Tu, T. M. Duman, M. Stojanovic and J. G. Proakis, “MultipleResampling Receiver Design for OFDM over DopplerDistorted Underwater Acoustic Channels,” IEEE Journal of Oceanic Engineering, Vol. 38, No. 2, 2013, pp. Copyright © 2013 SciRes. IJCNS
H. ESMAIEL, D. C. JIANG 375 333346. doi:10.1109/JOE.2012.2221812 [58] M. Uysal, “Cooperative Communications for Improved Wireless Network Transmission: Framework for Virtual Antenna Array Applications,” Information Science Ref erence, University of Waterloo, Waterloo, 2010. [59] K. Tu, “MultiCarrier Communications over Underwater Acoustic Channels,” Doctor of Philosophy, Arizona State University, Phoenix, 2011. [60] Q. Zhang, A. B. Kokkeler and G. J. Smit, “An Oversam pled Filter Bank Multicarrier System for Cognitive Ra dio,” 19th International Symposium on Personal, Indoor and Mobile Radio Communications, Cannes, 1518 Sep tember 2008, pp. 15. doi:10.1109/PIMRC.2008.4699865 [61] P. Amini, C. RongRong and B. FarhangBoroujeny, “Fil terbank Multicarrier for Underwater Communications,” 49th Annual Allerton Conference on Communication, Control, and Computing, Monticello, 2830 September 2011, pp. 639646. doi:10.1109/Allerton.2011.6120228 [62] B. Le Floch, M. Alard and C. Berrou, “Coded Orthogonal Frequency Division Multiplex,” Proceedings of the IEEE, Vol. 83, No. 6, 1995, pp. 982996. doi:10.1109/5.387096 [63] M. Alard, “Construction of a Multicarrier Signal,” CA Patent CA 2,220,107, 1996. [64] P. Amini, C. H. Yuen, R.R. Chen and B. Farhang Boroujeny, “Isotropic Filter Design for MIMO Filter Bank Multicarrier Communications,” IEEE Sensor Array and Multichannel Signal Processing Workshop, Jerusa lem, 47 October 2010, pp. 8992. doi:10.1109/SAM.2010.5606775 [65] L. Toni, L. Rossi, N. Agoulmine and J.G. Fontaine, “Virtual UEP for Progressive Image Transmission in Underwater Communication,” 2009, in press. [66] L. Baosheng, H. Jie, Z. Shengli, K. Ball, M. Stojanovic, L. Freitag and P. Willett, “MIMOOFDM for HighRate Un derwater Acoustic Communications,” IEEE Journal of Oceanic Engineering, Vol. 34, No. 4, 2009, pp. 634644. doi:10.1109/JOE.2009.2032005 [67] D. B. Kilfoyle, J. C. Preisig and A. B. Baggeroer, “Spatial Modulation Experiments in the Underwater Acoustic Channel,” IEEE Journal of Oceanic Engineering, Vol. 30, No. 2, 2005, pp. 406415. doi:10.1109/JOE.2004.834168 [68] S. Roy, T. M. Duman, V. McDonald and J. G. Proakis, “HighRate Communication for Underwater Acoustic Channels Using Multiple Transmitters and Space—Time Coding: Receiver Structures and Experimental Results,” IEEE Journal of Oceanic Engineering, Vol. 32, No. 3, 2007, pp. 663688. doi:10.1109/JOE.2007.899275 [69] R. F. Ormondroyd, “A Robust Underwater Acoustic Com munication System Using OFDMMIMO,” OCEANS, Ab erdeen, 1821 June 2007, pp. 16. [70] P. Bouvet and A. Loussert, “An Analysis of MIMO OFDM for Shallow Water Acoustic Communications,” OCEANS, Waikoloa, 1922 September 2011, pp. 15. [71] M. Stojanovic, “MIMO OFDM over Underwater Acous tic Channels,” IEEE 43rd Asilomar Conference on Sig nals, Systems and Computers, Piscataway, 14 November 2009, pp. 605609. [72] Y. Li, N. Seshadri and S. Ariyavisitakul, “Channel Esti mation for OFDM Systems with Transmitter Diversity in Mobile Wireless Channels,” IEEE Journal on Selected Areas in Communications, Vol. 17, No. 3, 1999, pp. 461 471. doi:10.1109/49.753731 [73] H. Minn and V. K. Bhargava, “An Investigation into TimeDomain Approach for OFDM Channel Estimation,” IEEE Transactions on Broadcasting, Vol. 46, No. 4, 2000, pp. 240248. doi:10.1109/11.898744 [74] M. Stojanovic, “OFDM for Underwater Acoustic Com munications: Adaptive Synchronization and Sparse Chan nel Estimation,” IEEE International Conference on Acou stics, Speech and Sign al Processing, Las Vegas, 31 March 4 April 2008, pp. 52885291. [75] B. Saltzberg, “Performance of an Efficient Parallel Data Transmission System,” IEEE Transactions on Communi cation Technology, Vol. 15, No. 6, 1967, pp. 805811. doi:10.1109/TCOM.1967.1089674 [76] C. A. Belfiore and J. H. Park Jr., “Decision Feedback Equalization,” Proceedings of the IEEE, Vol. 67, No. 8, 1979, pp. 11431156. doi:10.1109/PROC.1979.11409 [77] M. Stojanovic, L. Freitag and M. Johnson, “Chan nelEstimationBased Adaptive Equalization of Under water Acoustic Signals,” OCEANS, Seattle, 1316 Sep tember 1999, pp. 590595. [78] R. Coates, “Underwater Acoustic Communications,” OC EANS, Vol. 3, 1993, pp. 420425. [79] M. Stojanovic, “Underwater Acoustic Communication,” Wiley Encyclopedia of Electrical and Electronics Engi neering, 1999. [80] M. Chitre, S. Shahabudeen and M. Stojanovic, “Under water Acoustic Communications and Networking: Recent Advances and Future Challenges,” Marine Technology Society Journal, Vol. 42, No. 1, 2008, pp. 103116. doi:10.4031/002533208786861263 [81] P. Shukla and L. Turner, “ChannelEstimationBased Adaptive DFE for Fading Multipath Radio Channels,” IEE Proceedings of Communications, Speech and Vision, Vol. 138, No. 6, 1991, pp. 525543. [82] P. Zarrinkhat, M. Ardakani and R. Yazdani, “Channel Estimation Considerations for Iterative Decoding in Wireless Communications,” IEEE International Confer ence on Communications, Glasgow, 2428 June 2007, pp. 10061011. [83] E. Calvo and M. Stojanovic, “Efficient ChannelEstima tionBased Multiuser Detection for Underwater CDMA Systems,” IEEE Journal of Oceanic Engineering, Vol. 33, No. 4, 2008, pp. 502512.doi:10.1109/JOE.2008.2005355 [84] J. G. Proakis, E. M. Sozer, J. A. Rice and M. Stojanovic, “Shallow Water Acoustic Networks,” IEEE Communica tions Magazine, Vol. 39, No. 11, 2001, pp. 114119. doi:10.1109/35.965368 [85] Y. R. Zheng, “Channel Estimation and PhaseCorrection for Robust Underwater Acoustic Communications,” IEEE Military Communications Conference, Orlando, 2931 October 2007, pp. 16. [86] J. Ling, X. Tan, T. Yardibi, J. Li, H. He and M. L. Nor denvaad, “Enhanced Channel Estimation and Efficient Symbol Detection in MIMO Underwater Acoustic Com Copyright © 2013 SciRes. IJCNS
H. ESMAIEL, D. C. JIANG Copyright © 2013 SciRes. IJCNS 376 munications,” 43rd Asilomar Conference on Signals, Systems and Computers, Pacific Grove, 14 November 2009, pp. 600604. [87] K. Taehyuk and R. A. Iltis, “Matching Pursuits Channel Estimation for an Underwater Acoustic OFDM Modem,” IEEE International Conference on Acoustics, Speech and Signal Processing, Las Vegas, 31 March4 April 2008, pp. 52965299. [88] J. Huang, C. R. Berger, S. Zhou and J. Huang, “Com parison of Basis Pursuit Algorithms for Sparse Channel Estimation in Underwater Acoustic OFDM,” OCEANS, Sydney, 2427 May 2010, pp. 16. [89] Q. Fengzhong and Y. Liuqing, “On the Estimation of DoublySelective Fading Channels,” IEEE Transactions on Wireless Communications, Vol. 9, No. 4, 2010, pp. 12611265. doi:10.1109/TWC.2010.04.080631 [90] B. S. Sharif, J. Neasham, O. R. Hinton and A. E. Adams, “A Computationally Efficient Doppler Compensation System for Underwater Acoustic Communications,” IEEE Journal of Oceanic Engineering, Vol. 25, No. 1, 2000, pp. 5261. doi:10.1109/48.820736 [91] M. Stojanovic, J. A. Catipovic and J. G. Proakis, “Phase Coherent Digital Communications for Underwater Acou stic Channels,” IEEE Journal of Oceanic Engineering, Vol. 19, No. 1, 1994, pp. 100111. doi:10.1109/48.289455 [92] W. Zhaohui, Z. Shengli, J. C. Preisig, K. R. Pattipati and P. Willett, “Clustered Adaptation for Estimation of Time Varying Underwater Acoustic Channels,” IEEE Transac tion on Signal Processing, Vol. 60, No. 6, 2012, pp. 30793091. doi:10.1109/TSP.2012.2189769 [93] H. Yu and J. Guo, “Compressed Sensing: Underwater Acoustic Channel Estimation with Doppler Shifts,” 8th International Conference on Wireless Communications, Networking and Mobile Computing, Shanghai, 2123 September 2012, pp. 14. [94] I. F. Akyildiz, D. Pompili and T. Melodia, “Challenges for Efficient Communication in Underwater Acoustic Sensor Networks,” ACM Sigbed Review, Vol. 1, No. 2, 2004, pp. 38. doi:10.1145/1121776.1121779 [95] E. M. Sozer, M. Stojanovic and J. G. Proakis, “Underwa ter Acoustic Networks,” IEEE Journal of Oceanic Engi neering, Vol. 25, No. 1, 2000, pp. 7283. doi:10.1109/48.820738 [96] C. Pelekanakis, M. Stojanovic and L. Freitag, “High Rate Acoustic Link for Underwater Video Transmission,” OC EANS, Vol. 2, 2003, pp. 10911097. [97] I. Iglesias, B. Lu, J. GarciaFrias and G. Arce, “Non Linear Mappings for Transmission of Compressed Sens ing Images,” 48th Annual Allerton Conference on Com munication, Control, and Computing, Allerton, 29 Sep tember1 October 2010, pp. 726732. [98] S. A. Alkharabsheh and M. I. Younis, “Dynamics of MEMS Arches of Flexible Supports,” Journal of Micro electromechanical Systems, Vol. 22, No. 1, 2013, pp. 216224. [99] P. SarisarayBoluk, V. Gungor, S. Baydere and A. Har manci, “Quality Aware Image Transmission over Under water Multimedia Sensor Networks,” Ad Hoc Networks, Vol. 9, No. 7, 2011, pp. 12871301. doi:10.1016/j.adhoc.2011.02.007 [100] J.W. Han, S.Y. Kim, K.M. Kim, S.Y. Chun and K. Son, “Design of OFDM System for High Speed Under water Communication,” International Conference on Com putational Science and Engineering, Vancouver, 2931 August 2009, pp. 988992. [101] T. B. Santoso and G. Hendrantoro, “Image Transmission with OFDM Technique in Underwater Acoustic Envi ronment,” 7th International Conference on Telecommu nication Systems, Services, and Applications, Bali, 3031 October 2012, pp. 3741. [102] A. Said and W. A. Pearlman, “A New, Fast and Efficient Image Codec Based on Set Partitioning in Hierarchical Trees,” IEEE Transactions on Circuits and Systems for Video Technology, Vol. 6, No. 3, 1996, pp. 243250. doi:10.1109/76.499834 [103] C. Murphy and H. Singh, “Wavelet Compression with Set Partitioning for Low Bandwidth Telemetry from AUVs,” Proceedings of the 5th ACM International Workshop on UnderWater Networks, Woods Hole, 30 September1 Oc tober 2010, p. 1. doi:10.1145/1868812.1868813 [104] M. A. Kader, F. Ghani and R. B. Ahmed, “Unequal Error Protection for SPIHT Coded Image Transmission over Erroneous Wireless Channels,” Asian Transactions on Fundamentals of Electronics, Communication & Multi media, Vol. 1, No. 5, 2011, pp. 16. [105] B. Tomasi, L. Toni, P. Casari, J. Preisig and M. Zorzi, “A Study on the SPIHT Image Coding Technique for Un derwater Acoustic Communications,” Proceedings of the 6th ACM International Workshop on Underwater Net works, Seattle, 12 December 2011, pp. 19. doi:10.1145/2076569.2076578 [106] H. Esmaiel and D. Jiang, “Image Transmission over Un derwater Acoustic Environment Using OFDM Technique with HQAM Mapper,” 3rd IEEE International Confer ence on Information Science and Technology, Yangzhou, 2325 March 2013, pp. 15961601. [107] H. Esmaiel and D. Jiang, “SPIHT Coded Image Trans mission over Underwater Acoustic Channel with Unequal Error Protection Using HQAM,” 3rd IEEE International Conference on Information Science and Technology, Yangzhou, 2325 March 2013, pp. 13651371. [108] K. Pelekanakis, “Design and Analysis of a HighRate Acoustic Link for Underwater Video Transmission,” Master of Science, Massachusetts Institute of Technology, Cambridge, 2004. [109] J. Ribas, D. Sura and M. Stojanovic, “Underwater Wire less Video Transmission for Supervisory Control and In spection Using Acoustic OFDM,” OCEANS, Seattle, 20 23 September 2010, pp. 19. [110] D. F. Hoag, V. K. Ingle and R. J. Gaudette, “Low BitRate Coding of Underwater Video Using Wavelet Based Compression Algorithms,” IEEE Journal of Oce anic Engineering, Vol. 22, No. 2, 1997, pp. 393400. doi:10.1109/48.585958
