CSIT DOI 10.1007/s40012-017-0179-3
S.I. : VISVESVARAYA
Experimental studies on the performance of underwater optical communication link with channel coding and interleaving Prasad Naik Ramavath1 U. Shripathi Acharya1
•
Amardeep Kumar1 • Shrutkirthi Shashikant Godkhindi1
•
Received: 14 April 2017 / Accepted: 25 October 2017 Ó CSI Publications 2017
Abstract We have evaluated performance of underwater optical communication links under conditions of turbulence and blockage. A 3 m long water column filled with sea water containing sediments and organic matter has been set up. Additional water jets have been placed in the medium to create turbulence and a motor driven paddle running with a constant rpm has been introduced to simulate conditions of blockage and added turbulence. Turbulence in the medium can cause beam wander creating random errors, while blockage blocks the data for certain duration leading to burst errors. To mitigate the errors caused by turbulence and blockage turbulence we have employed a BCH (31, 11) error correcting code with error correcting capability t = 6. We have also introduced a 31 9 31 block interleaver to improve the burst error correcting capabilities of the code. The bit error rate performance of this system has been evaluated and plots describing the improvement in reliability of information transfer across the channel as a function of transmit power (in dBm) have been plotted. This experimental study provides insight into the nature of errors that can occur in under water optical communication and the means that can
& Prasad Naik Ramavath
[email protected] Amardeep Kumar
[email protected] Shrutkirthi Shashikant Godkhindi
[email protected] U. Shripathi Acharya
[email protected] 1
Electronics and Communication Engineering Department, National Institute of Technology Karnataka, Surathkal 575025, India
be adopted to improve the reliability of information transfer across such channels. Keywords UWOC Turbulence Blockage Random errors Burst errors BCH code Interleaved BCH code
1 Introduction Modern wireless communications is characterized by a relentless demand for high-end services like security, reliability and high-speed transmission. Reliable communication over wireless channels is achieved by employing channel coding and diversity. There seems to be a never ending demand for wireless services and this has led to the requirement of very high data rates to be sustained over wireless channels characterized by severe impediments (such as signal fading) and limited bandwidth. This has led to the design of terrestrial free space optic (FSO) systems which can sustain high data rates without any spectrum licensing issues. In addition to terrestrial FSO, there has also been an increased interest in under water optical communication (UWOC) due to its possible applications in strategic and military domains. Such a system can allow for reliable and high data rate short distance (several tens of meters) information transfer between under water vehicles or between under water vehicles and shore installations. This has motivated us to take up experimental studies on the design of high speed reliable under water communication systems. In recent years, interest in UWOC and its possible applications has grown enormously. Both theoretical [1] and experimental studies [2, 3, 13] have been reported. Djoordjevic and Wang [4, 5] have given a detailed description of the terrestrial free space optical channel and have suggested channel coding schemes for the same.
123
CSIT
Inspired by this work, we have started experimental studies on UWOC. We will work on various aspects of setting up of UWOC links such as the alignment and tracking problem, synchronization problem and the problem of designing appropriate channel coding/MIMO and interleaving schemes to improve the reliability of link. In this report, we have described the progress done so far in setting up the underwater channel, making provisions for introducing turbulence and blockage (to simulate such events in the actual channel) and some channel codes/interleaving that have been synthesized to improve the reliability of information transfer. The remaining part of the paper is organized as follows. In Sect. 2 the description of the system model is presented, Sect. 3 provides a description of the experimental setup and specifies the arrangements made to set up random and burst errors in the channel. Section 4 presents a discussion of the BCH code and interleaver arrangement used to minimize the effects of random and burst errors introduced by turbulence and blocking. Section 5 presents a discussion of the error rate performance for the given experimental setup. The paper is concluded in Sect. 6 with a discussion of the work that is planned so as to enable the design of a complete system for under water communication.
2 System model Optical wireless communication (OWC) requires the presence of line of sight (LOS). Information is usually modulated on an optical carrier wave at a suitable wavelength chosen to minimize losses due to absorption and scattering. The most commonly employed modulation scheme is On–Off Keying where the binary electrical signal directly drives the optical source. The block diagram in Fig. 1 describes the experimental set up of the underwater optical communication used.
digits are used to modulate (On–Off keying) the optical source. The transmitter section consists of driver circuit, the LED/ LASER source and a collimated lens arrangement to focus the light beam into the water column. An optical wavelength of 473 nm (which enables transmission with minimum attenuation and absorption) is employed. The details of the drive circuit are provided in Fig. 2, HFA 3096 op-amp has high slew rate (100 V/lsec) [6] and two PBLS 6024 D series transistors (NPN and PNP) are suitable for high data-rates [7]. The LED drive circuit comprising of a push pull amplifier is designed to drive currents as large as 400 mA into the LED source. The LED source is rated at 0.5 W [8]. The source is characterized by very small values of carrier recombination time. This will allow the system to support high modulation speeds (of the order of several MHz) in UWOC links. 2.2 Receiver The Received signal can be expressed as [12], r ¼ gðIr þ Ib Þ þ n
ð1Þ
In Eq. (1), g is the optical-to-electrical conversion coefficient, Ir is the received signal light intensity, Ib is the ambient light intensity, n is the additive white Gaussian noise with zero mean and variance of N20 , N0 is the power spectral density of noise. We have employed an optical band pass filter centered at 470 nm with a spectral width of 35 nm. The use of this filter prevents ambient light (which constitutes noise) from falling on the detector. A highly sensitive, low-noise, light intensity to voltage converter [9] is used, followed by trans-resistance amplifier. The reconstructed electrical signal provided by this unit is communicated to the comparator circuit which cleans up the distorted output and created a two level binary signal for further processing. The receiver circuit is illustrated in Fig. 3. 2.3 Underwater channel
2.1 Transmitter The data source consists of a text message which is converted into ASCII symbols and then into binary digits. These binary
For the evaluation and checking the performance of above system under conditions likely to be present in the field, (where turbulence and blockage can be experienced), we have established a 3-meter long experimental setup comprising of PVC pipes filled with sea water and glass tanks. Water jets have been employed to create turbulence and motor driven paddles rotating at fixed rpm have been placed to create conditions of additional turbulence and blockage.
3 Experimental setup and types of errors
Fig. 1 Block diagram of underwater optical communication
123
The first set of experiments were conducted with opaque PVC pipes filled with sea water. Since the pipe is opaque, ambient light is not able to enter the medium and thus the
CSIT Fig. 2 LED driving circuit
Fig. 3 Receiver circuit
Fig. 5 Experimental setup of 3-meter length
Fig. 4 1-meter fresh water filled PVC pipe
effects of background noise are largely eliminated. The experimental set up consisting of 1-m long PVC pipe is shown in Fig. 4. To include the effect of ambient light which is usually present in the underwater channel, the PVC pipe was replaced with glass aquarium type setup which is shown in Fig. 5. To generate turbulence, we have placed five water jets: one is close to transmitter, second one is near the receiver and remaining are placed between transmitter and receiver. With this set up, we have observed the occurrence of errors. We have observed that communication with a BER of 10-3 in the presence of turbulence was enabled at a transmit power level of 11.6 dBm. In Figs. 6 and 7, we have provided samples of the transmitted text and the text material as recovered at the receiver in the presence of turbulence and blockage at a transmit power level of 11.6 and 11.8 dBm respectively.
Fig. 6 Transmitted data
Next case to create blockage turbulence, we used a paddle. Paddle is rotated using motor of speed 120 rpm. So the paddle started blocking data for certain duration. The observed errors in this case are burst, hence these errors are burst errors. An example shown below of the received data in blocking case for the same transmitted data at 11.8 dBm of transmit power. The maximum burst error length observed is 10 characters (80 bits). Here rotating of paddle creates blocking as well as turbulences. Errors are marked with green and yellow colors (Fig. 8).
123
CSIT
a.
Construction of generator polynomial
BCH code of t = 4 bit error correcting code [10, 11] in GF(25) with minimum distance dmin = 9, block length n = 25 - 1=31 and the elements of group are 0; 1; a; a2 ; a3 ; a4 ; . . .a30 , these group elements are partitioned into conjugacy classes with conjugate roots, using these roots calculating the minimal polynomials using primitive polynomial in GF(25) is PðxÞ ¼ x5 þ x2 þ 1. Generator polynomial g(x) is least common multiple of minimal polynomials. gðxÞ ¼ x20 þ x18 þ x17 þ x13 þ x10 þ x9 þ x7 þ x6 þ x4 þ x2 þ 1
ð2Þ
Generator matrix (G) and parity check matrix (H) are obtained from generator polynomial g(x), minimum distance dmin is achieved using this construction is 13, so BCH (31, 11) code can correct t = 6 bit errors. Fig. 7 Received data under turbulence condition
b.
BCH encoding
Let data source m ¼ ½m0 ; m1 ; . . .mk1 is encoded and the corresponding code word is c ¼ ½c0 ; c1 ; . . .cn1 c ¼ ½m0 ; m1 ; . . .mk1 G:
ð3Þ
c. Syndrome decoding Let received vector r ¼ ½r0 ; r1 ; . . .rn1 at the output of the underwater channel and H is the parity check matrix. Syndrome S ¼ r H T
Fig. 8 Received data under blockage condition
ð4Þ
where ()T is Transpose of matrix. The syndrome matrix is S ¼ ½s0 ; s1 ; . . .snk . If S is all zero matrix, then received vector is code word, otherwise the received vector is not code word and it will detect the presence of error. The computed syndrome S from the received vector r depends on error pattern e ¼ ½e0 ; e1 ; . . .en1 .
4 Channel code
Table 1 Minimum transmit power under different conditions for error free transmission
It is observed that there is significant amount of error induced in the system due to underwater channel nature, induced turbulence and blocking. When analyzed this data the randomness in error was observed and the maximum bits corrupted consecutively were noted to be about eighty bits. Applying channel codes would reduce the data corruption, as the number of errors getting at a time is very high no channel could guarantee reconstruction of data. This leads to need of interleaving. Here BCH code with interleaving concept is adopted in order to recover the data.
Condition
123
Turbulence
Blockage and turbulence
Minimum transmit power to transmit error-free communication (dBm) Uncoded
11.6
BCH coded
10.4
Interleave BCH coded
8.8
Uncoded
14.8
BCH coded
12.3
Interleaved with BCH coding
10
CSIT
5 Experimental results
Fig. 9 Bit error rate of UWOC with turbulence uncoded, BCH coded and interleaved BCH coded
i:e; eH T ¼ S:
ð5Þ
The error pattern is smallest number of non-zero matrix; b ¼ r þ e. with this, we estimate the received data using C b is estimated data at receiver. Here C The (31, 11) BCH code can correct maximum of 6 bit error, but the maximum burst error length observed in blocking case is 80 bits. So Interleaved (31, 11) BCH code of interleave length 31 9 31 can correct maximum burst length of 186 (= 31 9 6 (Interleaver size) 9 (BCH error correcting capability)) bits.
In this Section, the BER of UWOC with respect to transmit power (dBm) under the influence of turbulence and blocking condition is presented. Later, the minimum power required to transmit error free data through underwater channel for different (normal, turbulence and blocking) conditions are tabulated in Table 1. Figure 9, shows the comparison between BER of uncoded data, BCH coded data and interleaved BCH coded data for turbulence environment. It is observed that for turbulence at BER of 10-3, BCH coded data has performance improvement of more than 1 dBm over uncoded data and more than 2.5 dBm advantage is achieved by Interleaved BCH coded data over uncoded data. Similarly Fig. 10 demonstrates BER performance for blocking condition. Performance improvement of 1.5 dBm is observed for BCH coded over uncoded data and interleaved BCH code achieves advantage of 3.5 dBm over uncoded data.
6 Conclusion and future work In this paper, we have analyzed the underwater and under sea water channel. The effect of turbulence and blocking is also analyzed. Beam dispersion was observed due to turbulence and a block errors was noted in the data due to blocking. This effect of turbulence can be equated to
Fig. 10 Bit error rate of UWOC with blocking uncoded, BCH coded and interleaved BCH coded
123
CSIT
moving wave structure while blocking can be considered due to aqua life or hard particles. In order to combat these effects BCH code with interleaving is employed and a significant improvement of about 2.5 dBm in BER was observed in turbulence condition and about 3.5 dBm improvement for blocking condition. We further propose to incorporate different channel codes in order to enhance performance of our system, acquisition and tracking point of LED and synchronization issues between transmitter and receiver in sea water.
4. 5.
6. 7. 8. 9.
References 1. Boucouvalas AC, Peppas KP, Yiannopoulos K, Ghassemlooy Z (2016) Underwater optical wireless communications with optical amplification and spatial diversity. IEEE Photonics Technol Lett 28(22):2613–2616 2. Simpson JA, Hughes BL, Muth JF (2012) Smart transmitters and receivers for underwater free-space optical communication. IEEE J Sel Areas Commun 30(5):964–974 3. Jamali MV, Khorramshahi P, Tashakori A, Chizari A, Shahsavari S, Abdollah Ramezani S, Salehi JA (2016) Statistical distribution
123
10. 11. 12.
13.
of intensity fluctuations for underwater wireless optical channels in the presence of air bubbles. In: 2016 Iran workshop on communication and information theory (IWCIT). IEEE, pp 1–6 Djordjevic I, Ryan W, Vasic B (2010) Coding for optical channels. Springer, Berlin Liu L, Wang Y, Li L, Zhang X, Wang J (2009) Design and implementation of channel coding for underwater acoustic system. In: IEEE 8th international conference on ASIC, ASICON’09. IEEE, pp 497–500 Intersil’s IC trans array HFA3096BZ96 datasheet, 11 Aug 2015 Lumileds LED luxeon rebel blue SMD LXML-PB01-0030 datasheet, 22 Jan 2017 Nexperia USA Inc., Trans npn prebias/pnp 6tsop PBLS6024D, datasheet, 14 Aug 2009 Ams’s IC light to voltage sensor 3 pin TSL254-R-LF-ND data sheet, 27 Jul 2016 Lin S, Costello DJ (2004) Error control coding. Pearson Education India, Bengaluru Blahut RE (1983) Theory and practice of error control codes, vol 126. Addison-Wesley, Reading Ghassemlooy Z, Wasiu P, Rajbhandari S (2012) Optical wireless communications: system and channel modelling with matlabÒ. CRC Press, Boca Raton Oubei HM, ElAfandy R, Ng TK, Alouini MS, Ooi BS (2017) Performance evaluation of underwater wireless optical communications links in the presence of different air bubble populations. IEEE Photon J 9:1–9
