Performance Analysis of Concatenated Convolutional and Reed Solomon

Coding with Interleaving in Optical Communication Channels

Ningombam Devarani Devi U. Sripati Acharya

Department of Electronics & Comm. Engineering Department of Electronics & Comm. Engineering
National Institute of Technology Karnataka National Institute of Technology Karnataka
Surathkal, Mangalore-575025 Surathkal, Mangalore-575025, INDIA
Mail Id : [email protected] Mail Id: [email protected]

Abstract—: For reliable and error free transmission of data in II. CODE DESCRIPTION
communication systems we require a system employing
forward error correction schemes. This paper analyze the Block diagram of the downstream channel is shown below,
performance comparison of Convolution code , Reed-Solomon
code as well as concatenation of Convolution code and Reed-
Solomon code over optical communication links with
interleaving in terms of their probability of Bit error rate and
signal to noise ratio with different code rate and identify with
code rate will give best performance .

Keywords— Convolutional Coding (CC), Reed-Solomon

Coding (RS), Quadrature Phase Shift Keying (QPSK),
Concatenated Codes, RS-CC, CC-RS.
Fig. 1.Theoritical block diagram of RS-CC coding.

I. INTRODUCTION A. Convolution Code

In optical communication systems, high bit error rates In our study, we have chosen a convolutional code which
require employing forward error correction (FEC) while is designed to provide several rates by the use of
transferring data. In satellite, space and optical communication puncturing. Consider a punctured Convolutional code of
system removing error is not a mare task. Even though optical
length 7 and rate 0.5 which is specified by parameters (7,
communication system provide high data rate, high
bandwidth, resistance to electromagnetic interference etc. but 1/2). The puncturing pattern is denoted by p and
in order to reduce the error due to transmission channel, we generator vectors is denoted by G. when G=1, if
have to choose a proper coding scheme. An implicative puncturing is done and equals 0 when puncturing is
enhancement in the performance aspect if the codes are deleted. The most frequently used punctured patterns of
properly chosen and decoding techniques are efficiently the Convolutional code are given in table below [1],
applied. In order to appropriate channel codes, for different
applications, we have to consider factors like transmission Table.1 Punctured Convolutional codes
channel characteristics, decoding performance etc.
In this paper we are analyzing the BER performance of the
Concatenation of two codes i.e. Reed-Solomon code and
Convolutional code using Viterbi decoding scheme by
considering half rate punctured Convolution code as the inner
code and shortened Reed-Solomon code as the outer code.
Puncturing phenomenon is done by periodically deleting
certain symbols from output stream to increase the code rate
of the encoder.

978-1-4799-8047-5/15/$31.00 2015
c IEEE 692
 B. Reed - Solomon Code routed into the branch corresponding to the null delay i.e. “0”
Consider shortened Reed Solomon (204,188). Here encoder of the interleaver. Same principle will be for de-interleaver.
process is done on every 188 bytes coordinate with t=8 i.e. we
can correct 8 enormous byes per transport packet. Here we are III.MODELING OF CHANNELS
using parity bytes to the packet which is going to transfer to
give the codeword. We are using 16 parity bytes to give 204 The channels models employed in FSO communication links
byte codeword. By using this shortened code, the amount of are briefly described in this section.
redundancy per information symbol increases because of this .
the amount of energy per information symbol is decreased.
A. Log-normal fading channel
Consider the following generator polynomials for Reed-
Solomon code,
If a random variable X is said to be log-normal distribution,
Code Generator Polynomial:
then the random variable Y = ln (x) is Gaussian distributed.
The log-normal circulation is applicable to standard
where. β=02 H continuous random quantities when the circulation is believed
(1) to be biased, such as certain proceeds and lifetime fickles. The
Field Generator Polynomial: log-normal fading probability density function is derived from
the first order Rytov approximation, where the logarithm of
(2) the power fluctuation can be approximated by the Gaussian
distribution. The probability density function (PDF) of the log-
normal distribution is given as,
C. Interleaving

Interleaving is a process which is based on an approach called

Forney approach. In this we are considering maximum
interleaving intensity, I=12. Each interleaved framework
composed of extending error protected envelope and (3) 
synchronization is bound by this action. Interleaver is
composed of I sections which is constantly connected with the
given bit stream by the selector. Here we followed the For spherical wave,
condition of First in First out (FIFO). So each branch is
connected in the FIFO shift register format with a depth of (R)
cells. The shift register cells contain 1 byte, and the input and
output selectors is synchronized.

Where, R= L/I, L=error guaranteed frame length (204)

I=maximum interleaving depth (12)
(4)
where, D is the receiver aperture diameter and
λ is the wavelength.

(5)

where, k is the wave number and 

Rytov Variance is given by.

Fig. 2 Convolutional Interleaver


(6)
For synchronization phenomenon, the bytes which used for
synchronization and the corresponding bytes should be always

2015 IEEE International Advance Computing Conference (IACC) 693

 B. Log –normally modulated exponential (LNME)
Fading channel
The LNME fading pdf model is a doubly stochastic model. In (10)
this fading, the pdf model (I|a) is assumed to follow the where, n= length of the Reed-Solomon code = 204 and 
exponential distribution and (a) is assumed to follow the log- = error correction capability of the code = 8.
normal distribution and given as, The interleaving is use between the inner code and outer code.
The common decoder for this concatenated scheme consists of
Viterbi algorithm which gives hard decisions in output. The
inner code has G = 177 and 133 in octal format and the outer
code is Reed-Solomon code which can correct both random
and burst errors. The fundamental RS code (N=255, K=239,
t=8) is shortened to support variable block size.
V. SIMULATION RESULTS
  The  performance
DE
of F 
the concatenated 
Reed Solomon and
Convolutional code with interleaving and without interleaving
= logarithmic variance of the modulation are shown,
and this variance is given as

is the variance of the normalized laser beam.

IV. ANALYTICAL PERFORMANCE
The performance of the Convolution code is approximately
consider by abbreviating the union bound after a symbolic
number of condition and is given as below,

Fig.3. Performance for concatenated RS-CC without interleaving.

From the above simulation results, we analyzed the

(8) performance improvement of the concatenated RS-CC code
Where, k is the number of encoder’s input bits, = free with the decrease of code rate.
distance, = total number of bit errors. This is occur in all
incorrect paths in trellis which alters from the correct path for
d positions. Here we use one significant term N for the
calculation purpose. = probability of choosing an incorrect
path. In this analysis we are assuming the interleaving reed
Solomon decoder and the Viterbi decoder is long enough to
achieve the phenomenon of breaking up of long burst errors
out of Viterbi decoder. So by union bound condition the
symbol error probability for Reed-Solomon is as follows [5],

For coherent QPSK modulation scheme, is given as below

[3],
Fig.4 Performance for concatenated RS-CC with interleaving.
(9)
The performance of the concatenated Code is much
where, , = signal to noise ratio per bit information improved by using the interleaving phenomenon. From the
and = overall code rate of the concatenated RS-CC. above comparison, concatenation with interleaving has
Therefore the overall BER out of Reed-Solomon decoder is around 4.5 dB improvement than the concatenation without
given as follows [4], interleaving code.

694 2015 IEEE International Advance Computing Conference (IACC)

 Table 2. BER Comparison with different code rate and free distance.

 
Code Free Total no. of bit errors in
Rate distance, all incorrect paths,
for BER= for BER=
RS Decoder output RS Decoder output for BER= RS
Decoder output
Without interleaving Without interleaving With
interleaving

1/2 10 {36,0,211,0,1404,0, 24.010 21.03 19.33

11633,0,77433,0,
502690,0,3322763,0,
21292910,0,1343659,0}

2/3 6 {3,70,285,1276,6160, 30 24 23.5

27128,117019}

3/4 5 {42,201,1492,10469, 32.5 26.5 27.5

62935,379644}

5/6 4 {92,528,8694,79453. 33.4 28 31

792114}

7/8 3 {9,500,7437,105707, 39 32.4 35.5

1402743}

REFERENCES
Interleaving phenomenon contribute to improve performance
of forward error correcting codes. As a result, concatenation of
Reed- Solomon and Convolution coding scheme with [1] J.Conan, “The Weight Spectra of Some Short Low-Rate
interleaving is appropriate for the downstream channel Convolutional codes”, IEEE Trans. On Commu., p.1050-1053,
frequency division system and it is popularly used in satellite Vol. COM-32, No. 9, Sept 1984
communication, such as in digital television broadcast [2]John M. Senior, “Optical Fiber Communications”, Third
systems. But the dependence of reduce in decoding probability Edition
error upon the difference in code rate and channel capacity is [3] J. Proakis, Digital Communications, Second Edition,
still in discussion. McGraw-Hill, 1989.
[4] A. Lapidoth, “On the Probability of Symbol Error in
CONCLUSION Viterbi Decoders”, IEEE Trans. On Commu., p. 152-155, Vol.
45, No.2
[5] R. Cideciyan, E. Eleftheriou, and M. Rupf, “Concatenated
In this paper we analyzed the BER performance characteristics Reed-Solomon/Convolutional Coding for Data Transmission”,
of the Convolutional code, Reed-Solomon code and IEEE Trans. On Commu., p. 1291-1303,Vol. 45, No. 10
concatenated Reed-Solomon/Convolutional code using
interleaving over optical communication channels. From our [6] Changlog Ju,”Decoding Algorithm for concatenated Reed-
analysis we verified that as compared to individual error Solomon/Convolutional code”. IEEE Trans on Commu.p.
correcting capability, concatenated code have better 1550-2230
performance. After interleaving the concatenated codes we got [7] S. Piazzolla, A. Biswas, “Deep Space optical
around 10 dB improvements than the uncoded information. Communication Downlink Budget, System Parameters”, IEEE
Trans on Commu., Vol. 2. No.42,

2015 IEEE International Advance Computing Conference (IACC) 695