Description
ABSTRACT
Error control codes are widely used in almost all digital systems as they provide a method for dealing with the unknown like noise. This research investigated the role of reducing error rates in image transmission over 3G systems using convolutional coding technique in MATLAB. The error correction code employed was the convolutional error correction codes. The performance of the codes are evaluated based on key performance indicators like Bit Error Rate (BER), number of symbols or data compared and number of errors detected. For the verification of proposed approach, computer simulation results are included. The results show a comparison of the performance terms of their Bit Error Rate (BER) of convolutional code with different code rate ( ½ and 1/3 ) used. Based on the results, between 60% and 65% improvement on coding was achieved between reference points of 10-2 and 10-4 respectively for the two code rates. The results also show that as the Bit Error Rate (BER) decreased, the coded system can transmit data signals with at least 3dB less power, so making the performance of the coded system better than the uncoded system.
TABLE OF CONTENTS
COVER PAGE
TITLE PAGE
APPROVAL PAGE
DEDICATION
ACKNOWELDGEMENT
ABSTRACT
CHAPTER ONE
- INTRODUCTION
- BACKGROUND OF THE STUDY
- PROBLEM STATEMENT
- AIM OF THE STUDY
- OBJECTIVE OF THE STUDY
- SCOPE OF THE STUDY
- PROJECT MOTIVATION
- SIGNIFICANCE OF THE PROJECT
- PROJECT ORGANISATION
CHAPTER TWO
- LITERATURE REVIEW
- INTRODUCTION
- DEFINITION OF AN ERROR
- TYPES OF ERRORS
- TYPES OF ERROR DETECTION
- REVIEW OF RELATED STUDIES
- SHANNON LIMIT
- ERROR CORRECTING CODES
- ERROR-CORRECTION CODE SELECTION
- TYPES OF ERROR CORRECTION CODES
- OVERVIEW OF CONVOLUTION CODES
- PROPERTIES OF CONVOLUTION CODES
CHAPTER THREE
3.0 METHODOLOGY
3.1 INTRODUCTION
3.2 IMPLEMENTATION OF CODES
3.3 ENCODER REPRESENTATION
3.4 ENCODER STRUCTURE
3.5 DECODING CONVOLUTION CODES
3.6 PERFORMANCE MEASURE
3.7 PERFORMANCE ANALYSIS
3.8 SIMULINK BLOCKS
CHAPTER FOUR
4.0 RESULTS AND ANALYSIS
4.1 SIMULATION RESULTS
4.2 RESULTS DISCUSSION
4.3 SIMULATION PLOTS
4.4 SIGNIFICANCE OF RESULT
CHAPTER FIVE
- INTRODUCTION
- SUMMARY
- LIMITATION
- CONCLUSION
- RECOMMENDATIONS FOR FURTHER WORK
- REFERENCES
APPENDIX A: Simulation results for 100 data points
APPENDIX B: Generated MATLAB®2008 Codes
ABBREVIATIONS
AWGN Additive White Gaussian Noise
BCH Bose-Chaudhuri-Hocquengham code1
BER Bit Error Rate
bpp Bit per pixel
BPSK Binary Phase Shift Keying
BSC Binary Symmetric Channel
ECC Error-Correcting Code
GF(2m) Galois Field with 2m elements
JSCC Joint Source-Channel Coding
KLT Karhunen-Loe`ve Transform
MAP Maximum a posteriori
ML Maximum Likelihood
MSE Mean Squared Error
SNR Signal-to Noise Ratio
RS Reed-Solomon code
UEP Unequal Error Protection
CHAPTER ONE
1.0 INTRODUCTION
The last thirty five years have seen a dramatic change in the way communication is achieved around the world. Wireless communication has evolved from being an expensive and rare technology for the few in the 70’s to becoming a wide spread and economical means of facilitating commercial as well as public service communications. One of the majors reasons for the continuous growth in the use of wireless communication is its increasing ability to provide efficient communication links to almost any location, at constantly reducing costs with increasing power efficiency (Jemibewon, 2000).
Wireless communication is one of the most active areas of technological development. This development is being driven primarily by the transformation of what has been a medium for supporting voice telephone into a medium for supporting other services such as transmission of video, images, text and data etc (Wang, 2003). Basically, a communication system deals with information or data transmission from one point to another (Du, 2009). Over the years, there has been a tremendous growth in digital communications especially in the fields of cellular, satellite and computer communications. In these communication systems, the information is represented as a sequence of binary bits. The binary bits are then mapped (modulated) to analog signal wareforms and transmitted over a communication channel. The communication channel introduces noise and interference to corrupt the transmitted signal. At the receiver end, the channel corrupted transmitted signal is mapped back to binary bits. The received binary information is an estimate of the transmitted binary information (Huang, 1997). Normally, during signal transmission through noisy channels errors can be detected and corrected using coding techniques (Huang, 1997). Noise is any undesired signal in a communication circuit. Noise can also be unwanted disturbances supper imposed on a useful signal, which tends to obscure its information content.
1.1 BACKGROUND OF THE STUDY
In the beginning of the third millennium, the growth of communication systems deeply affects human societies. Lots of data flow through net- works of various types set up all over the planet. With the emergence of multimedia applications, several types of data have to be transmitted, including text, speech or sound, and images, which are at the center of this work. All these data types have undergone the digital revolution, which enables a rich set of storage and processing techniques. More and more image sources are available, be they fixed or moving, computer-generated or based on sampled analog pictures. Television, movies, satellite observations, medical images or camera pictures are a few examples of applications linked to image transmission and processing.
Transmission channels are available in lots of different types, wired or wireless, packet or connection-oriented. TV-cables, telephone networks or power lines are expanding from their initial roles into large-capacity transmission channels. They all have their own characteristics in terms of available bandwidth and quality of service. Mobile applications offer more and more services, and image transmissions will truly appear in the third generation of mobile phones. Transforming all these channels suffering from various impairments into efficient transmission media requires a lot of signal processing.
In order to conquer the corresponding markets, communications engineers have to design systems always facing the same constraints: band- width and power. Bandwidth limitations impose an efficient compression of transmitted data. This is particularly crucial for digital images, which can be of very large size. All the corresponding techniques are known as source coding. Limitations in power determine the ability for useful signals to be recovered above noise, generally with a certain probability of errors or other impairments. Interfering signals from other sources or distortions caused by the nonideal response of the channel can also cause errors. The way to cope with them is called channel coding.
Emitter
Figure 1: Schematic representation of an image transmission chain including source and channel codes.Receiver
A typical image transmission system contains the elements depicted in figure 1, implementing the two coding steps we have just mentioned; the picture successively undergoes source and channel coding. The source coder is responsible for removing the redundancy of the picture, in order to lower the bit rate required to transmit it. Classically, this coder begins with a decorrelating transform, whose role is to remove the correlation between adjacent pixels by expressing them in a space based on the eigenvectors of the statistical distribution of the pixels, or close to these vectors. This can be proved to reduce the required bit rate to its minimum value. Then, a conversion from real values to bits has to be achieved. This consists in first quantizing the coefficients we have to code, i.e. expressing them into integer multiples of a base step, and secondly entropy coding them. This last step also contributes to reducing the bit rate, by taking into account the fact that some values are more frequent than others, e.g. values close to zero are more frequent in case of zero-mean distributions. By assigning shorter code words to these values, we reduce the average bit rate.
After all these source coding steps, we get a flow of bits to be transmit- ted through the channel. If this channel was perfect, we could simply put the bits on it, get them back at the channel output, and use them in order to recover the source picture. However, the channel generally introduces errors. In order to prevent them, we have to use what we call channel coding. The idea is somewhat opposite to source coding, as here we introduce redundancy. However, this redundancy is structured in such a way that we can use it to correct channel errors. Of course, the more redundancy we introduce, the more errors we can correct. We can already see that adapting the channel code we use to the channel characteristics will be an important issue in the design of a transmission system.
After receiving bits at the output of the channel, we have to implement decoders in order to reconstruct the transmitted picture. The channel de- coder will remove the redundancy introduced by the channel code and use it in order to correct channel errors. We will see later that different kinds of channel decoders exist for a given code, depending on the avail- able computing power in order to correct as many errors as possible. Then, the hopefully almost error-free bit stream will be used by the source de- coder in order to rebuild the picture. Basically, this decoder simply inverts the operations achieved by the source coder. However, when facing errors, we can try to develop more efficient strategies. This includes using some knowledge about the general shape of a picture in order to recover erroneous coefficients.
Both source and channel coding problems have led to many results and coding schemes. Behind them, Shannon’s information theory provides bounds on their performances [71]. This theory also tells us that both coding problems can be processed separately, a pretty nice result as both problems are already hard to tackle on their own. However, Shannon’s result only holds under ideal conditions of infinite data size and computing power. In practical systems, limitations on size and complexity make Shannon’s bounds unreachable. Better solutions are generally obtained by combining both coding problems, giving rise to the so-called joint source- channel coding philosophy. Its aim is to provide solutions not only to a given source or channel coding problem, but to a more general application including both subsystems. Such a way of designing transmission systems has led to better performances with respect to traditional design, by taking benefit from the mutual influences and relationships between source and channel coders and decoders.
In this thesis, we want to address some of the possible interactions between source and channel coders in a global system. A key issue is how to select a channel coding scheme for hierarchical image transmission, adapted to the type of data transmitted. Different parts of the data can affect more or less the received picture by being more or less important in the image representation, or more or less sensitive to errors. Of course, this requires a sufficient knowledge of the performances achieved by the available coding schemes, in order to accurately match the level of protection to the significance of the source data.
1.2 STATEMENT OF PROBLEM
Various researches have been carried out with a view to securing and optimizing the usage of wireless communication systems. With bandwidth requirements in communication systems, the system capacity is depending on the level of interference that can be tolerated, and it is inversely proportional to the signal to noise ratio (SNR) (Qian, 1999). In other words, the system capacity increases as SNR decreases. In order to reduce SNR, error correction codes need to be integrated into the system. Error correction codes decrease SNR by introducing the coding gain for the communication link. The coding gain measures the amount of additional SNR required to provide the same BER performance for an uncoded message signal (Qian, 1999).
In 1948, Shannon proved that for a band limited Additive White Gaussian Noise (AWGN) channel, with bandwidth B, there exist families of coding schemes that can achieve arbitrarily small probability of error at the receiving end at a communication rate less than the capacity of the channel, which can be described as follows:
Where C represents the capacity of the channel, S and N are the average signal power and noise power respectively. The implication of equation 1.1 is that if information rate can be dropped below the capacity of transmission channel, with proper error protection means, such as error correction codes, error free transmission is possible. In this research, error correction is achieved using convolution codes.
- AIM OF THE STUDY
The aim of this study is to stimulate error in image transmission over the 3g network using different types of error control codes.
1.4 OBJECTIVES OF THE STUDY
The main objectives chosen for this thesis are highlighted as follows:
- Understand the concept of error correction in wireless communication
- Implementation of error correction codes in a noisy channel using convolution codes to compensate transmission impairment in order to increase quality of
- Creation of the experimental simulation model of a communication system
in MATLAB®2008/Simulink to oversee the process of error correction implementation.
- Investigate the performance of the chosen codes (convolution codes) with different code
1.5 SCOPE OF THE STUDY
Digital communication systems have played a vital role in the growing demand for data communications. In communication systems when data is transmitted or received, error is produced due to unwanted noise and interference from the communication channel. For efficient data communications it is necessary to receive the data without error. Error-control coding technique is to detect and possibly correct errors by introducing redundancy to the stream of bits to be sent to a channel.
1.6 SIGNIFICANCE OF THE STUDY
One of the goals of wireless communication systems is to deliver information or data efficiently and reliably comparable to wired systems, at reasonable costs and convenience to as many users as possible. A system with good performance should be able to deliver clear, uncorrupted data with very little latency and minimal power consumption. In digital systems, the ability to achieve uncorrupted data transmission is directly proportional to a system’s probability of bit error, which is in turn inversely proportional to the power of data transmission. One of the major drives in wireless communication is the effort to increase power efficiency, thus delivering reliable links at lower power levels. Many practices have been developed to assist in the achievement of higher power efficiency, one of which is the implementation of error-correction codes.
1.7 PROJECT ORGANISATION
The work is organized as follows: chapter one discuses the introductory part of the work, chapter two presents the literature review of the study, chapter three describes the methods applied, chapter four discusses the results of the work, chapter five summarizes the research outcomes and the recommendations.
CHAPTER TWO
2.0 LITERATURE REVIEW
2.1 INTRODUCTION
In digital systems, the analog signals will change into digital sequence (in the form of bits). This sequence of bits is called as “Data stream”. The change in position of single bit also leads to catastrophic (major) error in data output. Almost in all electronic devices, we find errors and we use error detection and correction techniques to get the exact or approximate output.
Reviews
There are no reviews yet.