Digital Image transmission using enhanced turbo codes

Other Title(s)

إرسال الصور الرقمية باستخدام الترميز التعاقبي المحسن

Dissertant

al-Muslih, Ziyad Talal Salah

Thesis advisor

Bani Ahmad, Umar

University

Princess Sumaya University for Technology

Faculty

King Abdullah II Faculty of Engineering

Department

Department of Electrical Engineering

University Country

Jordan

Degree

Master

Degree Date

2018

English Abstract

Turbo codes have received a lot of attention for its ability to provide low error rates over many wireless applications resulting in excellent data quality and high data rate systems.

This ability of turbo coding arises from the fact that the codec system compromised of recursive systematic parallel or serial concatenated codes, resulting in an increase of the code hamming distance.

In addition, on the receiver side, the decoder implements iterative soft- input / softoutput decoding allowing the data to be decoded more than once.

In turbo code structure using convolution coding, two decoding algorithms are used, Soft Output Viterbi Algorithm (SOVA), and log Maximum A Posteriori (Log-MAP).

It is well known that the Log-MAP algorithm outperforms the SOVA algorithm, however, the MAP algorithm requires more system resources as compared to the (SOVA) algorithm.

In both cases, the data is processed as frames and the coding gain is mainly determined by the frame size, the code generating polynomials, and the number of decoding steps (number of iterations).

In this thesis, we propose a new coding scheme (enhanced turbo code), which is compromised of a hamming code followed by the classical turbo code structure.

The component codes used in the new structure are two convolution codes connected in parallel with generating polynomials (5,7), separated by 2048 bit random interleaver.

The proposed system was then used to transmit three different image signals over additive white Gaussian channel, where the transmission of the different image signals was tested over two coloring schemes prior to transmission, while the second case considered transmission of compressed image signals.

The compression routine used is this thesis is a three level wavelet decomposition followed by a uniform quantizer and long run encoder.

For each of the three compressed images, the image signals were processed using three quantization levels: 50,100, and 150, which resulted in different compression ratios for each of the quantization levels.

The received data was decoded using the new proposed decoding structure with 2048 data frame size and three decoding steps using the MAP algorithm.

In addition, for both cases of image transmission, the decoded data was filtered using a median filter and a slight gain in peak signal to noise ratio was attained for the different received images as compared to decoding the received image without filtering.

To evaluate the proposed system performance, the three different images were transmitted using the new proposed system, and it was found that the proposed system offers an average improvement of 0.8 dB in the signal to noise ratio and about 6 dB improvement in the peak SNR as compared to the classical turbo code structure.

However, the actual improvement in the peak to signal ratio of the different images depends on the nature of the colored signal and the processing type used.

Finally, for Eb⁄No= 1 dB, an average peak signal to ratio of 33 dB and 37.5 dB, were found for the HSI and RGB coloring methods, and 30 dB for a compression ratio of 29.1.

Main Subjects

Electronic engineering

Topics

• Digital images

No. of Pages

73

Table of Contents

• APA
• MLA
• AMA

Language

English

Data Type

Arab Theses

Record ID

BIM-833096