A Master of Science Thesis in Electrical Engineering submitted by Rimas Adnan Zrae entitled, "Joint Adaptive Modulation and Channel Coding for Wireless Multimedia Communications," submitted in June 2011. Available are both soft and hard copies of the thesis.
Wireless communication has witnessed a signi cant growth due to recent ad- vances in technology. As a result, there has been a strong demand for various multi- media services and applications over wireless channels. Typically, multimedia data is large in volume with high bandwidth requirements; therefore multimedia communi- cations over unreliable and bandwidth limited channels has been always a challenge. In addition, the compressed multimedia data is highly sensitive to information loss and channel bit errors. Transmission errors can cause decoding failure which will distort the reconstructed image or video. Forward error correction (FEC) techniques could be employed to reduce the e ects of errors on the quality of the reconstructed multimedia information. However, these error correction schemes in ict redundant bits on the transmitted data resulting in a loss in the bandwidth e ciency. Thus, the challenge resides in protecting the compressed multimedia data to ensure reli- able transmission against hostile channel conditions while having less impact on the bandwidth e ciency. In this thesis, we outline a new approach to tackle the problem of reliable image and video transmission over wireless channels. We propose an adaptive trans- mission scheme that attempts to meet the available channel conditions through the adaptation of the modulation modes and channel coding rates. The proposed scheme partitions the compressed image and video data into di erent bit streams with un- equal importance as perceived by the human visual system. Then, it jointly considers the estimated channel conditions and the importance of the transmitted bit stream when deciding on the modulation mode and the coding rate. The proposed scheme attempts to meet a prede ned target bit error rate (BER) and yet avoids over pro- tection, hence it improves the bandwidth e ciency which consequently increases the e ective transmission rate. The proposed scheme then decides on the transmission parameters which meet the prede ned target BER using an o -line lookup table. This prede ned BER is based on the sensitivity of the transmitted bit stream. The results show that image transmission over Rayleigh at fading channel using the pro- posed adaptive modulation scheme achieves peak signal-to-noise ratio (PSNR) values slightly less than those of a xed-modulation system resulting in an insigni cant im- pact on the subjective quality of the reconstructed images. However, the spectral e - ciency of the proposed scheme is improved up to 460%. Results also demonstrate that combining adaptive modulation and channel coding for image transmission provides a graceful trade-o between image quality and spectral e ciency. For video trans- mission, a modi ed version of the proposed scheme shows signi cant improvement in the spectral e ciency while maintaining an average PSNR value of 32 dB across the transmitted video frames with an acceptable perceptual quality and smooth playback. Furthermore, the overall performance of the proposed scheme degrades gracefully in the presence of imperfect channel estimation. In future work, the playback bu er occupancy could be considered along with the channel condition and the importance of the bit stream to reduce bu er starvation instants.