We've Moved!
Visit SDSU’s new digital collections website at https://digitalcollections.sdsu.edu
Description
Digital video technology has evolved from a low resolution (SD) to very high resolution (ultra HD or 4K) in this decade. Many digital multimedia devices, such as smartphones, smart television and tablets, can now support high resolution video content over wireless mediums. The main challenges in supporting the high resolution mobile video applications over wireless medium are the limited and variable channel bandwidth, and channel induced packet losses. Due to limited available channel bandwidth, the video streams are highly compressed by using the video compression technology, such as the recently developed High Efficiency Video Coding (HEVC) standard. However, the compressed video bitstream is very sensitive to the packet losses during transmission, which results in distorted video at the user end. In this thesis, we have designed a new scheme for transmitting HEVC video bitstreams over wireless channels by using unequal error protection (UEP). In UEP scheme, the HEVC video packets are divided in four priority classes, based on the distortion their loss would introduce to the video quality. In the next step, a suitable low density parity check (LDPC) based forward error correction (FEC) code is applied to each packet based on its priority class, video bit rate, available channel bandwidth and channel SNR, by using a genetic algorithm based optimization which minimizes the video distortion. The LDPC FEC codes are used as they give error-free performance near Shannon’s channel capacity limit. Effectively, the higher (lower) priority packets are transmitted with stronger (weaker) FEC protection. Finally, the error concealment schemes (including the temporal error concealment (TEC) and motion compensation error concealment (MCEC) algorithms are used to conceal the effect of lost video packets at the receiver. The use of UEP and error concealment schemes provide high video quality in the presence of channel losses. It validates our primary objective which was to better utilize the available channel bandwidth to improve the video quality.
