Error Tolerant Video Streaming in IEEE WLAN

Abstract

Broadband network proliferation at higher bit rates has spurred a host of multimedia streaming applications like internet video surveillance, live broadcast of international events, distant learning and entertainment etc. Wireless networks have been recognized as the most popular networking technology due to their flexible infrastructure, portability and easy deployment. Today's wireless internet user is truly global and can access content free of space or time constraints. But the caveat is that efficient streaming of multimedia contents over wireless links has proven to be an open research problem. In-time packet delivery and successful play out on receiver end are stringent quality requirements of any bandwidth intense, delay-sensitive real time multi-media application. However, wireless link characterized by fluctuating channel bandwidth and route obstacles results in transmitted signal attenuations or loss thus adversely affect the played back video quality. Ultimately, efficient streaming of multimedia contents over error prone wireless links has emerged as one of the most challenging problems of the current era of digital communication. However, applying unequal error protection strategies and avoiding unnecessary packet discard at various network levels yield valuable outcomes. In this research work, an error control mechanism has been proposed that enables a multimedia application to adapt the video bit rate gracefully with changing channel conditions based on intelligent network estimation. Intelligent link estimation triggers the rate adaptation on sender side through reducing the video bit rate by re-encoding at lower quantization scale. In addition, an error frame forwarding approach based on the idea of discriminating video streaming calls from the data packeting over IEEE WLAN channels through bit demarcation in network packet headers. Error computation at various network levels are evaluated and disabled in order to attain efficient utilization of channel data rates characterized by minimized re-transmissions and reduced delays with least error checksum computations and packet re-transmissions, increased throughput characterized by the higher number of packets available for decoding and enhanced multimedia visual quality due to gap elimination (appears as a consequence of some frame loss). Results have been supported through experimentation involving numerous video statistics as well as tools and techniques that can possibly be deployed in standard video codecs. Objective and Subjective quality assessments have yielded the encouraging outcomes for a mix of standard video clips (foreman, mother daughter, tennis) with variable mobility level (degree of variations in background and foreground frame contents). Reconstructed video assessment has further been improved through deploying perceptual quality tool with a rich set of parameters available for precise and complete QoE based video analysis. It has been observed that proposed error control strategy has outperformed the traditional error elimination techniques(particularly Automatic Repeat reQuest (ARQ)) for all Group Of Picture (GOP) structures of standard video codecs in terms of efficient utilization of available WLAN channel data rates even at higher Bit Error Rates (BERs). Selective re-transmission events have been incorporated for inevitable part of video bit stream without causing a noticeable drop in objective or subjective media quality. Moreover, proposed error frame forwarding approach exhibits higher link saving even at higher BER with acceptable visual quality.