QOS guaranteed multimedia transmission over wireless networks

Wireless multimedia communication has been extensively investigated in the past decades due to the rapid development of wireless devices and explosive demand from mobile users on high quality multimedia services. Despite recent advances, it is still challenging to provide media of pleasing quality for various wireless multimedia applications. In this dissertation, we develop a suite of novel algorithms and schemes to integrate application layer media representations with both physical layer (PHY) and media access control (MAC) layer solutions.;First, we design a novel multi-input multi-output (MIMO) multimedia transmission with spatial multiplexing configuration. Existing MIMO-based approaches attempted to match the status of each sub-channel with suitable power allocation. This is impractical in two aspects: RF circuits complexity and unequal importance of multimedia bit-streams. We develop a creative solution through the design of novel precoding that serves dual purpose: embedded power allocation and intrinsic matching of source and channel characteristics. This precoding scheme is shown to be capable of achieving enhanced quality-of-service (QoS) performance for MIMO multimedia services. This novel pr-coding scheme is further extended for fast moving MIMO terminals in order to meet the new needs in next generation mobile systems in which the terminals are no longer restricted to still or slow motion applications. To this end, we develop a blind constant modulus algorithm (CMA)-based channel equalization (CMACE) scheme to track the fast time-varying channel in order to assure the service quality of multimedia consumer with high moving speed.;Then, we turn our research focus to addressing technical challenges resulting from existing MAC layer solutions for 802.11 wireless systems that often lead to over-protection of multimedia (video) data by adopting virtual contention-free schemes. Such aggressive schemes achieved improved QoS for video transmission at the expense of best effort and background traffics within the same mobile systems. Aiming at overcoming the drawbacks of the existing approaches, we develop a deadline-aware virtual contention free (DVCF) scheduling algorithm for video over IEEE 802.11e wireless networks. This scheme jointly considers the delay profile of the video source, the prioritized 802.11e protocol, and network specifications by adjusting the parameters of the 802.11e protocol to achieve well-balanced transmission of all types of data traffic.;Next, we attempt to resolve the new challenges of multimedia applications arising from newly deployed LTE networks. In particular, the standardized hard hand-off (HO) procedure in LTE networks poses significant barriers for moving receivers to maintain multimedia QoS due to potential service interruption introduced by this procedure. In this dissertation, we develop a novel downlink scheduling scheme for video streaming that considers both QoS metrics derived from video packet deadlines and possible service degradation resulting from hard HO. The pro-posed scheduler is not only able to achieve satisfactory QoS for users during the HO period, but also guarantees fairness for both multimedia traffic and regular data traffic.;Finally, we exploit a futuristic scenario in which device-to-device (D2D) communication can be evoked as an underlay in LTE networks. Although LTE cellular infrastructure was designed to support broadband multimedia communications, it is still challenging to accommodate the ever increasing data traffics into the limited licensed spectrum due to explosive increase in mobile users. It is possible to create an additional transmission opportunity for multimedia transmission when the two communicating terminals are within the neighborhood of each other. We develop a new framework for D2D communication as an underlay in LTE networks to establish a direct link between them by designing a clever interference avoidance strategy. It has been illustrated that the total achievable data throughput for LTE networks can be dramatically enhanced by the proposed D2D communication scheme. (Abstract shortened by UMI.).