| Human desire to access information and to communicate with anyone at anytime and anywhere is the driving force behind the tremendous research efforts in the area of multimedia distribution in wireless networks. In this thesis we study some challenging issues involved in this area. Particularly, we investigate resource management issues in energy saving and Quality of Service (QoS), low complexity embedded image/video coding algorithms and complexity scalability design, and protection in multimedia distribution.; First, we address the resource management issue from the perspective of minimizing total power consumption of mobile terminals with a guaranteed QoS. We establish a general framework to model the dynamic interaction between transmission and processing power of a mobile terminal. The optimal tradeoff of transmission and processing power guarantees minimum total energy consumption. To achieve the energy saving goal, the multimedia coders used in our system have low computational complexities, support multimode operations, and provide a run-time complexity that is inversely proportional to the compressed data rate. We then develop two technologies, Multimode Adaptive Power Saving (MAPS) protocol and Multistage Coded Modulation (MCM) scheme, to integrate our energy saving strategy into current wireless networks and to make mobile communication systems be energy-saving and QoS-ensuring.; Next, we design a low complexity coding algorithm called MultiGrid Embedding (MGE) coding. We then use it to transcode and compress JPEG or MPEG bitstreams to fully embedded ones. This provides many advantages to media distribution across the Internet and wireless networks. We also build a low bit rate image coder based on sigma filtering and MGE coding. Afterwards, we construct a tri-mode complexity scalable coder by combining an MGE based coder, a JPEG coder, and a sigma filtered coder, to be used in our energy saving system presented above.; Finally, we deal with the data protection issue in multimedia distribution. We design a high capacity data embedding algorithm to self embed some important regions of an image into the image itself, If the image has been tampered with, one can detect and reverse changes in the image. |
