Advanced approaches in information transmission and access control for wireless communication networks.

摘要

Over the last two decades, wireless communication has seen tremendous growth and has been demonstrated as a robust voice and data transport mechanism. New wireless communication methods and services are enthusiastically adopted by people throughout the world. Driven by the ever increasing demand on high speed wireless multimedia services, development of highly reliable and more efficient wireless communication networks has become the ultimate goal of the research community. In this dissertation, we focus on efficient information transmission and medium access control (MAC) in wireless networks. More specifically, we aim to improve the reliability and efficiency of the wireless networks through the following research thrusts:; First, we investigate the channel tracking scheme in time-varying environments. Accurate channel estimation is essential in ensuring reliable information transmission. However, both time and frequency dispersions in mobile wireless channels cause significant challenges in channel estimation. Conventionally, for time-varying channels, pilot (training) signals are periodically transmitted to achieve accurate channel estimation. Such a scheme is not spectrally efficient due to the considerable overhead signals. In this dissertation, we propose a semi-blind approach to efficiently estimating fast fading channels and jointly detecting transmitted signals. The proposed scheme is shown to have high spectral efficiency and performance reliability. Additional efforts are devoted to studying the extreme case when there are unpredictable abrupt changes in the channel. We propose algorithms to detect such abrupt changes and suppress the possible error propagation. The proposed algorithms are demonstrated to be effective through simulations.; Next, we explore the MAC protocol design by taking into account the physical (PHY) layer channel capability. In conventional medium access control protocol designs, the physical layer is simply characterized using a binary collision model. Although the model provides a tractable path for network performance analysis, it fails to reflect the physical layer channel capability. Cross-layer medium access control protocol design, which exploits the physical layer signal processing capabilities for MAC performance improvement, has attracted considerable research attention. In this dissertation, taking a mutually interactive perspective, we propose to design an MAC protocol, named hybrid ALOHA, which is in favor of the physical layer, and the improved physical layer, in turn, improves the MAC performance in terms of throughput, stability and delay behavior. Both theoretical and simulation results show that significant performance improvement can be achieved by hybrid ALOHA, in comparison with traditional ALOHA.; Finally, from a mixed analog-digital perspective, we investigate a system with analog inputs and propose a source-aware information transmission scheme to minimize the average input-output distortion. After sampling and source coding, the digital bits could have different levels of significance, i.e., some bits could be more important than others. Such non-uniformity in source coding then calls for nonuniform information transmission to improve system performance. An unequal error protection scheme is proposed to minimize the average distortion. Simulation results demonstrate its effectiveness. Furthermore, a joint quantization-constellation design is investigated under the criterion of distortion minimization. The proposed method generalizes the concept of constellation design from the perspective of joint source-channel coding. The simplicity and power efficiency of the proposed schemes make them particularly attractive for systems with tight power constraints, such as wireless sensor networks and space communications.