Performance Enhancement in Heterogeneous Wireless Networks: Channel Assignment considering Switching Overhead, Query Processing using Event Signatures, and Uplink Traffic Analysis

As new Radio Access Technologies have been constantly developed and deployed with various emerging multimedia applications and multi-radio portable devices, the level of heterogeneity in wireless networks has been increasing. Since current wireless network technologies have their own unique characteristics and capabilities, radio resource sharing and information processing in heterogeneous environments is widely considered to be crucial in optimizing the network throughput and capacity. Furthermore, new trends in the use of the Internet due to the emergence of new services and changes in the propensity of mobile users make heterogeneous resource management problems increasingly difficult. In this dissertation, these three concerns are addressed in order to achieve significant performance enhancement in terms of network management and user satisfaction.;First, resource allocation and scheduling problems in multi-radio multi-channel Wireless Mesh Networks are addressed by considering the switching overhead incurred from switching radios dynamically from one channel to another. We explicitly model the switching delay that is incurred during channel switching and use that delay in the design of channel assignment algorithms. Both centralized and distributed channel assignment algorithms are provided. Performance of the developed channel assignment algorithms is analyzed through discrete-event simulations.;Second, the problem of information processing in Heterogeneous Wireless Sensor Networks is considered. As a powerful application domain of information processing, we consider the problem of identifying significant events using diverse sensors deployed in the area. We provide a mechanism by which sensors can exchange information using signatures of events instead of raw data to save on transmission costs. Further, we present an algorithm that dynamically generates phases of information exchange based on the cost and selectivity of each sensor filter. Simulation results show that the proposed algorithm detects events while minimizing the transmission and processing costs at sensors.;The new trend in wireless services is shifting from downlink-centric services to bidirectional and uplink centric services. Through the popularity of social networking services (e.g. Facebook, YouTube, and Flickr), we are observing an ever-increasing amount of user-generated content, also known as user-created content. This recent uplink traffic pattern is considered as a final problem in this dissertation. Live uplink traffic traces obtained by monitoring 3G networks of a mobile data service provider are analyzed. The results using six different self-similarity analysis algorithms suggest that this uplink traffic is self-similar. The impact of analyzed traffic characteristics on mobile data networks is evaluated in the WiMAX module available in OPNET software.;The contributions of this dissertation research lie in the area of radio resource management, distributed information processing, and new traffic pattern analysis in heterogeneous wireless networks. This work is the first to investigate the three crucial factors that limit network throughput and capacity, and analyze their impact on network performance in heterogeneous environments. Our consideration of switching overhead and use of sensory signatures are novel contributions and achieve significant performance enhancement in heterogeneous wireless networks.