Design and performance evaluation of qos-oriented wireless networks

According to CTIA's Semi-Annual Wireless Industry Survey, the U.S. wireless industry has grown from just under 350,000 subscribers in 1985 to currently more than 137,000,000 subscribers (a growth of more than 39,000%). The wireline networks are experiencing increased demand for better, faster, and easier connectivity to the Internet through broadband services such as cable modems and digital subscribers lines (DSL). Due to the exponential penetration of wireless services, many people consider their wireless service as much or more a part of their lives as the Public Switched Telephone Network (PSTN). It is predicted that wireless users will demand the same types of services from wireless networks in the future. Implicit in this assumption is that a lack of dependability and other QoS attributes, which have the potential to constrain services and applications (as well as the loss of critical data), will not be tolerated. Although the evolution of Third Generation (3G) wireless systems is projecting data rates of up to 2 Mbps, explicitly lacking are specifications concerning dependability and other QoS attributes. Although there has be a plethora of research done concerning the dependability and other QoS attributes of the PSTN, limited research has been done on wireless networks. Most of the recent QoS research concerning the performance or optimization of wireless networks assumes the wireless network is 100% dependable. However, real world wireless networks have components and links that experience failures. Wireless networks are inherently different from wired networks. There are several important issues related to dependability and other QoS attributes in wireless networks that prohibit wireline models from being transferable. These issues include user mobility and handoff, dynamic connections, and fluctuating channel availability. This study views dependability as one of the most crucial QoS attributes and seeks to investigate the dependability of current and future wireless network infrastructures and its impact other complimentary QoS attributes. The major objectives of this study are: (1) the design of QoS-oriented wireless networks using multi-level fault tolerance and (2) the performance evaluation of such networks using simulation and analytical modeling.