Algorithmic aspects of resource allocation in cognitive radio wireless networks
Loading...
Date
2013
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Montana State University - Bozeman, College of Engineering
Abstract
Wireless networking is a critical component of today's internet infrastructure. Two examples of important wireless internet infrastructure are long distance network backbone links and last-mile solutions to remote areas. Wireless technology already supplies a wide variety of consumer solutions including analog television channels (TVWS), cellular infrastructure for massive scale real-time communication, and computer networking for seamless global connectivity. Worldwide, there are an estimated 2.5B internet users and 6B cellular phone subscribers- and those numbers are steadily growing. Sufficient capacity for divergent wireless applications, along with their growing users, calls for a more efficient use of bandwidth. We present multiple resource allocation algorithms to address this challenge in various aspects of wireless networking. Each algorithm focuses on a single resource of wireless networking: antenna beam sector activation, directional antenna beam bearings and duration, joint routing and channel selection, and link-channel allocation. In terms of computation and memory, our topology control algorithms provide near optimal performance with significantly lower cost. For each algorithm, a rich set of simulation scenarios is presented that compare our novel algorithms performance to the optimal solution. Ultimately, we present a topology control algorithm that provides an efficient solution to the channel rental problem: finding the most cost-effective set of communication channels (for a wireless mesh network) at a minimum performance guarantee. This problem occurs in high-density traditional wireless networking, cellular networking, and rural sparse networking with last mile internet connectivity; topology control algorithms are well suited for all applications of wireless technology. These algorithms are shown to be robust against various network challenges including topology, frequency availability, and interference.