Document Type dissertation Author Name Rocke, Sean A URN etd-042513-124548 Title On Random Sampling for Compliance Monitoring in Opportunistic Spectrum Access Networks Degree PhD Department Electrical & Computer Engineering Advisors Professor Alexander M. Wyglinski, Advisor Professor Kaveh Pahlavan, Committee Member Professor Donald R. Brown III, Committee Member Dr. Kim I. Mallalieu, Committee Member Keywords spectrum monitoring compliance enforcement opportunistic spectrum access Date of Presentation/Defense 2013-04-24 Availability unrestricted
In the expanding spectrum marketplace, there has been a long term evolution towards more marketoriented mechanisms, such as Opportunistic Spectrum Access (OSA), enabled through Cognitive Radio (CR) technology. However, the potential of CR technologies to revolutionize wireless communications, also introduces challenges based upon the potentially nondeterministic CR behaviour in the Electrospace. While establishing and enforcing compliance to spectrum etiquette rules are essential to realization of successful OSA networks in the future, there has only been recent increased research activity into enforcement.
This dissertation presents novel work on the spectrum monitoring aspect, which is crucial to effective enforcement of OSA. An overview of the challenges faced by current compliance monitoring methods is first presented. A framework is then proposed for the use of random spectral sampling techniques to reduce data collection complexity in wideband sensing scenarios. This approach is recommended as an alternative to Compressed Sensing (CS) techniques for wideband spectral occupancy estimation, which may be difficult to utilize in many practical congested scenarios where compliance monitoring is required.
Next, a lowcost computational approach to online randomized temporal sensing deployment is presented for characterization of temporal spectrum occupancy in cognitive radio scenarios. The random sensing approach is demonstrated and its performance is compared to CSbased approach for occupancy estimation. A novel framebased sampling inversion technique is then presented for cases when it is necessary to track the temporal behaviour of individual CRs or CR networks. Parameters from randomly sampled Physical Layer Convergence Protocol (PLCP) data frames are used to reconstruct occupancy statistics, taking account of missed frames due to sampling design, sensor limitations and frame errors.
Finally, investigations into the use of distributed and mobile spectrum sensing to collect spatial diversity to improve the above techniques are presented, for several common monitoring tasks in spectrum enforcement. Specifically, focus is upon techniques for achieving consensus in dynamic topologies such as in mobile sensing scenarios.
Browse by Author | Browse by Department | Search all available ETDs
Questions? Email firstname.lastname@example.org