Document Type

Master's Research

Degree Name

Master of Science in Engineering




Todor Cooklev

Date of Award



With the rapidly-increasing amount of high data rate wireless devices, technologies and services appearing on the market today, there is an increasing demand for the wireless spectrum. Current wireless networks are characterized by a static spectrum allocation policy, where governmental agencies assign wireless spectrum to license holders on a long-term basis for large geographical regions. Recently, because of the increase in spectrum demand, this policy faces spectrum scarcity in particular spectrum bands. Dynamic spectrum access (DSA) shows promise to increase spectral efficiency. DSA aims at dynamically sharing spectrum that is licensed to primary users (PUs) with non-licensed secondary users (SUs). In order to effectively share spectrum the SUs must be sure to access the spectrum only when the PUs are not utilizing it, otherwise the SUs could cause interference to the PUs. One method to determine when a PU is accessing the spectrum is for a SU to identify if the spectrum is occupied or not through spectrum sensing. Spectrum probing is a key component in spectrum sensing and defines the policy for when the SU will perform a channel scan of the spectrum to collect spectral data to be used for spectrum sensing.

This work describes the development of a software defined radio (SDR) testbed based on the Universal Software Radio Peripheral (USRP) for research in DSA with a focus in spectrum probing methods. Spectrum probing methodology is an often overlooked component of spectrum sensing. Theoretical analysis and simulation results for comparing different spectrum probing methods are presented in [1]. This work expands on the work in [1] by using the developed SDR testbed to collect experimental data and compare the results. Different spectrum probing methods are implemented for the case of an independent SU and for the case of a cooperative network of SUs. Experimental results are compared to theoretical analysis and simulated results. The experimental findings further support the conclusions based on simulation in [1]. In particular, in the independent sensing scenario, periodic probing indeed achieves the smallest probing delay; however, in the cooperative sensing scenario randomization can drastically reduce the probing delay.