Benchmarking resource usage for spectrum sensing on commodity mobile devices

Effective management of various white space spectra may require spectrum sensing at finer spatial granularity than is feasible with expensive laboratory-grade spectrum sensors. To enable this, we envision a future where commodity mobile devices would be capable of spectrum sensing as needed, possibly via crowd-sourcing. However, since mobile devices are resource limited, understanding their resource usage in this set up is important, specifically in terms of overall latency and energy usage. In this work, we carry out a comprehensive performance benchmarking study using 4 different USB-powered software radios and 2 common smartphone/ embedded computers as mobile spectrum sensing platforms. The study evaluates latency and energy usage using a suite of commonly used sensing algorithms specifically targeting TV white space spectrum. The study shows that latency due to sensing and computation and related energy usage are both modest.