Spectrum sensing system in software-defined radio for determining spectrum availability

Spectrum sensing is an integral part of Cognitive radio, a radio that seeks to address perceived spectrum scarcity that studies have shown was rather caused by inefficient utilization of the available spectrum. In this paper, a spectrum sensing system using energy detection for analog TV and FM broadcast transmitters, and modified ISDB-T signal was implemented in a Software-defined radio platform using GNU Radio and USRP N200. Real-time implementation and experimental tests were conducted in Metro Cebu, a highly urbanized area in the southern part of the Philippines. Massive tests and measurements were necessary to determine spectrum availability particularly in the TV band. This is in support of the Philippine government's efforts to provide internet connectivity to rural areas. Experimental results have so far met the IEEE 802.22 requirements for energy detection spectrum sensing. The designed system was able to detect signals at -114 dBm with a sensing time of 100 ms. Further, the required Pd(≥ 90) and Pfa(≤ 10) of the standard were also achieved with different threshold for various signal sources representing primary users.