ROAR: An architecture for Real-Time Opportunistic Spectrum Access in Cloud-assisted Cognitive Radio Networks

摘要

Need for Radio Frequency (RF) spectrum is increasing along with increasing wireless subscriptions and devices that is causing scarcity in total available RF spectrum. Opportunistic spectrum access in cognitive radio networks is emerging for maximizing the RF spectrum efficiency where unlicensed Secondary Users (SUs) access idle spectrum bands without causing any harmful interference to licensed Primary Users (PUs). All SUs are required to scan/sense RF spectrum to find idle bands or search for idle bands in a spectrum database not to interfere with PUs while using those idle bands. In this paper, we propose a Real-time Opportunistic Spectrum Access in Cloud-assisted Cognitive Radio Networks (ROAR) architecture where SUs (i.e., USRP wireless devices) are equipped with wide-band (50 MHz ??? 6 GHz) antenna and GPS units. ROAR uses cloud computing platform for real-time processing of wide-band data since SUs' performance is considerably constrained by their limited power, memory and computational capacity. In ROAR architecture, there are two parts: spectrum sensing to create a database of idle channels and dynamic spectrum access for opportunistic SU communications using idle channels. For the spectrum database, RF sensors scan/sense RF bands to find idle bands and report the geo-location of idle band, channel frequency and time stamp to the database installed in the distributed cloud platform. For opportunistic spectrum access, each SU interested for opportunistic communication queries a spectrum database to find idle channels. Distributed cloud computing is used to find idle channels for the SU where geolocation and other demands (e.g., data rate) are checked to find whether the SU is admissible or not for a given geo-location and time. If the SU is admissible based on the admissibility criteria, spectrum server sends the list of channels available for a given location and time to the SU. Then SU chooses the best suited channel to communicate opportunistically. We e- aluate the ROAR architecture using numerical results obtained from extensive experiments.