Learning-Based Distributed Radio-Environment Observation for Dynamic Spectrum Sharing

Distributed real-time spectrum sensing is a key enabler for dynamic spectrum sharing; however, communication cost for observed-data transmission is crucial especially when massive numbers of spectrum sensors are deployed. In considering the use of low-cost wireless connections provided for machine-to-machine communication, data rates should be less than tens of kilobits per second (kbps). To cope with this issue, proposed spectrum sensors learn ambient radio environment in real-time and create a time-spectral model, of which parameters are shared with servers operating in the edge-computing layer. This process makes it possible to drastically reduce communication cost of the sensors because frequent data transmission is no longer needed while enabling the edge servers to keep up on current status of the radio environment. On the basis of the created time-spectral model, sharable spectrum resources are dynamically harvested and allocated in terms of geospatial, temporal, and frequency-spectral domains when accepting an application for secondary-spectrum use. A web-based prototype spectrum-management system is implemented using ten servers and dozens of sensors. The measurement results show that proposed approach can reduce data traffic between the sensors and servers by 97%, achieving an average data rate of 10 kbps.