Group-query-as-a-service for secure dynamic spectrum access in geolocation-enabled database-driven opportunistic wireless communications in ROAR framework

Dynamic spectrum access is regarded as an emerging approach to enhance the RF spectrum utilization in future cognitive wireless networks where unlicensed secondary users sense RF channels (that are licensed to primary users) to find idle channels to use opportunistically without creating any harmful interference to primary users. Recently, the US Federal Communications Commission (FCC) mandated that all unlicensed secondary users must query a spectrum database for channel opportunities instead of sensing by themselves to avoid any channel sensing uncertainties and harmful interference to primary users. However, database query process for secondary users could overwhelm the communication channel when there are many users requesting the same information. This excessive number of queries by numerous secondary users could lead to a denial-of-service attack at the spectrum database server. In this paper, we investigate a Group-Query-as-a-Service for dynamic spectrum access in geolocation-enabled database-driven ROAR (near real-time opportunistic spectrum access in cognitive radio network) framework. In a Group-Query-as-a-Service approach, secondary users form a group based-on location based grids and a limited number of secondary users of the given group, known as grid leaders, query the spectrum database and let other follower secondary users know about the idle channels. Grid leaders are selected based on their past level of interactions and trust they have in the group or grid. Grid leaders query the database on behalf of other secondary users periodically to get the updated idle channel information. Note that when a new user joins the grid, instead of querying the database, it gets the idle channel information from the grid leaders. To avoid any outliers in the grid, we use the majority voting among grid leaders. We evaluate the performance of the proposed approach using numerical results obtained from Monte Carlo simulations and numerical results show that the performance of the query process can be significantly enhanced with Group-Query-as-a-Service.