A Geolocation Database Middleware Sublayer for Minimizing Secondary User Waiting Times in TV White Space Spectrum Allocation

Abstract

In wireless communication, licensed primary users (PUs), such as TV broadcasters, are allocated channels in the UHF and VHF spectrum for their transmission. Television white spaces (TVWS) are the unused channels, or vacant channels (VCs), which could be beneficial to unlicensed secondary users (SUs). With the TVWS wireless technology, the allocation of VC to SUs is currently based on a geolocation database (GLDB) approach. The unlicensed SUs compete for VCs as soon as they are released by the GLDB, which generates contention and long waiting times (among the SUs). Developing an efficient channel allocation strategy for the SUs, using the GLDB method, continues to be a challenge, including long waiting times. This paper discusses a methodology to minimize long waiting times by integrating a middleware sublayer, called dynamic queue regulator (DQR), into the main GLDB architecture. Under this strategy, the DQR GLDB issues a pricing scheme to the SUs in three categories: New SUs, Returning SUs, and SUs in Waiting. By coordinating with the main GLDB, the DQR middleware uses a blind ticketing cost model based on the idle times of PUs, which requires all arriving SUs to submit price offers. In this research, we simulate the dynamics of VC allocation in TVWS spectrum using two types of databases: the standard or Normal GLDB, proposed by regulators, and our DQR middleware GLDB. The simulation results show that the DQR middleware, when integrated into the main GLDB architecture, minimizes the SU waiting times by 0.3 h, or a 35.3% improvement.