LAUSR

In 2010, federal communications commission released the final ruling to allow the unlicensed operation of TV white spaces, i.e., locally vacant TV channels. Since TV white spaces are open to all networks and all types of applications, it is likely that there will be multiple networks and application systems authorized to use the same TV channels at the same time. Currently, there is mechanism for different users to operate in the TV white spaces within one system, while TV spectrum sharing among heterogeneous networks and application systems is largely ignored. To address the problem, this paper designs a trading mechanism to study the TV spectrum market between Geolocation database and heterogeneous application systems. We formulate the spectrum sharing problem into a 0–1 integer optimization problem and prove that this problem is NP-complete. The proposed solution decomposes an optimal fractional solution of a NP-hard problem into a convex combination of internal solutions by randomized rounding algorithm. Simulation results show that the proposed approximation algorithm can achieve a close-to-optimal solution with far less complexity.