Throughput-Efficient Channel Allocation Algorithms in Multi-Channel Cognitive Vehicular Networks.

Many studies show that the dedicated short range communication band allocated to vehicular communications is insufficient to carry the wireless traffic generated by emerging vehicular applications. A promising bandwidth expansion possibility presents itself through the release of large TV band spectra (i.e., the TV white space spectrum) by the Federal Communications Commission for cognitive access. One primary challenge of the so-called TV white space (TVWS) spectrum access in vehicular networks is the design of efficient channel allocation mechanisms in face of spatial-temporal variations of TVWS channels. In this paper, we address the channel allocation problem for multi-channel cognitive vehicular networks with the objective of system-wide throughput maximization. We show that the problem is an NP-hard non-linear integer programming problem, to which we present three efficient algorithms. We first propose a probabilistic polynomial-time $(1-1/e)$ -approximation algorithm based on linear programming. Next, we prove that the objective function can be written as a submodular set function, based on which we develop a deterministic constant-factor approximation algorithm with a more favorable time complexity. Then, we further modify the second algorithm to improve its approximation ratio without increasing its time complexity. Finally, we show the efficacy of our algorithms through numerical examples. [ABSTRACT FROM PUBLISHER]