Broadcast Scheduling with Latency and Redundancy Analysis for Cognitive Radio Networks

Abstract

Cognitive radio networks (CRNs) introduce a new communication paradigm that enables unlicensed users to opportunistically access spectrum bands assigned to licensed users. Interestingly, the broadcast problem, which is one of the most fundamental operations in CRNs, has not been well studied. Existing works for the broadcast issue in CRNs are heuristic solutions either without a performance guarantee or with performance far from the optimal solution. In this paper, we study the minimum-latency broadcast scheduling (MLBS) issue for CRNs. Our contributions are threefold. First, we propose a mixed broadcast scheduling (MBS) algorithm under the unit disk graph (UDG) model, denoted by MBS-UDG. MBS-UDG finishes a broadcast task by employing mixed unicast and broadcast communication modes in two phases. We show that the latency performance of MBS-UDG is $O(\hbar+\Delta_{T})$ when $\Delta_{T}\leq \hbox{1}/p$ or $O(\hbar+\log_{1-p}(\hbox{1}/p\Delta_{T}))$ when $\Delta_{T}>\hbox{1}/p$ , where $\hbar$ and $\Delta_{T}$ are the height and the maximum number of leaf nodes connected by a second user (SU)of the broadcasting tree, respectively, and $p$ is the spectrum opportunity for a secondary communication. Furthermore, the redundancy performance of MBS-UDG is analyzed. Second, we extend MBS-UDG to a more general MBS algorithm under the protocol interference model and analyze its latency and redundancy performance. Finally, simulations are conducted to validate MBS, which indicate that MBS significantly improves existing algorithms with respect to both latency and redundancy.