Research Article
Sensitivity of the Packet Level Performance Metrics to the Primary Channel Holding Time Distribution in Cognitive Radio Networks with VoIP-traffic
@INPROCEEDINGS{10.4108/icst.crowncom.2014.255339, author={Sandra Castellanos-Lopez and Felipe A. Cruz-P\^{e}rez and Genaro Hernandez-Valdez and Mario Rivero-Angeles}, title={Sensitivity of the Packet Level Performance Metrics to the Primary Channel Holding Time Distribution in Cognitive Radio Networks with VoIP-traffic}, proceedings={9th International Conference on Cognitive Radio Oriented Wireless Networks}, publisher={IEEE}, proceedings_a={CROWNCOM}, year={2014}, month={7}, keywords={cognitive radio networks voip traffic primary channel holding time probability distribution session blocking probability packet dropping probability mean packet delay}, doi={10.4108/icst.crowncom.2014.255339} }- Sandra Castellanos-Lopez
Felipe A. Cruz-Pérez
Genaro Hernandez-Valdez
Mario Rivero-Angeles
Year: 2014
Sensitivity of the Packet Level Performance Metrics to the Primary Channel Holding Time Distribution in Cognitive Radio Networks with VoIP-traffic
CROWNCOM
IEEE
DOI: 10.4108/icst.crowncom.2014.255339
Abstract
In this paper, the performance of VoIP traffic-based cognitive radio networks (CRNs) with different primary channel holding time (CHT) distribution (specifically, negative exponential, log-normal, and Coxian) is investigated. From the mathematical point of view, a joint connection level and packet level analysis considering the primary CHT to be Coxian distributed is developed. The developed mathematical analysis allows evaluating system performance at both connection and packet levels. Then, system performance is evaluated in terms of packet dropping probability and mean packet delay considering a target VoIP session blocking probability of cognitive users. Numerical results for CHT log-normally and Coxian distributed are shown for different values of their first three standardized moments. Numerical results reveal that packet level performance metrics are practically insensitive to the probability distribution of the primary CHT. Thus, the analytical performance evaluation can be performed considering the negative exponential distribution for the sake of simplicity and mathematical tractability.