Dynamic Spectrum Access for Multimedia Transmission Over Multi-User, Multi-Channel Cognitive Radio Networks

The optimal spectrum access strategy is investigated for multi-user multi-channel scenario in cognitive radio networks. At first, an online learning method based on Dirichlet Process is adopted to predict the channel usage based on ACK/NACK feedbacks, which can avoid frequent information exchange am...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on multimedia 2020-01, Vol.22 (1), p.201-214
Main Authors: Huang, Xin-Lin, Tang, Xiao-Wei, Hu, Fei
Format: Article
Language:English
Subjects:
Algorithms
Bayesian online learning
Cognitive radio
Cognitive radio networks
Computer simulation
Delay
Delays
Dirichlet problem
Dynamic spectrum access
Heuristic algorithms
Information exchange
Interference
Machine learning
Multimedia
Optimization
Quality of service
Signal to noise ratio
Spectrum allocation
Streaming media
Wireless communication
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The optimal spectrum access strategy is investigated for multi-user multi-channel scenario in cognitive radio networks. At first, an online learning method based on Dirichlet Process is adopted to predict the channel usage based on ACK/NACK feedbacks, which can avoid frequent information exchange among users. Based on the prediction result, the delay performance can be computed when a user transmits certain percentage of multimedia packets on a specific channel. Second, the packet delivery ratio (PDR) is derived from the prediction result of channel usage to reflect the accessing competition among multiple users. Finally, the quality of service (QoS) of multimedia applications is defined as the joint delay and throughput performances. Moreover, a dynamic spectrum access scheme is proposed to optimize the QoS metrics. The simulation results demonstrate that the QoS and the peak-signal-to-noise ratio (PSNR) of the proposed spectrum access algorithm outperform the three existing spectrum access algorithms, i.e., cognitive cross-layer algorithm, dynamic learning algorithm, and dynamic least interference algorithm. The proposed algorithm achieves more than 21.8%, 5.4%, and 3.9% PDR enhancement and over 3.23 dB, 0.82 dB, and 0.50 dB PSNR gains, compared with those three algorithms, given the transmission power as 10, 20, and 30 units, respectively.
ISSN:1520-9210
1941-0077
DOI:10.1109/TMM.2019.2925960