Energy Detection based Spectrum Sensing over Two-wave with Diffuse Power Fading Channels

Eleftherios Chatziantoniou, Ben Allen, Vladan Velisavljevic, Petros Karadimas, Justin Coon

Research output: Contribution to journalArticlepeer-review

27 Citations (Scopus)


One of the most important factors that affects the performance of energy detection (ED) is the fading channel between the wireless nodes. This article investigates the performance of ED-based spectrum sensing, for cognitive radio (CR), over two-wave with diffuse power (TWDP) fading channels. The TWDP fading model characterizes a variety of fading channels, including well-known canonical fading distributions, such as Rayleigh and Rician, as well as worse than Rayleigh fading conditions modeled by the two-ray fading model. Novel analytic expressions for the average probability of detection over TWDP fading that account for single-user and cooperative spectrum sensing as well as square law selection diversity reception are derived. These expressions are used to analyze the behavior of ED-based spectrum sensing over moderate, severe and extreme fading conditions, and to investigate the use of cooperation and diversity as a means of mitigating the fading effects. Our results indicate that TWDP fading conditions can significantly degrade the sensing performance; however, it is shown that detection performance can be improved when cooperation and diversity are employed. The presented outcomes enable us to identify the limits of ED-based spectrum sensing and quantify the trade-offs between detection performance and energy efficiency for cognitive radio systems deployed within confined environments such as in-vehicular wireless networks.
Original languageEnglish
Pages (from-to)868-874
JournalIEEE Transactions on Vehicular Technology
Issue number1
Early online date21 Apr 2016
Publication statusPublished - Jan 2017


Dive into the research topics of 'Energy Detection based Spectrum Sensing over Two-wave with Diffuse Power Fading Channels'. Together they form a unique fingerprint.

Cite this