Abstract:
Spectrum sensing in cognitive radio (CR) is a critical process as it directly
influences the accuracy of detection. Noise uncertainty affects the reliability
of detecting vacant holes in the spectrum, thus limiting the access of that
spectrum by secondary users (SUs). In such uncertain environment; SUs
sense the received power of a primary user (PU) independently with
different measures of signal-to-noise ratio (SNR). Long sensing time serves
in mitigating the effect of noise uncertainty, but on the cost of throughput
performance of CR system. In this paper, the scheme of an asynchronous
and crossed sensing-reporting is presented. The scheme reduces energy
consumption during sensing process without affecting the detection
accuracy. Exploiting the included idle time (𝑇𝑖
) in sensing time slot; each SU
collects power samples with higher SNR directly performs the reporting
process to a fusion center (FC) consecutively. The FC terminates the sensing
and reporting processes at a specific sensing time that corresponds to the
lowest SNR (𝑆𝑁𝑅𝑤𝑎𝑙𝑙). Furthermore, this integrated scheme aims at
optimizing the total frame duration (𝑇𝑓). Mathematical expressions of the
scheme are obtained. Analytical results show the efficiency of the scheme in
terms of energy saving and throughput increment under noise uncertainty.
