Summary: | 碩士 === 國立清華大學 === 通訊工程研究所 === 99 === Spectrum sensing is a very important issue in cognitive radio networks. In this thesis, we propose an efficient cooperative spectrum sensing scheme which can not only reduce the number of sensing reports to the fusion center but also ensure a low miss detection probability. In the scheme, each unlicensed (secondary) user senses the licensed band and makes a local decision of the sensing result by two thresholds λh and λo (λh > λo). As long as the fusion center receives a high assured sensing result from a secondary user, it can immediately make a global decision that the licensed band is occupied. Otherwise, the fusion center need to receive the sensing energies from the secondary users and combines the received sensing energies to make a global decision. Then the fusion center broadcasts the final decision to all secondary users. Each secondary user compares the received sensing result with its local decision. If the global decision is equal to the local decision of a secondary user, the secondary user will increase its reputation value. Otherwise, the secondary user will decrease its reputation value. If the reputation value of a secondary user is smaller than a predetermined threshold, the secondary user cannot send its sensing result to the fusion center in next sensing slot. Simulation results show that the performance of our proposed method is better than previous work.
|