Cognitive radio network is a key to effectively utilize the underutilized radio spectrum. It works on the basis of Software Defined Radio. Spectrum sensing is the most important part of the cognitive radio system and it is used to detect the presence of signal in the air. There are many techniques to sense spectrum like Matched filter based detection, Energy detection, Cyclostationary detection and Wavelet based detection. In the proposed work, Energy detection is chosen in which spectrum sensing is done by considering static and dynamic threshold level. To evaluate the performance of the detection techniques, MATLAB software has been used for simulation. Simulations were carried out and graphs of probability of detection vs. the probability of false alarm were observed and analysed. The detection probability increases with respect to the increase in false alarm. Also significant reductions in probability of missed detection have been achieved with this sensing technique. The detection probability also varies with the SNR value. SNR has a great influence on the probability of detection. With an increase in SNR value, the probability of detection increases.