Spectrum sensing with programmable granularity bands using Cosine Modulated Filter Bank for Cognitive Radio

In this paper we propose Cosine Modulated Filter Bank (CMFB) for spectrum sensing with programmable granularity for Cognitive Radio (CR). The bandwidth efficiency of the detected spectral holes can be increased by varying the granularity of prototype filter depending on the spectrum to be sensed. Here we emphasize spectrum sensing can be done effectively by proper choice of prototype filter with specified subband attenuation. The energy detection after subband filtering detects the spectral holes with high confidence due to spectral containment in the subband sensing using CMFB. The significant advantage of using CMFB technique is that the same filter bank structure can also be used for reallocation of secondary user in the detected spectral holes. Through simulations it is shown that spectrum sensing with varying granularity maintains bandwidth efficiency and reliability in spectrum detection even at low SNR.