Joint sub-Nyquist wideband spectrum sensing and reliable data transmission for cognitive radio networks over white space

Abstract The emerging cognitive radio technology resolves the spectrum scarcity problem by exploiting the unutilised licenced spectrum on an opportunistic basis. Meanwhile, future cognitive devices require advanced sensing techniques for rapid and active identification of spectrum holes over a wideband. Incorporation of prior information from the geolocation database enhances the sensing performance, reduces the computation complexity and maximizes the spectrum utilisation by white space devices. However, unlike the TV white space database, other wideband spectrum databases are not yet available. Therefore for dynamic and fragile cognitive networks, the geolocation database cannot assist the spectrum sensing process, which necessitated database-independent real-time spectrum sensing. Constant interference from primary users is yet another major challenge with cognitive radio networks which makes reliable communication extremely difficult for secondary users. Forward error correction codes ensure the reliability of transmitted data, among which low-density parity check codes are the most appropriate. Lengthy low-density parity check codes, however, have certain drawbacks specifically high encoding and decoding complexity. Parallel concatenated Gallager code addresses these issues. This paper presents a reliable data transmission scheme for secondary users using the proposed non-zero syndrome parallel concatenated Gallager codes with interleaver for cognitive radio networks performing real-time spectrum sensing over white space. The joint sensing scheme along with prior knowledge from the geolocation database improves the spectrum sensing performance with reduced complexity and is more suitable for low-power cognitive devices. Further, simulation results affirm that proposed PCGC outperform a dedicated LDPC in terms of coding gain and bit-error rate.