Cyclostationary spectrum sensing based channel estimation using complex exponential basis expansion model in cognitive vehicular networks

Cyclostationarity sensing methods are appealing for spectrum sensing due to its strong robustness to noise uncertainty. However, in cognitive vehicular networks, the Doppler frequency shift induced by high mobility cognitive vehicle will bring the cyclic frequency offset (CFO) for cyclostationarity spectrum sensing. The CFO can cause significant detection performance degradation because of a difference between cyclic frequency aware of the cognitive vehicle and the actual cyclic frequency of primary signal. To address this issue, cyclostationary spectrum sensing based on channel estimation using complex exponential basis expansion model (CE-BEM) is established in this paper. We firstly establish a Doppler frequency shift estimation method based on in-vehicle information. Then an appropriate CE-BEM is given according to the value of Doppler frequency shift estimation. The cyclostationarity spectrum sensing based on CE-BEM model for single user and cooperative users are provided. Theoretical analysis show that new cyclostationary characteristics are produced on account of the cyclostationarity induced by the CE-BEM. Simulation results demonstrate that both the local cyclostationarity spectrum sensing (LCSS) and the cooperative cyclostationarity spectrum sensing (CCSS) provide substantial improvement on detection performance in the dynamic moving speed environment for cognitive vehicles.