NC-OFDM as one of the candidates for the Cognitive Radio (CR) system has many challenges in the synchronizing time. This is due to the frequency sharing which can cause Narrowband Interference (NBI), which causes a very significant time synchronization error. Large errors in time synchronization can reduce overall NCOFDM system performance. In this study, we evaluate the time synchronization performance of NC-OFDM systems on the multipath channel in the present of NBI. We provide an evaluation of the time synchronization performance by comparing the time synchronization algorithm that has a timing metric such as an impulse-like (Symmetric correlator). Here, we present three algorithms that use the symbol training to calculate the arrival time of the NC-OFDM symbol. The simulation conducted by considering the effects of multipath channel, Signal to Noise Ratio (SNR), Carrier Frequency Offset (CFO), and SIR (Signal to Interference Ratio). SIR is modeled with NBI. Algorithm performance is measured by MAE (Mean Absolute Error) and MSE (Mean Squared Error). Simulation results show that the autocorrelation-based method has very poor performance in MAE and MSE tests, so it is not suitable for NC-OFDMbased Cognitive Radio systems. The GLRT (Generalized Likelihood Ratio Test) method and Iterative Normalization method perform well in case the NC-OFDM Cognitive Radio systems.
Robust time synchronization for NC-OFDM cognitive radio systems
