This paper proposes a new blind channel identification and equalization method that leverages the cyclostationarity of oversampled communication signals to achieve accurate channel estimation and equalization without using training signals. Unlike traditional adaptive blind equalization methods, which often suffer from convergence issues and local extrema attractors, the proposed algorithm converges to the channel impulse response using only second-order statistics. This approach can achieve equalization with fewer symbols compared to methods based on higher-order statistics. The method is insensitive to timing recovery uncertainties and can be used to initialize adaptive schemes and facilitate decision feedback adaptation. The algorithm is theoretically sound and has been validated through simulations, demonstrating promising performance for blind equalization of a three-ray multipath channel. The paper also discusses the implementation details and provides a Monte Carlo simulation to evaluate the algorithm's performance at different signal-to-noise ratios (SNRs).This paper proposes a new blind channel identification and equalization method that leverages the cyclostationarity of oversampled communication signals to achieve accurate channel estimation and equalization without using training signals. Unlike traditional adaptive blind equalization methods, which often suffer from convergence issues and local extrema attractors, the proposed algorithm converges to the channel impulse response using only second-order statistics. This approach can achieve equalization with fewer symbols compared to methods based on higher-order statistics. The method is insensitive to timing recovery uncertainties and can be used to initialize adaptive schemes and facilitate decision feedback adaptation. The algorithm is theoretically sound and has been validated through simulations, demonstrating promising performance for blind equalization of a three-ray multipath channel. The paper also discusses the implementation details and provides a Monte Carlo simulation to evaluate the algorithm's performance at different signal-to-noise ratios (SNRs).