A sustainable production of electricity is essential for low carbon green growth in South Korea. Although wind energy is unlimited in potential, both intermittency and volatility should be tackled for smart grid integration in future. To cope with this, many works have been done for wind speed and power forecasting. It is shown that statistical techniques are useful for short-term forecasting of wind power. This paper presents a statistical wind speed forecasting. A wavelet decomposition is employed as a de-noising technique. An illustration will be given by real-world dataset. According to the result, the mean absolute deviation (MAD) is improved as much as 10%.