A high accuracy frequency synchronization method is proposed for the 3rd generation partnership project (3GPP) long term evolution advanced (LTE-A) downlink receiver in time division duplexing (TDD) mode.In general,cyclic prefix (CP) correlation based fractional frequency offset (FFO) estimation method and primary synchronization signal (PSS) differential correlation based integer frequency offset (IFO) estimation method are applied for LTE-A frequency synchronization.However,the polarity of CP based FFO estimation result may get reversed when system FFO is closer to the edge of frequency estimation range on account of noise interference;PSS based IFO estimation has performance degradation in low signal noise ratio (SNR).We propose polarity detection aided CP based FFO estimation and frequency domain enhanced differential correlation based IFO estimation to obtain higher accuracy of frequency synchronization.Computer simulation shows that the proposed method greatly outperforms the conventional methods,especially in low SNR scenario. A high accuracy frequency synchronization method is proposed for the 3rd generation partnership project (3GPP) long term evolution advanced (LTE-A) downlink receiver in time division duplexing (TDD) mode.In general, cyclic prefix (CP) (FFO) estimation method and primary synchronization signal (PSS) differential correlation based integer frequency offset (IFO) estimation method are applied for LTE-A frequency synchronization. to propose the method of estimating the performance of a low frequency noise ratio (SNR). We propose polarity based aided CP based FFO estimation and frequency domain enhanced differential correlation based IFO estimation to obtain higher accuracy of frequency synchronization. Computer simulation shows that the proposed method greatly outperforms the con ventional methods, especially in low SNR scenario.