In this paper, we propose to improve the performance of the channel estimation for LTE Downlink systems under the effect of the channel length. As LTE Downlink system is a MIMO-OFDMA based system, a cyclic prefix (CP) is inserted at the beginning of each transmitted OFDM symbol in order to mitigate both inter-carrier interference (ICI) and inter-symbol interference (ISI). The inserted CP is usually equal to or longer than the channel length. However, the cyclic prefix can be shorter because of some unforeseen channel behaviour. Previous works have shown that in the case where the cyclic prefix is equal to or longer than the channel length, LMMSE performs better than LSE but at the cost of computational complexity .In the other case, LMMSE performs also better than LS only for low SNR values. However, LS shows better performance for LTE Downlink systems for high SNR values. Therefore, we propose a hybrid LS-LMMSE channel estimation technique robust to the channel length effect. MATLAB Monte-Carlo simulations are used to evaluate the performance of the proposed estimator in terms of Mean Square Error (MSE) and Bit Error Rate (BER) for 2x2 LTE Downlink systems.