The rapid development of wireless communication technology makes Orthogonal Frequency Division Multiplexing (OFDM) technology and Multiple Input Multiple Output (MIMO) technology, key candidates for the next generation mobile communication system. OFDM system has high frequency utilization and effectively resists multipath and narrowband interference. MIMO technology can effectively improve system capacity and system separation. This paper presents a hybrid Linear-Quadratic Estimation (LQE) technique for Channel Estimation in MIMO-OFDM system. In channel estimation, the noise covariance matrix (Q) plays a key role. In general, the value of Q depends on the previous state values. In a dynamic system, the previous values are not always known. The proposed hybrid model uses two techniques to address this issue. When the previous state values are known, Fast Adaptive Hybrid - Linear Quadratic Estimation (FAH-LQE) is used and when the previous state values are unknown Discrete Multivariate Hybrid - Linear Quadratic Estimation (DMH-LQE) is used. This enables the channel estimation to perform effectively in both scenarios. The FAH-LQE method uses covariance channel noise estimation equations and updates the threshold limit. DMH-LQE finds previous values using discrete generate matrix. The switching between FAH-LQE and DMH-LQE, based on the availability of previous state values, has increased the performance of the system effectively. The proposed algorithm produced lower BER values. For a mobility of 55km/h at SNR 15, BER values of proposed algorithm are 22%, 10%, 7%, 6%, and 3% lower than existing methods namely CE with and without EEP-Kalman tracking, perfect channel estimation, FAH-LQE, and DMH-LQE respectively. Similarly for a mobility of 115km/h at SNR 15, BER values of the proposed algorithm are 45%, 35%, 20%, 15% and 5% lower than above mentioned existing methods namely CE with and without EEP-Kalman tracking, perfect channel estimation, FAH-LQE, and DMH-LQE respectively. The SER has been reduced by 5% when no compensation method is applied. By using the proposed method, the Peak to Average Power Ratio (PAPR) has also reduced. © 2021, International Journal of Intelligent Engineering and Systems. All Rights Reserved.