Improved channel quality indicator estimation using extended Kalman filter in LTE networks under diverse mobility models

Accurate channel quality indicator (CQI) estimation is crucial for optimizing resource allocation, improving link adaptation, and sustaining high performance in long term evolution (LTE) networks. In real-world scenarios, where channel conditions fluctuate rapidly due to user mobility, inaccurate CQI estimation can lead to suboptimal scheduling, degraded throughput, and reduced quality of service (QoS) for both users and network operators. Traditional Kalman filter (KF) approaches often struggle with the non-linear and time-varying nature of wireless channels, especially under unpredictable mobility patterns. This paper proposes an improved CQI estimation method based on the extended Kalman filter (EKF), which models non-linear system dynamics more effectively. The method is implemented in LTE-Sim, analyzed using MATLAB, and evaluated under random and Manhattan mobility models. Results show that across mobility regimes, KF outperforms EKF in the structured Manhattan model, while in the non-linear random-direction model, EKF yields markedly higher signal-to-interference-plus-noise ratio (SINR) stability and robustness to channel variation with SINR values above 10 dB between 300-450 s and a peak of approximately 60 dB. These results underscore the importance of mobility-aware estimation strategies in enhancing LTE network adaptability and throughput.