He also suggested the relationship:
2. The Kalman gain sequence automatically adapts to changing detection histo-ries. This includes a varying sampling interval as well as missed detection.
3. The Kalman filter provides a convenient measurement of the estimation accu-racy through the covariance matrix. This measure is required to perform the data association functions.
1. Target probability density function (pdf) is assumed Gaussian. But in many applications this assumption is inappropriate.
2. Higher computational load than the fixed-gain filter.