For online, real-time applications, WIndLS is implemented recursively (Recursive Least Squares with Forgetting Factor). The algorithm updates the estimate at each time step $t$ without reprocessing all past data.
Windowed Least Squares (WIndLS) is a robust extension of the standard LS method tailored for time-varying systems. By exponentially weighting data, it sacrifices the stability of long-term averaging for the agility required to track dynamic environments. It remains a cornerstone technique in modern adaptive signal processing and control engineering. windls
This creates a "window" of relevance, focusing the estimation on the most recent system behavior. windls