Which statement accurately distinguishes filtering and smoothing in navigation data processing?

• A. Filtering estimates the current state using present measurements.
• B. Smoothing estimates the current state using present measurements.
• C. Filtering refines past estimates using future data.
• D. Smoothing enables real-time operation and filtering improves retrospective accuracy.

Answer: (A)

In navigation data processing, filtering and smoothing describe different ways of using measurements over time to estimate the state. Filtering is a causal, real-time process: at each moment, you update your estimate of the current state using measurements up to and including the present measurement. The current observation directly informs the current state estimate, and you don’t rely on future data to decide what the state is now.

That’s why the statement that filtering estimates the current state using present measurements is the best description. It captures the essence that the current update is based on information available up to now, making filtering suitable for real-time operation.

Smoothing, by contrast, uses measurements from the entire window, including future data relative to a given time, to refine estimates—typically producing better estimates of past states after the full data set is available. It’s not inherently real-time, since it relies on future observations to improve past estimates.

So the correct idea is that filtering estimates the current state using present measurements, while smoothing leverages future data to improve past estimates and is not limited to real-time use.