KALMAN FILTERING IS a state estimation technique
used in many application areas such as spacecraft
navigation, motion planning in robotics, signal
processing, and wireless sensor networks because
of its ability to extract useful information from
noisy data and its small computational and memory
12, 20, 27–29 Recent work has used Kalman
filtering in controllers for computer systems.
5, 13, 14, 23
Although many introductions to Kalman filtering are
available in the literature,
1–4, 6–11, 17, 21, 25, 29 they are usually
focused on particular applications such as robot motion
or state estimation in linear systems, making it difficult to
see how to apply Kalman filtering to other problems. Other
presentations derive Kalman filtering as an application
of Bayesian inference, assuming that
noise is Gaussian. This leads to the
common misconception that Kalman
filtering can be applied only if noise
Abstractly, Kalman filtering can be
seen as a particular approach to combining approximations of an unknown value to produce a better approximation.
Suppose we use two devices of different
Demystifying the uses of a powerful tool
for uncertain information.
BY YAN PEI, SWARNENDU BISWAS,
DONALD S. FUSSELL, AND KESHAV PINGALI
˽ This article presents an elementary
derivation of Kalman filtering, a classic
state estimation technique.
˽ Understanding Kalman filtering is
useful for more principled control
of computer systems.
˽ Kalman filtering is used as a black box
by many computer scientists.