What is a goal of Galileo's magnetometer recursive filter

Question

I'm designing the basics of space magnetometer instrument for academic project and I came across a Galileo mission investigation document, with data flow described in chapter 6. As a first magnetometer readings filter they use recursive filter. To quote the document:

The data, sampled at 30 samples s-1, are filtered to provide 16-bit words, tests have demonstrated that these words are accurate to the 15th bit. This filtering is performed by the microprocessor system using a recursivefilter, a form selected for economy of implementation. The recursive filter algorithm is of the form,

$$ V^j_0 = (1 - A)V^{j-1}_0 + AV^j_i $$

where $V^j_0$ is the (output) value calculated from measurements prior to measurement $j$ and $V^j_i$ is the (input) value read in at measurement $j$. The value of $A$ is set at $\frac{1}{4}$ for the filter at this stage of the data processing.

What is the goal of this filter? To my understanding, it only makes the influence of the next reading smaller, which just changes data.

Answer 1

The formula is an implementation of an exponential smoothing filter, https://en.m.wikipedia.org/wiki/Exponential_smoothing

Answer 2

It is an exponentially weighted moving average (EWMA) filter: a simple lowpass filter intended to eliminate high frequency noise. See my answer here for some analysis and a Matlab/Octave script to make some plots of the filter's frequency response: https://dsp.stackexchange.com/a/40465/28112

The $-3$ dB corner frequency of this particular filter appears to be:

$$f_{3dB} = 30 \mathrm{Hz}\dfrac{\cos^{-1}\left[1-\dfrac{0.25^2}{2(1-0.25)}\right]}{2\pi} = 1.38 \mathrm{Hz}$$