I am confused by the following statement from the paper Beamforming: A Versatile Approach to Spatial Filtering by Barry D. and Kevin M.:
There are two basic adaptive approach: 1) block adaption ..., 2) continuous adaption.... If a nonstationary environment is anticipated, block adaption can be used, provided that the weights are recomputed periodically....Continuous adaption is usually preferred when statistics are time-varying or (for computational reasons) when the number of adaptive weights M is moderate to large (values of M>50 is not uncommon).
This statement gives the conditions when the two kinds of adaption methods should be used. But the condition NONSTATIONARY and TIME-VARYING, in my opinion, are the same. Can you point out in what situations they are different? Illustrative examples would be greatly appreciated.