Skip to main content
. 2024 Oct 15;24(20):6633. doi: 10.3390/s24206633
Algorithm 1: Improved GMM User Clustering Algorithm
1: Input user coordinates and base station location, then compute the initial values for the number of clusters M and means μ.
2: Initialize parameters m and m.
3:  Repeat
4: E-Step: Keep parameters m,m,μm unchanged. Compute the posterior probability l(zim) of user ui by (10).
5: M-Step: Keep l(zim) unchanged. Update parameters μm,m,m by (12), (13), and (14).
6: Repeat until the objective likelihood function L converges or the maximum number of iterations is reached.