Channel estimation by symmetrical clustering
Jugurta Montalvão, Bernadette Dorizzi, João Mota · IEEE Transactions on Signal Processing · 2002
A new blind channel estimation algorithm is presented. This algorithm comes from the well-known maximum likelihood estimation approach. However, we intentionally "smooth" the joint probability density function (PDF) of a finite set of observations in order to reduce the computational burden. As a result, we obtain an online clustering algorithm whose main characteristic is the constraint of symmetry among cluster centers. Computational simulations are used to evaluate this algorithm.