New Bussgang methods for blind equalization
João Batista Destro Filho, Gérard Favier, João Marcos Travassos Romano · 2002
This paper analyses Bussgang blind algorithms in terms of the minimization of an associated nonlinear cost function. Then, two new blind algorithms (modified decision directed and modified Bellini's (1986)) are proposed and evaluated by computer simulation. The simulation results show the good behaviour of these new blind algorithms and that they converge faster than their standard counterparts. The Bussgang techniques are the oldest methods for blind channel equalization. These methods can be interpreted as stochastic gradient algorithms.