Channel Equalization Based on Decision Trees
David Baptista, Rafael Ferrari, Romis Attux · Journal of Communication and Information Systems · 2020
This paper analyzes the application of decision trees to the problem of communication channel equalization. Decision trees are interesting structures because they are nonlinear and relatively simple from a computational standpoint. They are tested for channel models that give rise to classification tasks of different complexity and compared to the Bayesian equalizer and the Wiener linear equalizer. The results are quite encouraging, as they show that the tree-based equalizer reaches, in many cases, a performance similar to that of the Bayesian filter at a lower computational cost.