A novel polynomial filtering method for data smoothing in cognitive radio applications

The smoothing process is a fundamental task in many application fields. This paper proposes a novel method to smooth raw data, based on the concept of polynomial fitting. It is thought to be effective in Cognitive Radio applications, especially focused on spectrum sensing tasks. The method is intended to be used instead of today's traditional smoothing filters, because of some advantages in terms of shaping retainment, data shifting problem avoidance, acceptable computational intensity, appreciable noise reduction property. The goodness of the proposal has been proved considering the H1 norm operator as performance index.