Regresar

Comparative Performance Analysis of Training Sequences in Equalization

Abstract:

During the last decade we have experienced an increasing demand for wireless communication systems, that require the implementation of different functions, including equalization, to establish a reliable connection between nodes of the network. The correlation and statistical properties of the training sequences have been exploited in wireless systems in the past, however there are few studies that compare the impact of these sequences on the efficacy of the equalization process where they are implemented. Our experimental design considers five well known training sequences of the same length, including Barker, Willard, Neuman-Huffman, a Random binary sequence, and the Zadoff-Chu (ZC) code, and a new proposed sequence. The simulation results show that Random code is the best performing training sequence, especially when it is used in environments with strong levels of inter-symbol interference (ISI), ZC performs better for conditions with lower ISI and our proposed sequence named EoP performs in second place for any ISI level.