Compressão de sinais para fronthaul em arquitetura CRAN utilizando algoritmo evolutivo

Abstract

Centralized radio access networks are present as a potential alternative for next generation of wireless networks, because they are able to provide high data rates and allow the reduction of structural and operational costs in the network. The centralized architecture implements the concept of fronthaul, and enables the challenge to increase the capacity of data transmission in hese links. In this way, the study of digital signal compression techniques presents itself as an alternative to reduce the cost of implementing centralized radio access networks.This work investigates the use of vector quantization methods in the compression of complex samples of baseband LTE signals. We propose the use of Genetic Algorithms in the training of sub-optimal dictionaries for the process of vector quantization in order to reduce the errors imposed in this process and consequent increase in fronthaul capacity. Results showed that the proposed compression algorithm allows reduction of fronthaul data rates associated with acceptable errors. It has been shown to be possible data rate compression factors of 5:4 times, with errors of approximately 4:4%, proving the effectiveness of codebook training process in LTE signals present in the downlink of centralized radio access networks.