Analysis of Baseband IQ Data Compression Methods for Centralized RAN
Résumé
Through recent wireless technologies, such as Centralized Radio Access Network, baseband unit and remote radio heads are physically separated and connected using fronthaul links. Compressing complex baseband signal samples prior transmission over fronthaul link is an effective way to satisfy the pressing need to decrease the huge required transported data rates. In this paper, we analyze the existing IQ data compression schemes exploiting time and spectral signal characteristics. We consider compression system evaluation parameters to have a smooth trade-off between required signal quality and complexity performance while achieving an acceptable compression gain. We propose an optimized uniform quantization technique combined with entropy coding achieving non-uniform quantization performance in exploiting signal temporal statistical characteristics with much less computational complexity. We also present a comparison between simulation results analyzing the trade-off between the removal of the signal spectral redundancies and vector quantization in terms of performance and complexity.
Origine : Fichiers produits par l'(les) auteur(s)
Loading...