ICYTE 26279
INSTITUTO DE INVESTIGACIONES CIENTIFICAS Y TECNOLOGICAS EN ELECTRONICA
Unidad Ejecutora - UE
congresos y reuniones científicas
Título:
Design of a Neural Network Based Successive Cancellation Decoder for Polar Codes
Autor/es:
MONICA C. LIBERATORI; JORGE CASTIÑEIRA MOREIRA; IVAN EXEQUIEL GELOSI; LEONARDO OSCAR COPPOLILLO; RAMIRO ÁVALOS RIBAS; ALEJANDRO J. URIZ
Lugar:
Bahía Blanca
Reunión:
Congreso; XVIII Reunión de trabajo en Procesamiento de la Información y Control; 2019
Institución organizadora:
Instituto de Investigaciones en Ingeniería Eléctrica ?Alfredo Desages? (IIIE, UNS-CONICET) y el Departamento de Ingeniería Eléctrica y de Computadoras de la Universidad Nacional del Sur (DIEC?UNS)
Resumen:
In this paper, a Neural Network based decoder is presented, which is based on the Successive Cancellation decoder for Polar Codes. The purpose is to test a classic decoding scheme, while replacing the functions which characterize the structure of the Polar Code Decoder by Neural Networks. The topology of the resulting system was studied by simulation, and the best parameters, selected. The Frame Error Rate was computed as well, for transmission over three different channels (Additive White Gaussian Noise, Rayleigh fading and Impulsive Noise channels).Notably, the proposed decoder does not require channel side information, and shows a good performance independently of the characteristics of the channel.
