DFT and ANN based channel estimation techniques combined to 3-tap equalization for a proposed FBMC scattered pilot frame model

This work presents a Filter Bank Multi Carrier (FBMC) based digital TV frame model combined to novel 3-tap Artificial Neural Network (ANN) and statistical Discrete Fourier Transform (DFT) channel estimation techniques. This arrangement improves the spectral efficiency and robustness of the transmission of data by avoiding the use of Cyclic Prefix (CP). The results show that pilot aided FBMC frame can be employed in digital TV applications. The Bit Error Rate (BER) curves demonstrate that ANN and DFT with statistical threshold 3-tap methods enhance the performance in multipath scenarios in comparison to traditional techniques.