ADSLEqualization-HomepagesofESAT.pptVIP

Content Implementation of theoretical schematic Optimisation algorithm (LMS) Simulation results Real channel Transmitter Channel and TEQ Receiver TEQ Determine W and B through different algorithms Choose length W and B Goal: h?w shorter than h Least Mean Squares Simulation results More or less ideal channel Simulation results Real channel Simulation results Influence CP Optimised LMS (FSTEQ) Every iteration step: B = FFT(b) Check if B lies between boundaries Scale if needed New b = IFFT(B) Conclusion Implementing TEQ doesn’t reduce error percentage, but does increase efficiency Still errors in received symbols: error coding is needed ADSL Equalization Van Herck Hans Verheyden Jonas QAM demod decoder S/P quadrature amplitude modulation (QAM) encoder mirror data and N-IFFT P/S D/A + transmit filter N-FFT and remove mirrored data S/P TRANSMITTER RECEIVER N/2 subchannels N real samples N real samples N/2 subchannels receive filter + A/D 00110 Bits P/S QAM demod decoder S/P quadrature amplitude modulation (QAM) encoder mirror data and N-IFFT add cyclic prefix P/S D/A + transmit filter N-FFT and remove mirrored data S/P remove cyclic prefix TRANSMITTER RECEIVER N/2 subchannels N real samples N real samples N/2 subchannels receive filter + A/D 00110 Bits channel P/S QAM demod decoder invert channel = FEQ S/P quadrature amplitude modulation (QAM) encoder mirror data and N-IFFT add cyclic prefix P/S D/A + transmit filter N-FFT and remove mirrored data S/P remove cyclic prefix TRANSMITTER RECEIVER N/2 subchannels N real samples N real samples N/2 subchannels receive filter + A/D 00110 Bits channel P/S QAM demod decoder invert channel = FEQ S/P quadrature amplitude modulation (QAM) encoder mirror data and N-IFFT add cyclic prefix P/S D/A + transmit filter N-FFT and remove mirrored data S/P remove cyclic prefix TRANSMITTER RECEIVER N/2 subchannels N real samples N real samples N/2 subchannels time domain equalizer (FIR filter) receive filter + A/D 00110 Bits