9A_Comms_slides.ppt

Lab 9 - Performing Bit Error Rate Measurements BER Measurements Two types of BER Sinks: berMC and berIS ... Monte Carlo offers a generalized topology advantage, but takes longer ... IIS makes some assumptions but takes much less time ... Monte Carlo BER Sinks Two types: berMC and berMC4 Monte Carlo berMC Sink berMC only has only two inputs, Test and Reference measures the PE at the operating Es/No of the system the Es/No needs to be determined by other means, if necessary Monte Carlo berMC4 Sink berMC4 has two additional inputs, Signal and Noise Signal input – samples the signal without noise Noise input – samples the noise at the input of the system under test measures the PE and the operating Es/No of the system Monte Carlo (MC) BER Sink Parameters Start sampling in the middle of the first valid symbol NumThreshold defines the decision boundaries (e.g. One boundary at 0V for QPSK) Delay boundary to align Test and Ref signals in time (-1 for automatic synchronization off) Display BER only or BER vs. Time Automatically or Manually determine Threshold Boundry(s) Does this element control simulation run time? Yes/No How long does this sink collect data? Setting Up the berMC Sink The berMC just counts errors ... Ref data is the ideal non- filtered signal Data Symbol Splitter Filter Filter Modulator Up-converter/ Propagation Channel Receiver/ Demodulator Q Ref I Ref Q Test I Test Test is the demodulated signal Es/No = ? PE= 1 E-3 BerMC yields a single PE value at the actual operating Es/No of the system ... The actual Es/No of the system is not known Setting Up the berMC4 Sink berMC4 allows PE vs Es/No to be measured ... Data Symbol Splitter Filter Filter Modulator (noiseless) Up-converter/ Propagation Channel (noiseless) Receiver/ Demodulator Q Ref I Ref Q Test I Test Es/No BerMC4 uses a sample of the signal before noise is added and a sample of the noise injected to compute Es/No ... RF Gaussian Noise Source Splitter Inject noise here to set the Es/No into t