WARP Project Forums - Wireless Open-Access Research Platform

One important clarification - the WARPLab MIMO OFDM example is a very different design from the OFDM Reference Design. These designs do not implement identical schemes. It can be useful to understand how both designs work, but you don't expect the implementations to be similar. There is no configuration of the OFDM Reference Design PHY (i.e. value for tx_controlBits) that will make it equivalent to the M code PHY implementation in the WARPLab example.

In warplab,after IFFT I,Q divided into two RFA and RFB ! but simulink don't divided.

In general a multiplexing MIMO OFDM receiver must compute one FFT per OFDM symbol per receive antenna. The WARPLab MIMO OFDM example does this explicitly with two calls to MATLAB's fft() function. The real-time PHY in the OFDM Reference Design uses a single FFT core to compute all transforms. The single FFT core is shared between receive antennas, alternating between samples from each antenna.

Why "Workspace Transmitter" in "ofdm_txrx_supermimo_coded" input only for RXA ?

When using the OFDM Reference Design PHY in simulation, you typically use the Tx PHY to generate a waveform for the Rx PHY. If you do not want to use the Tx PHY you can use the Workplace Transmitter to feed arbitrary waveforms into the Rx PHY. By default that block only has "From Workspace" blocks for the RF A receive antenna. You can add another for RF B if needed.

We do not have an M code implementation of the OFDM Reference Design PHY. This PHY was designed in Simulink/System Generator from scratch. You can simulate the PHY model in Simulink, but there is no M code equivalent.

If you're just getting started with System Generator, I would encourage you to use the Tx and Rx PHY models from the 802.11 Reference Design. These PHY designs are much newer than the old OFDM Reference Design. The 802.11 PHYs are also cleaner and will probably be easier to understand than the OFDM Reference Design PHY.