You are not logged in.
Pages: 1
Hi, I have a modulation problem by running WARPLAB 6 on warp v3 board.
I used OFDM with Binary-ASK before and it works well. Then I tried to add QAM modulation to my OFDM codes(Just use QAM in payloads and keep rest pilots same as before), then I have some interesting findings.
First, if I just use 4 QAM, it works well, especially after I added convolution code. But when I use 8 QAM or 16 QAM, it's not right any more. At these cases, when I use convolution code, the BER just get worse. Take 8 QAM for example, when all payload is 0 then the symbol after 8 QAM is 0(in the range 0-7), which is -3+1i in my method. But at receiver, all 0 changed to 2(in the range 0-7) and the signal is like -0.5125 + 0.1439i, -0.3325 + 0.2623i, -0.3343 + 0.2563i, etc. I guess in some sense, transmitter or receiver cannot recognize 3 in the amplitude. That's why 4 QAM works well, because 4QAM just use +/- 1 +/- i.
I'm wondering is it related to the bitstream used in FPGA, or even some hardware configuration? Can you tell me what should I to solve this problem?
Thanks!
For your reference, I use QAM methods in website: http://www.mathworks.com/help/comm/gs/c … atlab.html . It's qammod and qamdemode in Matlab comm toolbox.
Offline
Does your OFDM code implement channel estimation and equalization? What about carrier frequency offset estimation/correction? Modulation schemes which encode information in the phase of the complex baseband signal will require equalization/phase correction at the receiver.
I would suggest studying our SISO OFDM example for WARPLab 7: wl_example_siso_ofdm_txrx.m. Most of the code is not specific to WARPLab (in fact, set USE_WARPLAB_TXRX = 0 to use the code in simulation with no WARP hardware). This example implements a SISO OFDM Tx and Rx PHY inspired by the 802.11 a/g waveform. The channel training, equalization, CFO estimation and phase correction processes would be useful building blocks for your own OFDM design.
Offline
Pages: 1