WARP Project Forums - Wireless Open-Access Research Platform

I'm a little confused on the description of the problem. To me, "channel hopping" refers to a single RF interface tuning to different RF frequencies over time. That isn't something that can be done on short time scales using the WARPLab Reference Design since changing channels requires interactions with the MATLAB instance on the host PC. But it doesn't sound like that's what you mean by channel hopping anyway. If you want to tune RFA, RFB, RFC, and RFD all to different channels then you can do that without modifying the WARPLab Reference Design. To achieve something similar to the behavior you are describing you could load up each Tx buffer with its own packet and stuff zeros during the times where the other buffers are transmitting. For example:

TX_A = [A, 0, 0, 0]
TX_B = [0, B, 0, 0]
TX_C = [0, 0, C, 0]
TX_D = [0, 0, 0, D]

You could then set each buffer to continuously transmit using the continuous_tx Tx command. If each interface is tuned to a different channel, you'd effectively achieve a rotating transmission of [A, B, C, D, A, B, C, ...] with each transmission being on its own channel. Note: one subtlety of this solution is that each transmitter is left enabled all the time. Transmitting 0-valued I/Q samples isn't the same thing as disabling transmission. There will be some LO leakage from your "inactive" transmitters since they are still actually enabled. I'm not sure what your application is, but if possible you might consider a real-time implementation using the 802.11 Reference Design. See the Frequency Hopping appnote for a description of how to modify the design to tune rapidly between different channels between transmissions. Since this design re-tunes a single RF interface to many different channels, it does not suffer the LO leakage issue of the proposed WARPLab solution.