WARP Project Forums - Wireless Open-Access Research Platform

I have a small question about rx_length parameters. In reference, they said rx_length is Receive length- reads or sets the duration of each receive cycle, in sample periods

I'm not fully understand definition of receive cycle and sample periods.

In my mind, rx_length is the number of sample that programs read from buffer to computer via Ethernet.

Can you explain the definition of this parameter and how to set this parameter along with tx_length ?

I agree the term "rx cycle" here is ambiguous. In this case "cycle" means the full process of triggering a reception and capturing Rx samples. We called this a cycle since this process is typically repeated many times in an experiment.

The tx_length and rx_length parameters are integers which set the number of Tx/Rx samples. When you set rx_length, you are controlling the number of samples that will be recorded in the next reception. The duration of the capture is (rx_length / sampling frequency). The sampling frequency is fixed at 40MSps in the current WARPLab ref design.

So it means in each cycle this is the number of samples will be store in buffer? What happen if we set tx_length value larger than buffer size? It will divide transmitted signal into many parts?

The latest WARPLab Reference Designs use the WARP v3 node's DRAM for storing Tx/Rx samples. This allows very large buffers - refer to the user guide for specifics. The WARPLab m code enforces the maximum supported buffer size. You will get an error in MATLAB if you set tx_length or rx_length to an unsupported value.

And in here cycle = the full process of triggering a reception + rx_length / sampling frequency?

Yes.

Most WARPLab experiments follow the basic flow:

1) Initialize and configure nodes
2) Write Tx samples (write_iq)
3) Enable Tx/Rx modes
4) Send trigger
5) Disable Tx/Rx
6) Read Rx samples (read_iq)
7) Cleanup

Many experiments repeat the "Tx/Rx cycle" (i.e. steps 2-6) many times.