5 | | ''NOTE: The WARPLab 7.5.0 MEX transport has been optimized for a single buffer read_iq:'' |
6 | | * `rx_IQ = wl_basebandCmd(node, [RF_RX], 'read_IQ', 0, rx_length);` |
7 | | ''vs'' |
8 | | * `rx_IQ = wl_basebandCmd(node, [RF_RX_0, RF_RX_1], 'read_IQ', 0, rx_length);` |
9 | | ''Currently, the iteration over the interfaces occurring in M code vs MEX C code. This will be fixed in future releases, but for now, if you do multiple buffer reads in a single Read IQ command, you will see a significant memory and performance penalty.'' |
| 5 | ''NOTE: The Ethernet performance benchmarks below were done on a single buffer read_iq:'' |
| 6 | * `rx_IQ = wl_basebandCmd(node, [RF_RX], 'read_IQ', 0, rx_length);` |
| 7 | |
| 8 | ''NOTE: Memory requirements will rise faster than the number of interfaces due to internal copying that Matlab performs during concatenation of arrays. In future releases, we will look at pushing the iteration over interfaces from M code to MEX C code to help with memory requirements.'' |