Currently the sniffer implementation is not very fast and loses a lot of packets. It should be improved. There is potentially three bottleneck in the current implementation:
The packets are printed as ASCII using printf, this is inefficient and might take a lot of CPU time
The USB bandwidth might not be enough
We can improve the 'off' time by not calculating the received packet power when getting the timestamp.
The first two problem can be addressed by implementing a binary protocol to send the packets over USB.
Eventually we should try the double buffering functionality of the DW1000 chip, lets keep that out of scope for this ticket since it will require much more work both in this project and in the libdw1000 project.
Currently the sniffer implementation is not very fast and loses a lot of packets. It should be improved. There is potentially three bottleneck in the current implementation:
The first two problem can be addressed by implementing a binary protocol to send the packets over USB.
Eventually we should try the double buffering functionality of the DW1000 chip, lets keep that out of scope for this ticket since it will require much more work both in this project and in the libdw1000 project.