Hey all,
I’ve been playing around with GMSK and reading through the code of the
processing blocks. I see that the dot product code for the Gaussian
filter is optimized such that it requires the data be 16-byte aligned.
Where does this alignment actually occur in the processing chain?
On the modulation path, the flow is NRZ->GF->FM:
gr_bytes_to_syms()->gr_firdes::gaussian()->gr_frequency_modulator_fc()
I’m assuming the data coming out of gr_bytes_to_syms() is not 16-byte
aligned, yet the data that comes in to the Gaussian filter is 16-byte
aligned. So, where does this alignment actually occur?
Thanks!
George