Add lower sideband measurement
We currently only simulate one of the two sidebands, the upper one. In reality, we'll have both upper and lower sidebands.
The assumption is that including both sidebands will mostly result in a factor sqrt(2) for uncorrelated noises in them, which is currently accounted for by directly adjusting the noise level. However, to properly test we didn't miss anything, we should add the lower sideband to the simulation, which will also allow us to update the post-processing algorithms to make use of them.