We present our view of an optimal implementation of EISCAT_3D transmission
modulations and data processing chain in plasma parameter observations. We
start from aperiodic transmission modulations that allow the same transmitted
pulses to be used for D, E, and F region observations, and lag profile
deconvolution by means of statistical inversion. We then continue with the
concept of multistatic incoherent scatter analysis that combines data from all
radar sites already in the plasma parameter fit step, and demonstrate
importance of this approach using data recorded with the existing EISCAT radars
and KAIRA. We discuss also possible implementation of Bayesian filtering in
EISCAT_3D data analysis and the requirements it would introduce to the plasma
parameter fit software. For each analysis step, we give also references to our
existing software tools, which could be used as starting points when developing
the EISCAT_3D data analysis tools.