Adding channel compensation dual stream correlation and layer demapping for nr IA decoder

OAI gNb-CI build (376): Some Changed files in Merge Request DO NOT follow OAI Formatting Rules

OAI gNb-CI build (376): Some modified files in Merge Request MAY have INTRODUCED WARNINGS (openair1/PHY/NR_UE_TRANSPORT/nr_dlsch_demodulation.c)

OAI gNb-CI build (376): failed (https://open5glab.eurecom.fr:8083/jenkins/job/gNb-CI/376/)

why do we need layer_llr? We already have llr, which can hold 2 layers. We can easily extend this to 4.

Hi Hongzhi, we should take this opportunity to get rid of "mimo_mode" and "beamforming_mode" which does not make sense anymore in NR.

Hi Florian, I agree with you. But it's not yet clear to me which modes are supported for mimo in NR. I will have a look at it. If it's clear to you, we can discuss. I think llr is for codeword. So I created layer_llr which is mapped to to one codeword when layers are less than 4.

there is only one transmission mode in NR. one codeword can use multiple layers, but I don't think there is a reason to create extra memory for that (we already have a very big memory footprint). Can't we do the layer unmapping at the same time as the llr computation?

added 75 commits

• f421e730...67e1afd3 - 74 commits from branch develop-nr
• ff388661 - Merge remote-tracking branch 'origin/develop-nr' into integration_nr_IA_receiver

Compare with previous version

OAI gNb-CI build (402): Some Changed files in Merge Request DO NOT follow OAI Formatting Rules

OAI gNb-CI build (402): Some modified files in Merge Request MAY have INTRODUCED WARNINGS (openair1/PHY/INIT/nr_init_ue.c openair1/PHY/NR_UE_TRANSPORT/nr_dlsch_demodulation.c)

OAI gNb-CI build (402): failed (https://open5glab.eurecom.fr:8083/jenkins/job/gNb-CI/402/)

merged