Since eNodeB can have information about channel quality with finer granularity through selected
subband CQI, wouldn’t it make sense to have MCSs defined on a per‐RB (or a group of RB) basis rather
than a single MCS per UE across all allocated RBs for a codeword?
