Correction of demodulation errors based on machine learning
Abstract:
A method, a device, and a non-transitory storage medium having instructions to receive data and first channel information pertaining to a transmission of the data; store first constellation data and demodulate the data; determine whether any error exists pertaining to the data; receive retransmitted data and second channel information when an error exists; demodulate the retransmitted data; calculate error vector magnitude data; generate corrective constellation data based on the error vector magnitude data, the first channel information, and the second channel information, wherein the corrective constellation data includes at least one reference constellation point that is repositioned on a constellation plane relative to at least one corresponding reference constellation point of a default constellation data; and use the corrective constellation data when demodulating additional data that is subsequently received.
Public/Granted literature
Information query
Patent Agency Ranking
0/0