Sub-band selection at cellular base station for non-overlapped or partially overlapped full duplex operation
摘要
說明書
In at least some embodiments, machine learning may be applied. For example, an artificial intelligence (AI) module may be trained to recognize a self-interference channel response based on certain inputs, such as an operational frequency and/or frequency range, a sub-band frequency and/or bandwidth, received UL signal levels and/or bandwidths, environmental conditions, time of day, day of week, and the like. The AI module, once suitably trained, e.g., according to a training set, may be employed to predict a favorable sub-band. Alternatively or in addition, the AI module may be trained to recognized receiver level adjustments or settings according to transmitter operational power levels and/or sub-band width and/or frequency range. Once trained, the AI module may be employed to predict a receiver level adjustment give operational parameters of the collocated transmitter, e.g., its power level and/or sub-band width and/or frequency range.
According to the illustrative sub-band selection process 270, a transmit frequency sub-band is allocated at 272 based on self-interference channel. Allocation of the sub-band may be based on one or more of an absolute minimum value and/or relative minimum value(s) in the self-interference channel 248, bandwidth(s) and/or frequency range(s) of receiver sub-band and/or sub-bands, application(s) supported by the full-duplex communications, subscription levels, channel availability, channel assignments, and so on.
