On the Cancellation of Modulated Spurs in 4G Transceivers via Advanced Learning based Architectures
Sprache des Vortragstitels:
Englisch
Original Tagungtitel:
Eurocast 2022
Sprache des Tagungstitel:
Englisch
Original Kurzfassung:
Due to the limited transmitter-to-receiver stop-band isola- tion of the duplexers in modern long term evolution (LTE) and LTE- adcanced (LTE-A) frequency division duplex transceivers, leakage signals may appear in the receiver (Rx). These leakage signals are the cause of self interference (SI) in the Rx paths, diminishing the receiver?s sensi- tivity. One type of such SI signals are modulated spurs, which occur in downlink carrier-aggregation in LTE-A transceivers. This work proposes an advanced learning architecture (AL-Arch), based on concepts from machine learning, to combat this type of interference. It will be shown that the proposed solution not only out-preforms current state-of-the-art (SOTA) models, but also requires less complexity to do so. Further, un- like SOTA architectures, the AL-Arch can deal with varying allocation patterns which might prove difficult for traditional methods.
Forschungs-