Adaptive self-interference cancelation in LTE-A carrier aggregation FDD direct-conversion transceivers

Modern frequency division duplex radio frequency transceivers experience transmitter-to-receiver leakage due to the limited isolation of the duplexer. In Long Term Evolution-Advanced (LTE-A) carrier aggregation receivers the coupling between the local oscillators creates harmonics on the chip which can lead to the downconversion of this leakage signal to the receive (Rx) baseband. Thereby, this so-called modulated spur interference reduces the signal-to-noise ratio of the Rx signal. In this paper, the modulated spur interference is modeled and several adaptive algorithms are compared regarding their convergence and cancelation performance. To maximize the data throughput, the adaptive filter is required to converge within the short time period of one orthogonal frequency-division multiplexing (OFDM) symbol. Out of the investigated concepts the proposed variable step-size least-mean-square algorithm turns out to be the most favorable choice. It satisfies the required constraints of convergence time and cancelation performance, and it features a reasonable low complexity.