Abstract:

Efficiency and linearity of the microwave power amplifier
are critical elements for mobile communication systems. A
memory polynomial baseband predistorter based on an indirect
learning architecture is presented for improving the linearity of
an envelope tracing (ET) amplifier with application to a wireless
transmitter. To deal with large peak-to-average ratio (PAR) problem,
a clipping procedure for the input signal is employed. Then
the system performance is verified by simulation results. For a
single carrier wideband code division multiple access (WCDMA)
signal of 16-quadrature amplitude modulation (16-QAM), about
2% improvement of the error vector magnitude (EVM) is achieved
at an average output power of 45.5 dBm and gain of 10.6 dB,
with adjacent channel leakage ratio (ACLR) of -64.55 dBc at offset
frequency of 5 MHz. Moreover, a three-carrier WCDMA signal and
a third-generation (3G) long term evolution (LTE) signal are used
as test signals to demonstrate the performance of the proposed
linearization scheme under different bandwidth signals.