Investigation of least mean square adaptive algorithm to mitigate the non-linearity in power amplifier

Power amplifiers are necessary in all communication systems. Power amplifiers, in transmitters, operating at near maximum of their output power, exhibit non-linearity. This causes an increase in noise floor beyond signal bandwidth and may lead to interference in adjacent channels. Distortions caused by co-channel interference in the signal bandwidth reduce the Bit Error Rate (BER) at the receiver. One method of mitigating the non-linearity in power amplifiers (PA) is to lower the input signal to operate within the linear region. However, this method significantly reduces the power efficiency of the PA. Another method is to apply linearization techniques, such as Digital Pre-Distortion (DPD). DPD predicts the inverse of PA non-linearity and adds it to the input signal to eliminate non-linearity. DPD purpose is to make use of power amplifier in the non-linear range by increasing the output power and also the efficiency of the PA. Least Mean Squares (LMS) algorithm is chosen as it is simple to implement in FPGA and does not require complex mathematical computations. The Power Amplifier’s linear region of operation is extended by applying Digital Pre-Distortion which can be observed in reduction of error as the LMS algorithm converges. The performance of DPD is measured in terms of improvement in error vector magnitude and it is observed to be in range of 2dB to 10dB.Master of Science (Integrated Circuit Design