Analysis of spectral regrowth of weakly nonlinear power amplifiers

Power amplifiers (PAs) are important components in a wireless communications system. Linearity and efficiency are two desirable but conflicting requirements of a PA. All PAs exhibit varying degrees of nonlinearity which causes spectral regrowth and adjacent channel interference. Based on the behavior model of the PA and statistics of the input, it is possible to infer spectral regrowth in the output. Existing analytical results on spectral regrowth assume that the input is Gaussian, which is unrealistic for many digital communications signals. We derive a closed form expression between the covariance function (and hence the power spectrum) of the output and statistics of the input, based on a polynomial behavior model of the PA, without assuming that the input is Gaussian. Such analytical results are useful in finding optimal operating conditions of the PA so that maximum efficiency is achieved without violating the out-of-band emission requirement.