This article discusses the reduction of nonlinearities in analog-to-digital (A/D) converters using digital signal processing (DSP). Also modeling of certain essential nonlinearities is considered in detail. The main focus is on wideband radio receivers, such as the emerging cognitive radio applications, where a collection of signals at different frequency channels is converted to digital domain as a whole. Therefore, the overall dynamic range can easily be in the order of tens of dBs and thus even mild nonlinear distortion can cause strong carriers to block weaker signal bands. In this article, a mathematical model for clipping distortion due to improper input signal conditioning is derived through Fourier analysis. Additionally, stemming from the analysis an adaptive DSP-based post-processing method for reducing the effects of clipping and integral nonlinearity (INL) in A/D converters is presented with illustrative examples using both computer simulations and laboratory radio signal measurements.
