ACPR prediction of CDMA systems through statistical behavioural modelling of power amplifiers with memory

Behavioural models for power amplifiers (PA) have traditionally been developed based on conventional AM-AM and AM-PM curves from one-tone measurements, which discounts the presence of memory in the system. However, as signal bandwidth increases in spread spectrum systems, memory effects become more severe. These effects result from the frequency response of matching networks, nonlinear capacitances of the transistors and the response of the bias networks. Multi-stage high power amplifiers require more accurate behavioural models to provide a better description of memory effects and highly nonlinear characteristics than memoryless models based on single-tone transfer characteristics. This is achieved using measured two-tone transfer characteristics of the amplitude and phase of the fundamental, IM3 and IM5 components, which have highly nonlinear components that represent the amplifier's significant memory. A statistical technique is then applied which allows accurate, fast and efficient prediction of the adjacent channel power ratio (ACPR) of the communication system without time-consuming time-domain simulations.