Power and Linearity Optimization of a 24.2 GHz Transmitter Chain Through Behavioral Modeling and Look-Up Tables

Abstract

The optimization of the trade-off between power efficiency and linearity is a challenge in the design of millimeterwave (mmWave) transmitters for <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{5 G}$</tex> systems. This paper presents a systematic methodology for experimental characterization and subsequent optimization of a 24.2 GHz transmit chain through the development of a predictive behavioral model. A multidimensional, automated sweep of the chain's key control parameters—intermediate frequency (IF) input power, mixer gain, and IF stage attenuation—is performed to exhaustively map the operational space. For each configuration point, the RF output power and Error Vector Magnitude (EVM) are empirically determined. The extensive dataset is then used to construct a multi-dimensional look-up table (LUT) that functions as an accurate, data-driven behavioral model of the system. This model facilitates the prediction of transmitter performance and the identification of optimal operating points that maximize output power subject to a specific linearity constraint (EVM) without necessitating manual, iterative tuning. The utility of the model as a powerful design-time optimization tool is demonstrated, effectively decoupling performance tuning from the constraints of real-time hardware implementation.