Motivations to develop performance prediction for adaptive radar

Abstract

Performance prediction, or the capability to forecast system behavior, is a widespread technique used to advance the comprehension of how systems, and systems of systems, will react under certain circumstances and assumptions. In this paper, we discuss the motivating factors for development of performance prediction (2P) for radar. In the case of adaptive radar, the optimal unconstrained transmit waveform (for detection) is easily computed. However, it is generally not suitable for practical use. Therefore, we apply constraints on the waveform design and are forced to sacrifice signal-to-interference-and-noise ratio (SINR) to meet the constraints, i.e., there is no free lunch. Understanding the consequences of applying constraints in arbitrary waveform design can benefit the decision making process of an adaptive system by providing insight into selection of the transmit signal. We develop a use-case for 2P and give examples of successful 2P models. We discuss the trade-space between optimality and speed in the waveform design process. Lastly, we mention several current areas of promising research, selected results and comment on future needs to realize effective 2P for radar.