Orthogonal frequency division multiplexing phenomenology: Radar technique combining genetic algorithm-based pulse design and energy detector for target recognition

This study investigates the benefit offered by the orthogonal frequency division multiplexing (OFDM) structure to enhance the recognition functionality in OFDM radar. The authors propose a novel phenomenology-based framework whereby the pulse is composed of several symbols each matched to a particular target and the recognition function in the receiver reduces to an energy detector. In this preliminary study, they make use of simple target models for which they have exact analytical expressions and apply genetic algorithm-based methods to design the symbols. They emphasise on the need to work within a less undeterministic environment and demonstrate the merit of the authors' technique with simulated results. Finally, they show how the integration of multiple echoes improves significantly the overall probability of correct classification despite additive white Gaussian noise.