MIMO radar array thinning optimization exploiting almost difference sets

Thinning techniques based on almost difference sets (ADSs) are presented to design MIMO radar antenna arrays with low peak sidelobe levels (PSLs). The standard ADS approach allows one to analytically determine the element locations of the transmitting and receiving arrays and estimate the upper bound of the radiation pattern sidelobes of MIMO virtual arrays. The peak sidelobe performances of ADS-based MIMO arrays turn out to be comparable with those of cyclic difference sets (CDSs) based MIMO arrays, at the same time obtainable in a wider set of array configurations. In order to overcome the limitations of purely ADS-based methodology in terms of flexibility and sidelobe performances, a genetic algorithm (GA)-enhanced ADS-based procedure is also proposed. Representative numerical results are illustrated to point out the features and performances of the proposed ADS approaches, and to validate their effectiveness and efficiency in comparison with other similar design methodologies.