Constant Modulus Precoded MIMO Radar Based on Zadoff-Chu Sequences

Abstract

Multiple-input-multiple-output (MIMO) radar systems have become a heavily researched topic in recent years due to the improved diversity techniques when compared to that of 1-D radar waveforms. Although there exist a myriad of techniques to design the transmit power distribution for MIMO radar systems, stringent constant modulus power limitations imposed by modern high-power amplifiers (HPAs) complicate their practical implementation. Ideally, a technique that emulates the spatial filtering flexibility of precoded MIMO communications is desired. To achieve such flexibility, an MIMO radar framework must be introduced that guarantees constant modulus for all combinations of signal sets and precoders, allowing simple interchangeability between components of the waveform—whether that be a new transmit power distribution or an adjusted dynamic MIMO radar waveform. In this article, we show that designing a constant modulus precoded MIMO radar can be achieved in a practical manner, utilizing alphabet-based waveform construction. This technique leverages the use of two sets for which the product of any pair of vectors, one from each alphabet, guarantees a fixed constant. By utilizing these sets as alphabets to design the waveform, it is possible to implement MIMO radar waveforms of any rank and enable the decoupling of the precoder and MIMO radar waveform design. This article presents a framework that achieves the aforementioned requirements by utilizing the properties of Zadoff-Chu (ZC) sequences and the discrete Fourier transform (DFT) matrix. By restricting construction of the precoder and signal set to be in part formed from finite alphabets, it is shown that the constant modulus constraint is achieved for any precoder and any signal set combination.