SIRE: A MIMO Radar for Landmine and IED Detection
This radar provides an efficient, cost-effective method of detecting landmines and IEDs.
Low-frequency ultra-wideband (UWB) radar has garnered attention for the detection of landmines and improvised explosive devices (IEDs) in recent years. The low frequencies used by these radars provide the necessary ground penetration capabilities for detection, and the wide bandwidth signals used are necessary for range resolution. Cross-range resolution that depends on the size of the antenna aperture can be improved by generating a synthetic aperture. Typical airborne synthetic aperture radars (SAR) that can provide high resolution in cross range are not practical for this problem due to cost limitations.
Multiple-input multiple-output (MIMO) radars can also be used to create a virtual array aperture larger than their single-input single-output (SIMO) counterparts, allowing for improved cross-range resolution. MIMO radars operate by using multiple antennas to transmit waveforms that could be linearly independent, and also use multiple antennas (receivers) to receive the reflected signals from targets in a given scene.
A MIMO radar with collocated antennas can provide advantages over its SIMO counterpart by exploiting waveform diversity. Some of the advantages afforded by this radar include improved parameter identifiability (i.e., maximum number of targets that can be uniquely identified), and improved cross-range resolution. This improved resolution can help resolve desired targets such as landmines from clutter.
The Synchronous Impulse Reconstruction (SIRE) ultra-wideband (UWB) radar was designed as a 2×16 (2 transmitters and 16 receivers) MIMO radar with collocated antennas. This radar operates in forward-looking mode and is built for landmine and IED detection. By transmitting orthogonal waveforms, improved cross-range resolution compared to using a single transmitter can be observed, showing this radar to be a working example of a MIMO radar. This radar employs cost-effective analog-to-digital (A/D) converters to sample its large signal bandwidth using an equivalent sampling scheme, making it practical for actual ground missions.
The use of low frequencies in UWB radar is necessary for foliage/ground penetration, whereas the use of UWB pulses is necessary for good resolution. Downrange resolution of this radar is determined by the bandwidth of the transmitted pulse that occupies a frequency range of 0.3 to 3 GHz. The crossrange resolution will be determined by the physical aperture of this radar. The conventional method for imaging is performed using the standard back-projection/ delay-and-sum (DAS) algorithm in forward-looking mode.
The SIRE radar system has a physical aperture (2 m) consisting of 16 receive antennas; 14 timing and control cards are also present to provide the necessary clock references for the radar. Each antenna consists of a digitizer that integrates the radar returns from a number of pulses that it passes to the system’s personal computer (PC), which acts as the operator control and display. The radar consists of two transmitters at the ends of the receive array. The returned radar signals collected from the 2D aperture can be used for imaging the scene. The images are formed 8 m (standoff range) ahead of the truck on which the radar is mounted.
Due to the large bandwidth of the returned radar signals, conventional sampling will require high-rate analog-todigital (A/D) converters to digitize the returned radar signals. These high-speed A/D converters are expensive to build and make practical implementation improbable. The goal was to develop a radar capable of landmine detection that is affordable and in a lightweight package for practical applications. Therefore, each of the receivers consists of a low-rate (40 MHz), commercially available A/D converter. The digitizers are used to sample the large bandwidth (≈ 3GHz) of the returned signals using an equivalent sampling scheme termed the SIRE sampling scheme.
This work was done by Lam Nguyen of the Army Research Laboratory; and Ode Ojowu Jr., Yue Wu, and Jian Li of the University of Florida, Gainesville. ARL-0187