This paper presents a study on target track accuracy improvement for air-to-air (A/A) radar. A Multi-hypothesis Post-processing (MHPP) approach is proposed for improving the air target track accuracy. The MHPP approach consists of a Target Maneuver Detector (TMD) and a Target Maneuver-adapted Track Smoother (TMATS). TMD applies a simple tracking filter to the Kalman Filter (KF) outputs for making a decision on the presence of target maneuver. The TMD’s decision is then utilized to guide TMATS to improve the overall track accuracy. TMATS is a filterbank that takes into account multiple hypotheses on the target maneuvering status. In particular, TMATS is constructed with multiple smoothing/tracking filters, each of which is dedicated to a different target maneuvering scenario. The final track outputs are selected from a particular TMATS component according to the target maneuvering status. Monte Carlo simulations are conducted to demonstrate the effectiveness of the proposed MHPP approach. The robustness of the proposed MHPP approach against the degree of target maneuvering is also verified with simulations.
This paper presents a study on the target position and velocity estimation for air-to-air radar. Two types of post-Kalman filtering (post-KF) processing methods, weighted least-square (WLS) estimators and tracking filters, are investigated to improve the filtering accuracy and thus to improve the target position and velocity estimation accuracy. The WLS estimators are considered up to the second order in order to effectively handle target acceleration. For this type of post- KF processing, the KF outputs within a sliding window are passed through a WLS estimator to refine the estimates of current target position and velocity and their variances. For the tracking filter method, an alpha-beta (α-β) filter and an alpha-beta-gamma (α-β-γ) filter are utilized; both dynamically smooth the KF outputs. For the linear motion scenarios where the target flies with constant velocity, the first-order WLS estimator and the α-β filter are expected to be a good fit, which permits accurate projection of the target position to a future time of interest. The second-order WLS estimator and the α-β-γ filter are capable of handling more general scenarios where the targets may be accelerating. The performance and effectiveness of these proposed post-KF processing methods are demonstrated by using Monte Carlo simulations.
In this paper, target position and velocity estimation is investigated for track declaration using an air-to-air radar. A post-Kalman filtering processing method is proposed to improve the filtering accuracy and thus to improve the target position
and velocity estimation accuracy. The proposed method passes the outputs of the Kalman filters (KFs) within a sliding
window through a weighted least squares (WLS) estimator to refine the estimates of current target position and velocity
and their variances. It is therefore referred to as the post-KF-WLS method. The post-KF-WLS estimates of the current
target position and velocity are utilized to project the target position in a future time of interest. The uncertainty of the
target position projection is derived and a closed-form solution is formulated. The effectiveness of the proposed method
is demonstrated by using Monte Carlo simulations. Impacts of contributing factors to the target position projection
uncertainty are quantified via simulations and the dominating factor is identified as well.
BAE Systems recently developed a rotorcraft brownout landing aid system technology (BLAST) to satisfy the urgent
need for brownout landing capability. BLAST uses a W-band monopulse (MP) radar in conjunction with radar signal
processing and synthetic display techniques to paint a three-dimensional (3-D) perspective of the landing zone (LZ) in
real time. Innovative radar signal processing techniques are developed to process the radar data and generate target data
vectors for 3-D image synthesis and display. Field tests are conducted to characterize the performance of BLAST with
MP and non-MP (only using the sum channel of the MP radar) modes in clear and brownout conditions. Data processing
and analysis are performed to evaluate the system's performance in terms of visual effect, signal-to-noise ratio (SNR),
target height estimation, ground-mapping effect, and false alarm rate. Both MP and non-MP modes reveal abilities to
sufficiently display the 3-D volume of the LZ; the former shows advantage over the latter in providing accurate ground
mapping and object height determination.
KEYWORDS: Target detection, Radar, Signal to noise ratio, Extremely high frequency, Super resolution, Image resolution, Signal processing, Image enhancement, Antennas, Resolution enhancement technologies
In this paper, two-dimensional (2D) (range and azimuth) resolution enhancement is investigated for millimeter wave
(mmW) real-beam radar (RBR) with linear or non-linear antenna scan in the azimuth dimension. We design a new
architecture of super resolution processing, in which a dual-mode approach is used for defining region of interest for 2D
resolution enhancement and a combined approach is deployed for obtaining accurate location and amplitude estimations
of targets within the region of interest. To achieve 2D resolution enhancement, we first adopt the Capon Beamformer
(CB) approach (also known as the minimum variance method (MVM)) to enhance range resolution. A generalized CB
(GCB) approach is then applied to azimuth dimension for azimuth resolution enhancement. The GCB approach does not
rely on whether the azimuth sampling is even or not and thus can be used in both linear and non-linear antenna scanning
modes. The effectiveness of the resolution enhancement is demonstrated by using both simulation and test data. The
results of using a 94 GHz real-beam frequency modulation continuous wave (FMCW) radar data show that the overall
image quality is significantly improved per visual evaluation and comparison with respect to the original real-beam radar