Enhanced Hybrid Underwater Target Tracking: A Virtual Array MIMO SONAR Framework with CA-EKF-TOMHT

Abstract

Tracking submerged targets in shallow-water acoustic environments poses severe operational challenges due to low Signal-to-Noise Ratio (SNR), ambient clutter, and range unobservability in passive bearings-only tracking. Conventional Single-Input Multiple-Output (SIMO) SONAR setups combined with Track-Oriented Multiple Hypothesis Tracking (TOMHT) often suffer from heavy computational loads because wide validation gates admit excessive clutter measurements. To address this gap, this paper investigates a hybrid target tracking framework combining a 6D Constant Acceleration Extended Kalman Filter with TOMHT (6D CA-EKF-TOMHT) integrated into a Virtual Array MIMO SONAR architecture ( ). The operational method employs a dual-phase strategy: covert passive bearings-only tracking ( ) followed by active range-bearing interrogation ( ). All reported quantitative performance metrics are evaluated as ensemble statistical averages across 100 independent Monte Carlo simulation runs in MATLAB R2024b. Virtual array synthesis expands the effective receiving aperture to 32 virtual elements, yielding an active signal processing gain of  (increasing SNR from  to ) and reducing measurement noise standard deviations by . Statistical evaluation confirms that the active position Root Mean Square Error (RMSE) decreases from  to  (  error reduction), while the average active track hypothesis counts drops from  to  (  tree reduction). The resulting contracted Mahalanobis validation gates suppress clutter hypothesis proliferation, accelerating processing latency from  to  (  speedup) under the simulated conditions.