TL;DR
This paper introduces a novel Sigmoid transform-based approach for parameter estimation in wideband bistatic MIMO radar systems operating in impulsive noise environments. By applying a Sigmoid transform, the method effectively suppresses impulsive noise without requiring prior knowledge of its characteristics, leading to improved accuracy in estimating target time delay, Doppler stretch, direction-of-departure, and direction-of-arrival.
What Problem Does This Paper Address?
Traditional radar parameter estimation methods often assume Gaussian noise, which is unrealistic in many practical scenarios where impulsive noise is prevalent. Existing methods designed for impulsive noise, such as those based on fractional lower-order statistics (FLOS), require prior knowledge of the noise's characteristic exponent, which is often unavailable. This leaves a gap in robust parameter estimation for wideband bistatic MIMO radar systems operating under such challenging noise conditions. The paper specifically targets the estimation of crucial parameters like time delay (TD), Doppler stretch (DS), direction-of-departure (DOD), and direction-of-arrival (DOA) in these environments.
Methods
The core of this paper lies in the introduction of two novel techniques:
- Sigmoid Wideband Ambiguity Function (Sigmoid-WBAF): This method extends the traditional Wideband Ambiguity Function (WBAF) by incorporating a Sigmoid transform. The Sigmoid transform is applied to the received signals, transforming the impulsive noise into a process with finite second-order moments. This allows the WBAF to effectively suppress impulsive noise interference and accurately estimate the time delay and Doppler stretch by locating the peak of the Sigmoid-WBAF.
- Sigmoid Correlation Function and Sigmoid-MUSIC Algorithm: To address the estimation of DOD and DOA, the authors propose a Sigmoid correlation function. This function, also leveraging the Sigmoid transform, provides a robust correlation measure in the presence of impulsive noise. This Sigmoid correlation is then used to develop a Sigmoid-MUSIC algorithm. This algorithm replaces the standard correlation matrices in the MUSIC algorithm with Sigmoid correlation matrices, enabling more accurate DOD and DOA estimation in impulsive noise.
The paper also derives the Cramér-Rao Bound (CRB) for parameter estimation to provide a theoretical benchmark for the performance of the proposed methods.
Datasets
Not discussed in the paper. The paper relies on simulated data for its experiments.
Key Insights and Ideas
- Sigmoid Transform as a Noise Suppressor: The Sigmoid transform is a powerful tool for converting impulsive noise (like alpha-stable distributions) into a process with finite second-order moments. This is crucial because many traditional radar signal processing techniques rely on these moments.
- No A Priori Noise Knowledge Required: A significant advantage of the proposed Sigmoid-based methods is that they do not require any prior knowledge about the characteristics of the impulsive noise, making them more practical for real-world applications.
- Unified Approach for Multiple Parameters: The paper presents a cohesive framework that addresses the estimation of both time delay/Doppler stretch (using Sigmoid-WBAF) and direction-of-departure/arrival (using Sigmoid-MUSIC), offering a comprehensive solution for wideband bistatic MIMO radar systems.
- Robustness to Impulsive Noise: The simulations clearly demonstrate that the proposed Sigmoid-based methods significantly outperform traditional WBAF and even FLOS-based methods in the presence of impulsive noise, especially at lower signal-to-noise ratios.
Key Results
- Improved Parameter Estimation Accuracy: Simulation results show that the Sigmoid-WBAF method achieves lower Root-Mean-Square Errors (RMSEs) for time delay and Doppler stretch estimation compared to WBAF and FLOS-WBAF, particularly under low GSNR and varying characteristic exponents.
- Accurate DOD/DOA Estimation: The Sigmoid-MUSIC algorithm demonstrates superior performance in estimating DOD and DOA compared to conventional MUSIC, FLOM-MUSIC, and lp-MUSIC algorithms, especially in highly impulsive noise environments.
- Comparable Complexity: The computational complexity of the proposed Sigmoid-WBAF and Sigmoid-MUSIC algorithms is comparable to existing methods (WBAF, FLOS-WBAF, MUSIC, FLOM-MUSIC), meaning the performance gains come without a significant computational cost increase.
- Theoretical Validation: The derived Cramér-Rao Bound (CRB) provides a theoretical lower bound on the estimation error, and the simulation results show that the proposed methods approach this bound, indicating their effectiveness.
Limitations and Future Work
- Parameter Selection for FLOS: The paper highlights that the performance of FLOS-based methods is sensitive to the choice of the fractional lower-order moment parameter 'p'. When 'p' is not appropriately chosen (e.g., not within the range related to the noise's characteristic exponent), FLOS methods can degrade significantly. The Sigmoid method, however, is less sensitive to this.
- Impulsiveness Threshold: While the Sigmoid transform is effective, the paper notes that for extremely high levels of impulsiveness, the suppression capacity might be limited compared to the ideal case. Future work could explore further enhancements for such extreme scenarios.
- Experimental Validation: The current study relies on simulations. Future work could involve experimental validation with real-world radar data to confirm the practical effectiveness of the proposed methods.
Summary updated