The traditional target tracking is a process of estimating the state of a moving target using measurement information obtained by sensors. However, underwater passive acoustic target tracking will confront further challenges, among which the system incomplete observability and time delay caused by the signal propagation create a great impact on tracking performance. Passive acoustic sensors cannot accurately obtain the target range information. The introduction of Doppler frequency measurement can improve the system observability performance; signal time delay cannot be ignored in underwater environments. It varies with time, which has a continuous negative impact on the tracking accuracy. In this paper, the Gauss–Helmert model is introduced to solve this problem by expanding the unknown signal emission time as an unknown variable to the state. This model allows the existence of the previous state and current state at the same time, while handling the implicit equations. To improve the algorithm accuracy, this paper further takes advantage of the estimated state and covariance for the second stage iteration and propose the Gauss–Helmert iterated Unscented Kalman filter under a three-dimensional environment. The simulation shows that the proposed method in this paper shows superior estimation accuracy and more stable performance compared with other filtering algorithms in underwater environments.

No abstract

No abstract

No abstract

In black-box adversarial attacks, attackers query the deep neural network (DNN) and use the query results to optimize the adversarial samples iteratively. In this paper, we study the method of adding white noise to the DNN output to mitigate such attacks. One of our unique contributions is a theoretical analysis of gradient signal-to-noise ratio (SNR), which shows the trade-off between the defense noise level and the attack query cost. The attacker's query count (QC) is derived mathematically as a function of noise standard deviation. This will guide the defender to find the appropriate noise level for mitigating attacks to the desired security level specified by QC and DNN performance loss. Our analysis shows that the added noise is drastically magnified by the small variation of DNN outputs, which makes the reconstructed gradient have an extremely low SNR. Adding slight white noise with a very small standard deviation, e.g., less than 0.01, is enough to increase QC by many orders of magnitude yet without introducing any noticeable classification accuracy reduction. Our experiments demonstrate that this method can effectively mitigate both soft-label and hard-label black-box attacks under realistic QC constraints. We also prove that this method outperforms many other defense methods and is robust to the attacker's countermeasures. INDEX TERMS deep learning, adversarial machine learning, black-box attack, noise perturbation, performance analysis VOLUME 4, 2016