Murat Efe

1 mJ pulse bursts from a Yb-doped fiber amplifier

We demonstrate burst-mode operation of a polarization-maintaining Yb-doped fiber amplifier capable of generating 60 μJ pulses within bursts of 11 pulses with extremely uniform energy distribution facilitated by a novel feedback mechanism shaping the seed of the burst-mode amplifier. The burst energy can be scaled up to 1 mJ, comprising 25 pulses with 40 μJ average individual energy. The amplifier is synchronously pulse pumped to minimize amplified spontaneous emission between the bursts. Pulse propagation is entirely in fiber and fiber-integrated components until the grating compressor, which allows for highly robust operation. The burst repetition rate is set to 1 kHz and spacing between individual pulses is 10 ns. The 40 μJ pulses are externally compressible to a full width at half-maximum of 600 fs. However, due to the substantial pedestal of the compressed pulses, the effective pulse duration is longer, estimated to be 1.2 ps.

An Adaptive Extended Kalman Filter with Application to Compartment Models

Abstract In this paper the ingestion and subsequent metabolism of a drug in a given individual, are investigated through the use of compartmental models. An adaptive formulation of the widely used extended Kalman filter (EKF) has been derived in order to solve the resulting nonlinear estimation problem at the output of the compartments. The adaptive EKF employs a forgetting factor to emphasize artificially the effect of current data by exponentially weighting the observations. The dependence of EKFs performance on the selection of appropriate values of the arbitrary matrices, the measurement covariance R and the process noise covariance Q has been demonstrated through simulations. With the appropriate choice of the matrices the EKF provides a very useful tool for online estimation of both the state and the parameters.

An adaptive Kalman filter with sequential rescaling of process noise

An approach to adaptive Kalman filtering is presented. The filter utilizes a scale factor, which represents the target unpredictability at any time, as estimated from the available data. The performance of the algorithm is compared with that of an IMM algorithm and also with that of a standard Kalman filter. The proposed filter does not rely on a priori knowledge about the target motion and it produces better estimates than the IMM algorithm during manoeuvring periods.

Stability of the Extended Kalman Filter When the States are Constrained

In this note, stability of the projection-based constrained discrete time extended Kalman filter (EKF) when applied to deterministic nonlinear systems has been studied. It is proved that, like the unconstrained case, under certain assumptions, the EKF with state equality constraints is an exponential observer, i.e., it keeps the dynamics of its estimation error exponentially stable. Also, it has been shown that a simple modification to the general definition of the EKF with exponential weighting increases the filters degree of stability and convergence speed with or without state constraints.

Improved assignment with ant colony optimization for multi-target tracking

Detecting and tracking ground targets is crucial in military intelligence in battlefield surveillance. Once targets have been detected, the system used can proceed to track them where tracking can be done using Ground Moving Target Indicator (GMTI) type indicators that can observe objects moving in the area of interest. However, when targets move close to each other in formation as a convoy, then the problem of assigning measurements to targets has to be addressed first, as it is an important step in target tracking. With the increasing computational power, it became possible to use more complex association logic in tracking algorithms. Although its optimal solution can be proved to be an NP hard problem, the multidimensional assignment enjoyed a renewed interest mostly due to Lagrangian relaxation approaches to its solution. Recently, it has been reported that randomized heuristic approaches surpassed the performance of Lagrangian relaxation algorithm especially in dense problems. In this paper, impelled from the success of randomized heuristic methods, we investigate a different stochastic approach, namely, the biologically inspired ant colony optimization to solve the NP hard multidimensional assignment problem for tracking multiple ground targets.

Multi-target tracking in clutter with histogram probabilistic multi-hypothesis tracker

This study presents a recently developed tracking algorithm, namely histogram probabilistic multi-hypothesis tracker (H-PMHT), a modified version of PMHT, for multi-target tracking. Even though the theory of H-PMHT could be easily extended to multi-dimensional case, its applications have only been realized for one-dimensional cases. In this work the theory of H-PMHT has been extended into two-dimensional case and its performance has been compared to that of interacting multi-model probabilistic data association filter (IMMPDAF) with amplitude information (IMMPDAF-AI). Simulation results reveal that H-PMHT algorithm outperforms the IMMPDAF-AI under various conditions explained in the following sections.

Radar Target Detection Using Hidden Markov Models

Standard radar detection process requires that the sensor output is compared to a predetermined threshold. The threshold is selected based on a-priori knowledge available and/or certain assumptions. However, any knowledge and/or assumptions become inadequate due to the presence of multiple targets with varying signal return and usually non stationary background. Thus, any fixed predefined threshold may result in either increased false alarm rate or increased track loss. Even approaches where the threshold is adaptively varied will not perform well in situations when the signal return from the target of interest is too low compared to the average level of the background. Track-before-detect techniques eliminate the need for a detection threshold and provide detecting and tracking targets with lower signal-to-noise ratios than standard methods. However, although track-before-detect techniques eliminate the need for detection threshold at sensor’s signal processing stage, they often use tuning thresholds at the output of the filtering stage. This paper presents a Hidden Markov Model based target detection method that avoids any thresholding at any stage of the detection process. Moreover, since the proposed Hidden Markov Model method is based on the target motion models, the output of the detection process can easily be employed for maneuvering target tracking.

Performance Evaluation of Track Association and Maintenance for a Mfpar with Doppler Velocity Measurements

This study investigates the efiects of incorporating Doppler velocity measurements directly into track association and maintenance parts for single and multiple target tracking unit in a multi function phased array radar (MFPAR). Since Doppler velocity is the major discriminant of clutter from a desired target, the measurement set has been expanded from range, azimuth and elevation angles to include Doppler velocity measurements. We have developed data association and maintenance part of a well known tracking method, Interacting Multiple Model Probabilistic Data Association

An Alternative Model for Target Position Estimation in Radar Processors

In this paper, we analyze the target position estimation errors induced by conventional radar signal processors, which assume a point-target model in matched filtering-based detection and tracking. As we demonstrated through simulations, the performance degradation under the point-target assumption can be significant for high-resolution radars, where targets extend across several detection cells. One of the main contributions of this paper is to provide an alternative model for reflections from extended targets and clutter. We model the events of backscatters from illuminated targets and clutter as a nonhomogenous Poisson process. The corresponding maximum likelihood estimator and the Cramer-Rao lower bound have been derived. Typical target detection process has been simulated and the validity of the model has been verified by using the Kullback-Leibler distance. It has been shown that the new method significantly reduces target position estimation errors compared to the peak picking method.

Peak picking losses in radar detectors

In this paper we analyze the peak picking losses induced by conventional radar signal processors, which assume a point target model for detection and tracking. As demonstrated through simulations, the performance degradation under the point target assumption can be significant for high-resolution radars, where targets extend across several detection cells. Interpolation of nearby data around the detected peak provides only a slight improvement. This paper presents a maximum likelihood estimator (MLE) to reduce the peak picking losses. By comparing the variance of the estimator with the Cramer Rao lower bound derived in this paper, it has been shown that the maximum likelihood estimator significantly reduces peak picking losses