In the field of asymptotic performance characterization conditional maximum-likelihood (CML) estimator, generally refers to either number samples or signal-to-noise ratio (SNR) value. The first case has been already fully characterized, although second only partially investigated. Therefore, this correspondence aims provide a sound proof result, i.e., (in SNR) Gaussianity and efficiency CML estimator in multiple parameters case, regarded as trivial but not so far demonstrated

It is well known that the ML estimator exhibits a threshold effect, i.e., rapid deterioration of estimation accuracy below certain signal-to-noise ratio (SNR) or number snapshots. This effect caused by outliers and not captured standard tools such as Cramer-Rao bound (CRB). The search SNR value (where CRB becomes unreliable for prediction maximum likelihood variance) can be achieved with help Barankin (BB), proposed many authors. major drawback BB, in comparison CRB, absence general...

The Cramér-Rao bounds (CRB) is a lower bound of great interest for system analysis and design in the asymptotic region [high signal-to-noise ratio (SNR) and/or large number snapshots], as it simple to calculate usually possible obtain closed form expressions. first part paper generalization complex parameters Barankin rationale deriving MSE bounds, that minimization norm under set linear constraints. With approach study Fisher information matrix (FIM) singularity, constrained CRB regularity...

We consider the classical radar problem of detecting a target in Gaussian noise with unknown covariance matrix. In contrast to usual assumption deterministic amplitudes, we assume here that latter are drawn from distribution. The generalized likelihood ratio test (GLRT) is derived based on multiple primary data and set secondary containing only. new GLRT shown be product Kelly's corrective, dependent term. also investigate two-step approaches where for known disturbance matrix first derived....

The derivation of tight estimation lower bounds is a key tool to design and assess the performance new estimators. In this contribution, first, authors derive compact Cramér–Rao bound (CRB) for conditional signal model, where deterministic parameter's vector includes real positive amplitude phase. Then, resulting CRB particularised delay, Doppler, phase, band-limited narrowband signals, which are found in plethora applications, making such broad interest. This expression particularly easy...

Ultra-Wide Band (UWB) communication systems can be used to design low cost, power efficient and precise navigation for mobile robots, by measuring the Time of Flight (ToF) messages traveling between on-board UWB transceivers infer their locations. Theoretically, decimeter level positioning accuracy or better should achievable, at least in benign propagation environments where Line-of-Sight (LoS) maintained. Yet, practice, even such favorable conditions, one often observes significant...

This contribution analyzes the fundamental performance limits of traditional two-step Global Navigation Satellite System (GNSS) receiver architectures, which are directly linked to achievable time-delay estimation performance. In turn, this is related GNSS baseband signal resolution, i.e., bandwidth, modulation, autocorrelation function, and sampling rate. To provide a comprehensive analysis standard point positioning techniques, we consider different GPS Galileo signals available, as well...

Abstract Images are an important source of information for spacecraft navigation. Based on image and a known attitude, triangulation techniques (intersection or resection) often used positioning In the resection problem, observer estimate its unknown location by using angle measurements to points at locations (i.e., landmarks), localization performance depending accuracy measurements. As contribution navigation, we considers dynamic estimation problem based radio interferometry, i.e., power,...

As aerial traffic becomes greater, it is more difficult to locate and recognise aircraft in the neighbourhood of civil airports. The technique proposed here resorts a particular device, monopulse radar, recent tool called independent component analysis (ICA) separate messages falling same radar beam. algorithms used compute ICA use fourthorder cumulants observed signals.

Monopulse is an established array processing technique for fast and accurate angle estimation. This has been generalized to space-time of any dimension. The statistical performance this monopulse parameter estimation characterized several target fluctuation models but not all cases. gap filled by paper. We derive the mean variance complex real imaginary part averaged ratio Swerling models, deterministic targets (Swerling 0 case), χ <sup xmlns:mml="http://www.w3.org/1998/Math/MathML"...

We consider Direction of Arrival (DoA) estimation in the case two closely spaced sources. In this case, most high resolution techniques fail to estimate DoAs if waveforms are highly correlated. Maximum Likelihood Estimators (MLE) known be more robust, but their excessive computational load limits use practice. paper, we propose an asymptotic approximation Unconditional (UML) procedure a Uniform Linear Array (ULA) and targets. This is based on asymptotically (in number observations)...

Summary Standard state estimation techniques, ranging from the linear Kalman filter (KF) to nonlinear extended KF (EKF), sigma‐point or particle filters, assume a perfectly known system model, that is, process and measurement functions noise statistics (both distribution its parameters). This is strong assumption which may not hold in practice, reason why several approaches have been proposed for robust filtering, mainly because performance particularly sensitive different model mismatches....

The derivation of estimation lower bounds is paramount to designing and assessing the performance new estimators. A lot effort has been devoted range-velocity problem, a fundamental stage on several applications, but very few works deal with acceleration, being key aspect in high dynamics applications. Considering generic band-limited signal formulation, we derive general compact form Cramér–Rao bound (CRB) expression for joint time-delay, Doppler stretch, acceleration estimation. This...

<h3>Abstract</h3> Global navigation satellite systems (GNSSs) play a key role in plethora of applications, ranging from and timing to Earth observation space weather characterization. For purposes, interference scenarios are among the most challenging operation conditions, with clear impact on maximum likelihood estimates (MLEs) signal synchronization parameters. While several mitigation techniques exist, an approach for theoretically analyzing GNSS MLE performance degradation under...

Abstract Time-delay and Doppler estimation is crucial in various engineering fields, as estimating these parameters constitutes one of the key initial steps receiver’s operational sequence. Due to its importance, several expressions Cramér–Rao Bound (CRB) Maximum Likelihood Estimation (MLE) have been derived over years. Previous contributions started from assumption that transmission process introduces an unknown phase, which hindered explicit consideration time-delay parameter carrier-phase...

A fairly general class of Bayesian "large-error" lower bounds the Weiss-Weinstein family, essentially free from regularity conditions on probability density functions support, and for which a limiting form yields generalized Cramér-Rao bound (BCRB), is introduced. In large number cases, BCRB appears to be Bobrovsky-Mayer-Wolf-Zakai (BMZB). Interestingly enough, regularized Bobrovsky-Zakai (BZB), applicable when support prior constrained parameter set, obtained. Modified BZB BMZB are...

For linear discrete state-space models, under certain conditions, the least-mean-squares filter estimate has a convenient recursive predictor/corrector format, aka Kalman filter. The purpose of this paper is to show that minimum variance distortionless response (MVDR) shares exactly same recursion, except for initialization which based on weighted least-squares estimator. If MVDR suboptimal in mean-squared error sense, it an infinite impulse (a deconvolver) does not depend prior knowledge...