As automated classification systems become increasingly prevalent, concerns have emerged over their potential to reinforce and amplify existing societal biases. In the light of this issue, many methods have been proposed to enhance the fairness guarantees of classifiers. Most of the existing interventions assume access to group information for all instances, a requirement rarely met in practice. Fairness without access to demographic information has often been approached through robust optimization techniques, which target worst-case outcomes over a set of plausible distributions known as the uncertainty set. However, their effectiveness is strongly influenced by the chosen uncertainty set. In fact, existing approaches often overemphasize outliers or overly pessimistic scenarios, compromising both overall performance and fairness. To overcome these limitations, we introduce SPECTRE, a minimax-fair method that adjusts the spectrum of a simple Fourier feature mapping and constrains the extent to which the worst-case distribution can deviate from the empirical distribution. We perform extensive experiments on the American Community Survey datasets involving 20 states. The safeness of SPECTRE comes as it provides the highest average values on fairness guarantees together with the smallest interquartile range in comparison to state-of-the-art approaches, even compared to those with access to demographic group information. In addition, we provide a theoretical analysis that derives computable bounds on the worst-case error for both individual groups and the overall population, as well as characterizes the worst-case distributions responsible for these extremal performances.

Supervised learning with large-scale data usually leads to complex optimization problems, especially for classification tasks with multiple classes. Stochastic sub gradient methods can enable efficient learning with a large number of samples for classification techniques that minimize the average loss over the training samples. However, recent techniques, such as minimax risk classifiers (MRCs), minimize the maximum expected loss and are not amenable to stochastic sub gradient methods. In this paper, we present a learning algorithm based on the combination of constraint and column generation that enables efficient learning of MRCs with large-scale data for classification tasks with multiple classes. Experiments on multiple benchmark datasets show that the proposed algorithm provides upto a 10x speedup for general large-scale data and around a 100x speedup with a sizeable number of classes.

The Median-of-Means (MoM) is a robust estimator widely used in machine learning that is known to be (minimax) optimal in scenarios where samples are i.i.d. In more grave scenarios, samples are contaminated by an adversary that can inspect and modify the data. Previous work has theoretically shown the suitability of the MoM estimator in certain contaminated settings. However, the (minimax) optimality of MoM and its limitations under adversarial contamination remain unknown beyond the Gaussian case. In this paper, we present upper and lower bounds for the error of MoM under adversarial contamination for multiple classes of distributions. In particular, we show that MoM is (minimax) optimal in the class of distributions with finite variance, as well as in the class of distributions with infinite variance and finite absolute $(1+r)$-th moment. We also provide lower bounds for MoM's error that match the order of the presented upper bounds, and show that MoM is sub-optimal for light-tailed distributions.

Methods for split conformal prediction leverage calibration samples to transform any prediction rule into a set-prediction rule that complies with a target coverage probability. Existing methods provide remarkably strong performance guarantees with minimal computational costs. However, they require the use calibration samples composed by labeled examples different to those used for training. This requirement can be highly inconvenient, as it prevents the use of all labeled examples for training and may require acquiring additional labels solely for calibration. This paper presents an effective methodology for split conformal prediction with unsupervised calibration for classification tasks. In the proposed approach, set-prediction rules are obtained using unsupervised calibration samples together with supervised training samples previously used to learn the classification rule. Theoretical and experimental results show that the presented methods can achieve performance comparable to that with supervised calibration, at the expenses of a moderate degradation in performance guarantees and computational efficiency.

Boosting methods often achieve excellent classification accuracy, but can experience notable performance degradation in the presence of label noise. Existing robust methods for boosting provide theoretical robustness guarantees for certain types of label noise, and can exhibit only moderate performance degradation. However, previous theoretical results do not account for realistic types of noise and finite training sizes, and existing robust methods can provide unsatisfactory accuracies, even without noise. This paper presents methods for robust minimax boosting (RMBoost) that minimize worst-case error probabilities and are robust to general types of label noise. In addition, we provide finite-sample performance guarantees for RMBoost with respect to the error obtained without noise and with respect to the best possible error (Bayes risk). The experimental results corroborate that RMBoost is not only resilient to label noise but can also provide strong classification accuracy.

Accurate wireless localization is crucial for multiple real-world applications such as modern autonomy, Internet-of-Things, and smart cities. Machine learning (ML) methods can significantly improve localization performance, especially through soft information (SI) techniques. However, existing ML methods are designed to determine positions within the same environment (network) where training data are collected and cannot exploit data from environments with assorted types. In addition, the usage of existing approaches across multiple environments would significantly increase privacy risks, as they require transferring large amounts of data to a central server. This paper presents a SI-based method for wireless localization that can leverage information from multiple environments and adapt to the specific characteristics of each environment. Differently from existing techniques, we propose methods that ensure privacy by only transfering summary statistics of the training data. The performance of the proposed method is evaluated using real-world datasets obtained in multiple localization environments. The results show that the proposed method significantly outperforms existing techniques and provides adaptation to assorted environments together with privacy protection of location data.

Multiple supervised learning scenarios are composed by a sequence of classification tasks. For instance, multi-task learning and continual learning aim to learn a sequence of tasks that is either fixed or grows over time. Existing techniques for learning tasks that are in a sequence are tailored to specific scenarios, lacking adaptability to others. In addition, most of existing techniques consider situations in which the order of the tasks in the sequence is not relevant. However, it is common that tasks in a sequence are evolving in the sense that consecutive tasks often have a higher similarity. This paper presents a learning methodology that is applicable to multiple supervised learning scenarios and adapts to evolving tasks. Differently from existing techniques, we provide computable tight performance guarantees and analytically characterize the increase in the effective sample size. Experiments on benchmark datasets show the performance improvement of the proposed methodology in multiple scenarios and the reliability of the presented performance guarantees.

The accurate labeling of datasets is often both costly and time-consuming. Given an unlabeled dataset, programmatic weak supervision obtains probabilistic predictions for the labels by leveraging multiple weak labeling functions (LFs) that provide rough guesses for labels. Weak LFs commonly provide guesses with assorted types and unknown interdependences that can result in unreliable predictions. Furthermore, existing techniques for programmatic weak supervision cannot provide assessments for the reliability of the probabilistic predictions for labels. This paper presents a methodology for programmatic weak supervision that can provide confidence intervals for label probabilities and obtain more reliable predictions. In particular, the methods proposed use uncertainty sets of distributions that encapsulate the information provided by LFs with unrestricted behavior and typology. Experiments on multiple benchmark datasets show the improvement of the presented methods over the state-of-the-art and the practicality of the confidence intervals presented.

Accurate location awareness is essential for various context-based applications. This calls for efficient methodologies to collect, communicate and process position-dependent measurements, especially in situations with limited computational resources. The soft information (SI) approach has recently shown significant improvements in accuracy over conventional localization methods. By developing efficient SI-based techniques, it is possible to achieve higher precision also in case of stringent computational constraints. This paper proposes new SI-based localization techniques that utilize belief condensation and maximum entropy methods to reduce both communication burden and computational complexity. In addition, the techniques presented enable the use of generic sensing measurements, including those taking discrete and categorical values. Through two case studies involving time and angle measurements, we demonstrate how the proposed approach can significantly improve localization accuracy and computational efficiency.

Supervised learning is often affected by a covariate shift in which the marginal distributions of instances (covariates $x$) of training and testing samples $p_\text{tr}(x)$ and $p_\text{te}(x)$ are different but the label conditionals coincide. Existing approaches address such covariate shift by either using the ratio $p_\text{te}(x)/p_\text{tr}(x)$ to weight training samples (reweighted methods) or using the ratio $p_\text{tr}(x)/p_\text{te}(x)$ to weight testing samples (robust methods). However, the performance of such approaches can be poor under support mismatch or when the above ratios take large values. We propose a minimax risk classification (MRC) approach for covariate shift adaptation that avoids such limitations by weighting both training and testing samples. In addition, we develop effective techniques that obtain both sets of weights and generalize the conventional kernel mean matching method. We provide novel generalization bounds for our method that show a significant increase in the effective sample size compared with reweighted methods. The proposed method also achieves enhanced classification performance in both synthetic and empirical experiments.

For a sequence of classification tasks that arrive over time, it is common that tasks are evolving in the sense that consecutive tasks often have a higher similarity. The incremental learning of a growing sequence of tasks holds promise to enable accurate classification even with few samples per task by leveraging information from all the tasks in the sequence (forward and backward learning). However, existing techniques developed for continual learning and concept drift adaptation are either designed for tasks with time-independent similarities or only aim to learn the last task in the sequence. This paper presents incremental minimax risk classifiers (IMRCs) that effectively exploit forward and backward learning and account for evolving tasks. In addition, we analytically characterize the performance improvement provided by forward and backward learning in terms of the tasks’ expected quadratic change and the number of tasks. The experimental evaluation shows that IMRCs can result in a significant performance improvement, especially for reduced sample sizes.

High-dimensional data is common in multiple areas, such as health care and genomics, where the number of features can be tens of thousands. In such scenarios, the large number of features often leads to inefficient learning. Constraint generation methods have recently enabled efficient learning of L1-regularized support vector machines (SVMs). In this paper, we leverage such methods to obtain an efficient learning algorithm for the recently proposed minimax risk classifiers (MRCs). The proposed iterative algorithm also provides a sequence of worst-case error probabilities and performs feature selection. Experiments on multiple high-dimensional datasets show that the proposed algorithm is efficient in high-dimensional scenarios. In addition, the worst-case error probability provides useful information about the classifier performance, and the features selected by the algorithm are competitive with the state-of-the-art.

Supervised classification techniques use training samples to learn a classification rule with small expected 0-1 loss (error probability). Conventional methods enable tractable learning and provide out-of-sample generalization by using surrogate losses instead of the 0-1 loss and considering specific families of rules (hypothesis classes). This paper presents minimax risk classifiers (MRCs) that minimize the worst-case 0-1 loss with respect to uncertainty sets of distributions that can include the underlying distribution, with a tunable confidence. We show that MRCs can provide tight performance guarantees at learning and are strongly universally consistent using feature mappings given by characteristic kernels. The paper also proposes efficient optimization techniques for MRC learning and shows that the methods presented can provide accurate classification together with tight performance guarantees in practice.

Location-awareness has become a fundamental requirement for multiple emerging applications with the rapid development of wireless technologies. The high-accuracy ranging enabled by ultra-wide bandwidth (UWB) signals is often deteriorated by clocks imperfections and non-line-of-sight (NLOS) propagation. Existing supervised learning methods for NLOS identification and mitigation are time-consuming, labor-intensive, and cost-inefficient due to the need for training data acquisition and label assignment. This paper presents an indoor localization system that enables NLOS mitigation based on self-training. The system provides a general information fusion framework that integrates map, inertial sensors, and UWB measurements, where the weak labels for UWB measurements are produced and iteratively refined by multi-sensory information fusion for self-training. In addition, the system utilizes the maximum likelihood ranging estimator that considers the impact of clock drift. The effectiveness of the proposed system is demonstrated via extensive experimentation in multiple real-world environments, e.g., the proposed methods reduce the NLOS ranging error by 80% and result in a 90th localization error percentile of 0.5 meters in a complex indoor environment.

Wireless propagated signals encapsulate rich information for high-accuracy localization and environment sensing. However, the full exploitation of positional and environmental features as well as their correlation remains challenging in complex propagation environments. In this paper, we propose a methodology of variational inference over deep neural networks for concurrent distance estimation and environmental identification. The proposed approach, namely inter-instance variational auto-encoders (IIns-VAEs), conducts inference with latent variables that encapsulate information about both distance and environmental labels. A deep learning network with instance normalization is designed to approximate the inference concurrently via deep learning. We conduct extensive experiments on real-world datasets and the results show the superiority of the proposed IIns-VAE in both distance estimation and environmental identification compared to conventional approaches.

Deep generative models (DGMs) and their conditional counterparts provide a powerful ability for general-purpose generative modeling of data distributions. However, it remains challenging for existing methods to address advanced conditional generative problems without annotations, which can enable multiple applications like image-to-image translation and image editing. We present a unified Bayesian framework for such problems, which introduces an inference stage on latent variables within the learning process. In particular, we propose a variational Bayesian image translation network (VBITN) that enables multiple image translation and editing tasks. Comprehensive experiments show the effectiveness of our method on unsupervised image-to-image translation, and demonstrate the novel advanced capabilities for semantic editing and mixed domain translation.

The statistical characteristics of instance-label pairs often change with time in practical scenarios of supervised classification. Conventional learning techniques adapt to such concept drift accounting for a scalar rate of change by means of a carefully chosen learning rate, forgetting factor, or window size. However, the time changes in common scenarios are multidimensional, i.e., different statistical characteristics often change in a different manner. This paper presents adaptive minimax risk classifiers (AMRCs) that account for multidimensional time changes by means of a multivariate and high-order tracking of the time-varying underlying distribution. In addition, differently from conventional techniques, AMRCs can provide computable tight performance guarantees. Experiments on multiple benchmark datasets show the classification improvement of AMRCs compared to the state-of-the-art and the reliability of the presented performance guarantees.

The maximum entropy principle advocates to evaluate events’ probabilities using a distribution that maximizes entropy among those that satisfy certain expectations’ constraints. Such principle can be generalized for arbitrary decision problems where it corresponds to minimax approaches. This paper establishes a framework for supervised classification based on the generalized maximum entropy principle that leads to minimax risk classifiers (MRCs). We develop learning techniques that determine MRCs for general entropy functions and provide performance guarantees by means of convex optimization. In addition, we describe the relationship of the presented techniques with existing classification methods, and quantify MRCs performance in comparison with the proposed bounds and conventional methods.

The overall global death rate for COVID-19 patients has escalated to 2.13% after more than a year of worldwide spread. Despite strong research on the infection pathogenesis, the molecular mechanisms involved in a fatal course are still poorly understood. Machine learning constitutes a perfect tool to develop algorithms for predicting a patient’s hospitalization outcome at triage. This paper presents a probabilistic model, referred to as a mortality risk indicator, able to assess the risk of a fatal outcome for new patients. The derivation of the model was done over a database of 2,547 patients from the first COVID-19 wave in Spain. Model learning was tackled through a five multistart configuration that guaranteed good generalization power and low variance error estimators. The training algorithm made use of a class weighting correction to account for the mortality class imbalance and two regularization learners, logistic and lasso regressors. Outcome probabilities were adjusted to obtain cost-sensitive predictions by minimizing the type II error. Our mortality indicator returns both a binary outcome and a three-stage mortality risk level. The estimated AUC across multistarts reaches an average of 0.907. At the optimal cutoff for the binary outcome, the model attains an average sensitivity of 0.898, with a 0.745 specificity. An independent set of 121 patients later released from the same consortium attained perfect sensitivity (1), with a 0.759 specificity when predicted by our model. Best performance for the indicator is achieved when the prediction’s time horizon is within two weeks since admission to hospital. In addition to a strong predictive performance, the set of selected features highlights the relevance of several underrated molecules in COVID-19 research, such as blood eosinophils, bilirubin, and urea levels.

Received waveforms contain rich information for both range information and environment semantics. However, its full potential is hard to exploit under multipath and non-line-of-sight conditions. This paper proposes a deep generative model (DGM) for simultaneous range error mitigation and environment identification. In particular, we present a Bayesian model for the generative process of the received waveform composed by latent variables for both range-related features and environment semantics. The simultaneous range error mitigation and environment identification is interpreted as an inference problem based on the DGM, and implemented in a unique end-to-end learning scheme. Comprehensive experiments on a general Ultra-wideband dataset demonstrate the superior performance on range error mitigation, scalability to different environments, and novel capability on simultaneous environment identification.

Different types of training data have led to numerous schemes for supervised classification. Current learning techniques are tailored to one specific scheme and cannot handle general ensembles of training samples. This paper presents a unifying framework for supervised classification with general ensembles of training samples, and proposes the learning methodology of generalized robust risk minimization (GRRM). The paper shows how current and novel supervision schemes can be addressed under the proposed framework by representing the relationship between examples at prediction and training via probabilistic transformations. The results show that GRRM can handle different types of training samples in a unified manner, and enable new supervision schemes that aggregate general ensembles of training samples.

Supervised classification techniques use training samples to find classification rules with small expected 0-1 loss. Conventional methods achieve efficient learning and out-of-sample generalization by minimizing surrogate losses over specific families of rules. This paper presents minimax risk classifiers (MRCs) that do not rely on a choice of surrogate loss and family of rules. MRCs achieve efficient learning and out-of-sample generalization by minimizing worst-case expected 0-1 loss w.r.t. uncertainty sets that are defined by linear constraints and include the true underlying distribution. In addition, MRCs' learning stage provides performance guarantees as lower and upper tight bounds for expected 0-1 loss. We also present MRCs' finite-sample generalization bounds in terms of training size and smallest minimax risk, and show their competitive classification performance w.r.t. state-of-the-art techniques using benchmark datasets.

Recent advancements in signal processing and communication systems have resulted in evolution of an intriguing concept referred to as Internet of Things (IoT). By embracing the IoT evolution, there has been a surge of recent interest in localization/tracking within indoor environments based on Bluetooth Low Energy (BLE) technology. The basic motive behind BLE-enabled IoT applications is to provide advanced residential and enterprise solutions in an energy efficient and reliable fashion. Although recently different state estimation (SE) methodologies, ranging from Kalman filters, Particle filters, to multiple-modal solutions, have been utilized for BLE-based indoor localization, there is a need for ever more accurate and real-time algorithms. The main challenge here is that multipath fading and drastic fluctuations in the indoor environment result in complex non-linear, non-Gaussian estimation problems. The paper focuses on an alternative solution to the existing filtering techniques and introduces/discusses incorporation of the Belief Condensation Filter (BCF) for localization via BLE-enabled beacons. The BCF is a member of the universal approximation family of densities with performance bound achieving accuracy and efficiency in sequential SE and Bayesian tracking. It is a resilient filter in harsh environments where nonlinearities and non-Gaussian noise profiles persist, as seen in such applications as Indoor Localization.

Location awareness is vital for emerging Internet-of-Things applications and opens a new era for Localization-of-Things. This paper first reviews the classical localization techniques based on single-value metrics, such as range and angle estimates, and on fixed measurement models, such as Gaussian distributions with mean equal to the true value of the metric. Then, it presents a new localization approach based on soft information (SI) extracted from intra- and inter-node measurements, as well as from contextual data. In particular, efficient techniques for learning and fusing different kinds of SI are described. Case studies are presented for two scenarios in which sensing measurements are based on: 1) noisy features and non-line-of-sight detector outputs and 2) IEEE 802.15.4a standard. The results show that SI-based localization is highly efficient, can significantly outperform classical techniques, and provides robustness to harsh propagation conditions.

The month’s regular papers issue covers machine learning at the wireless network edge, soft-informationbased localization techniques, and Antenna-in-Package technology.

Network navigation, encompassing both spatial and temporal cooperation to locate mobile agents, is a key enabler for numerous emerging location-based applications. In such cooperative networks, the positional information obtained by each agent is a complex compound due to the interaction among its neighbors. This information coupling may result in poor performance: algorithms that discard information coupling are often inaccurate, and algorithms that keep track of all the neighbors' interactions are often inefficient. In this paper, we develop a principled framework to characterize the information coupling present in network navigation. Specifically, we derive the equivalent Fisher information matrix for individual agents as the sum of effective information from each neighbor and the coupled information induced by the neighbors' interaction. We further characterize how coupled information decays with the network distance in representative case studies. The results of this paper can offer guidelines for the development of distributed techniques that adequately account for information coupling, and hence enable accurate and efficient network navigation.

High-accuracy localization in harsh environments is a challenging research problem, mainly due to non-line-of-sight (NLOS) propagation, multipath effect, and multiuser interference. Many techniques have been proposed to address this problem; most of them focus on improving the accuracy of ranging estimation, e.g., NLOS identification and mitigation. In this paper, we take ranging one step further by introducing the concept of ranging likelihood (RL), showing that RL is the essential element for localization. Moreover, we present effective techniques for real-time RL estimation. We focus on ultra-wide bandwidth (UWB) localization systems and assess the performance of the proposed approach by using the data from an extensive indoor measurement campaign. The results show that the proposed approach can significantly improve the performance of wireless localization in harsh environments.

Accuracy and complexity represent fundamental aspects of localization and tracking systems. In this manuscript the impact of a priori knowledge about agent position on the accuracy and the complexity of localization algorithms is investigated. In particular, first Cramer-Rao bounds on localization accuracy are derived under the assumption that a priori information is described by a map restricting the agent position to a specific region. Then, the computational complexity of optimal map-aware and map-unaware localization techniques is assessed. Our results evidence that: a) map-aware localization accuracy can be related to some geometrical features of the map but usually exhibits a complicated dependence on them; b) in some scenarios map-aware localization algorithms provide better accuracy than their map-unaware counterparts at comparable computational complexity.

Establishing bounds on the accuracy achievable by localization techniques represents a fundamental technical issue. Bounds on localization accuracy have been derived for cases in which the position of an agent is estimated on the basis of a set of observations and, possibly, of some a priori information related to them (e.g., information about anchor positions and properties of the communication channel). In this paper, new bounds are derived under the assumption that the localization system is map-aware, i.e., it can benefit not only from the availability of observations, but also from the a priori knowledge provided by the map of the environment where it operates. Our results show that: a) map-aware estimation accuracy can be related to some features of the map (e.g., its shape and area) even though, in general, the relation is complicated; b) maps are really useful in the presence of some combination of low SNRs and specific geometrical features of the map (e.g., the size of obstructions); c) in most cases, there is no need of refined maps since additional details do not improve estimation accuracy.

The availability of reliable positional information is a key enabler for numerous emerging location-based applications. Network navigation via joint spatial and temporal cooperation can provide mobile nodes with high-accuracy and robust positional information. Meanwhile, this joint cooperation incurs intricate information acquisition due to the correlation in inferred nodes' position, referred to as information coupling. In this paper, we quantify the information coupling in four representative scenarios by Fisher information analysis. We show that the information obtained by each node is a sum of the contribution from its own spatio-temporal cooperation and information coupling due to the cooperation of its neighbors. Our results shed lights on the complex information acquisition in network navigation, and can serve as a design guideline for efficient network navigation algorithms.

The success of location-based services in open areas, mainly driven by Global Navigation Satellite Systems (GNSS), has pushed the research community towards the development of robust systems that provide a similar localization accuracy in harsh environments, where there is no line-of-sight (LOS) to satellites. Techniques based on time-of-arrival (TOA) measurements are the most promising since they can offer an appropriate balance between accuracy and complexity. However, their performance is commonly compromised by the knowledge of an accurate model relating TOA measurements and distance, which can involve an arduous task of calibration. In this paper, we present a method for real-time self-calibration of the model relating TOA and distance, based on TOA measurements exchanged among the anchors in the wireless network. Simulation and empirical results show the appropriateness of the derived models for a TOA-based localization system, since the obtained errors are remarkably close to the ones achieved with pre-calibrated parameters.

In cooperative localization networks, nodes estimate their locations through inter-node ranging, location information exchange, and information fusion. These operations cause packet collision, large communication overhead, and high computational complexity, which can be alleviated by proper scheduling techniques. In this paper, we design distributed scheduling algorithms for cooperative localization networks, which improve the efficiency of localization through neighbor selection and collision control. Then, we determine the evolution of each node's location error through cooperation and movement using Fisher information analysis. Furthermore, we analyze the convergence of the location error under the proposed scheduling algorithm. Simulation results show that the efficiency of localization is significantly improved by using the proposed scheduling algorithm.

Real-time and reliable location information of mobile nodes is a key enabler for many emerging wireless network applications. Such information can be obtained via network navigation, a new paradigm in which nodes exploit both spatial and temporal cooperation to infer their positions. In this paper, we establish a theoretical foundation for network navigation and determine the fundamental limits of navigation accuracy using equivalent Fisher information analysis. We then introduce the notion of carry-over information and provide a geometrical interpretation for the evolution of navigation information. Our framework unifies the navigation information obtained from spatial and temporal cooperation, leading to a deep understanding of information evolution and cooperation benefits in navigation networks.

Real-time navigation capability is a key enabler for many emerging applications in wireless networks. Localization of moving nodes via network navigation gives rise to a new paradigm, where nodes exploit both temporal and spatial cooperation to determine their positions based on intra- and inter-node measurements. In this paper, we establish a theoretical foundation for network navigation to determine the fundamental limits of navigation accuracy. In particular, we derive the accuracy limits in terms of navigation information by equivalent Fisher information analysis. Our framework unifies the navigation information obtained from temporal and spatial cooperation, leading to a deep understanding of information exchange in the network and benefit of cooperation.

Traditional techniques for navigation such as the Kalman filter cannot capture the nonlinear and non-Gaussian models appearing in wireless localization systems deployed in harsh environments. Nonparametric filters as particle filters can cope with such models at the expense of a computational complexity beyond the reach of low-cost navigation devices. In this paper, we establish a general framework for parametric filters based on belief condensation (BC), which can express highly nonlinear and non-Gaussian system and measurement models. Our methodology exploits the specific structure of the problem and decomposes it in such a way that the linear and Gaussian part can be solved efficiently. The set of parameters for the posterior distribution is updated by an optimization process, referred to as BC. The simulation results show that the performance of the proposed parametric filter is close to that of the particle filter, but with a much lower complexity.

Several sources of error disturb the range estimates utilized to infer the position of a mobile user (MU). This paper presents a method which includes two stages of error reduction for a round-trip time (RTT)-based indoor location system. First, systematic error due to non-line of sight (NLOS) propagation is mitigated by adapting the prior-NLOS- measurements-correction (PNMC) technique to the input of a particle filter (PF). Thereupon, the latter is used to filter out the random error in the range estimates which is not assumed to be Gaussian distributed, on the contrary to what is commonly used in the literature. Both techniques are described and their performance is assessed in a real indoor environment achieving a reduction of more than 50% in the total error.

The hybrid localization techniques have attracted significant research interest since a variety of localization metrics can be easily obtained by most wireless devices. This paper presents a hybrid localization scheme that combines received signal strength (RSS) and round-trip time (RTT) information. It is based on a RSS ranging technique that dynamically update the model that best fit the RSS information to the actual distance. RTT information, among other heuristics, are used to refine the search of that model. Once distances have been estimated, the position of the mobile station (MS) is estimated using a trilateration technique that combines the RSS and RTT ranging estimates after applying a median filter to remove outliers. By means of simulations and measurements, this paper demonstrates that combining RSS and RTT information it is possible to outperform the conventional RSS-based and RTT-based localization schemes, without using either a tracking technique or a previous calibration stage of the environment.

Prior to the trilateration process, different sources of error disturb the range estimates in any localization system, especially in dense cluttered environments where the non-line-of-sight (NLOS) between the mobile user (MU) and the access points (AP) becomes the main problem. Common error mitigation approaches are based on linear or Gaussian assumptions that are not fulfilled in this kind of scenarios, such as indoor or dense urban outdoor areas. This paper points out the better performance of a round-trip time (RTT)-based localization system when combining a prior NLOS measurements correction (PNMC) method with particle RTT-only tracking, since hard decisions are only made in the last stage of the positioning process, and neither linear nor Gaussian models are assumed. The final performance leads to a reduction of more than 57% of the error obtained without any mitigation technique, keeping the requirement of no calibration stage.

Due to their availability and cost, Received Signal Strength (RSS) based techniques are the most spread localization schemes when a person who carries an RSS meter is going to be located. However, this techniques are subject to errors associated with perturbations of the fields by the presence of the human body. Although these alterations are complex they are not completely unpredictable. This paper presents a few simple case studies to assess the E-field strength errors caused by the presence of the human body on RSS based localization schemes in a theoretical and experimental approach. It can be concluded that in a common but conservative worst case scenario this error could reach up to 15 dB and therefore it is of concern for analysis. Also, useful insights into the overall problem based on finite-difference time-domain (FDTD) simulations are given.

The addition of positioning capabilities to widespread communications such as IEEE 802.11 compliant networks could open up interesting markets, in particular if a low-cost hardware has to be added to the existing one and standard RTS/CTS control frames exchange could be used for this purpose. In this paper various statistical estimators of the delay profile observed, derived from Round-Trip Time measurements, are analyzed and a linear regression of the statistical estimators is applied in order to improve the accuracy of distance estimation between a Mobile User and an Access Point. A precision in distance estimation with errors in the range of a meter is achieved.

The main drawback to improve the precision in cellular location systems based on time delays is the presence of non-line-of-sight (NLOS) propagation, introducing significant and hardly predictable errors in the measurements used in location. In this paper we propose a technique to previously detect and estimate the NLOS ratio present in the measurements. A comprehensive research on the Time of Arrival (TOA) dependence on the NLOS ratio existing in the measurements leads to estimate that ratio from the deviation calculated directly from the available measurements. NLOS detection and ratio estimate is very useful to classify and weight the measurements available in order to perform the following location estimate with the most accurate ones. Several simulations have been conducted to show the precision of ratio estimate and the measurements rating based on it.