Robin Wentao Ouyang

Received Signal Strength-Based Wireless Localization via Semidefinite Programming: Noncooperative and Cooperative Schemes

The received signal strength (RSS)-based approach to wireless localization offers the advantage of low cost and easy implementability. To circumvent the nonconvexity of the conventional maximum likelihood (ML) estimator, in this paper, we propose convex estimators specifically for the RSS-based localization problems. Both noncooperative and cooperative schemes are considered. We start with the noncooperative RSS-based localization problem and derive a nonconvex estimator that approximates the ML estimator but has no logarithm in the residual. Next, we apply the semidefinite relaxation technique to the derived nonconvex estimator and develop a convex estimator. To further improve the estimation performance, we append the ML estimator to the convex estimator with the result by the convex estimator as the initial point. We then extend these techniques to the cooperative localization problem. The corresponding Cramer-Rao lower bounds (CRLB) are derived as performance benchmarks. Our proposed convex estimators comply well with the RSS measurement model, and simulation results clearly demonstrate their superior performance for RSS-based wireless localization.

Indoor Location Estimation with Reduced Calibration Exploiting Unlabeled Data via Hybrid Generative/Discriminative Learning

For indoor location estimation based on wireless local area networks fingerprinting, how to reduce the offline calibration effort while maintaining high location estimation accuracy is of major concern. In this paper, a hybrid generative/discriminative semi-supervised learning algorithm is proposed that utilizes a large number of unlabeled samples to supplement a small number of labeled samples. This hybrid method allows us to combine the modeling power and flexibility of generative models with the superior performance of discriminative approaches. Other related issues, such as learning efficiency enhancement and distribution estimation smoothing, are also discussed. Extensive experimental results show that our proposed method can effectively reduce the calibration effort and exhibit superior performance in terms of localization accuracy and robustness.

GPS Localization Accuracy Improvement by Fusing Terrestrial TOA Measurements

This paper explores the use of terrestrial time of arrival (TOA) measurements to improve the initial Global Positioning System (GPS) location fix accuracy. First, we present a geometric approach when a GPS location fix and one TOA measurement are available. Then, a more general hybrid GPS/TOA method via the Weighted Least Square Estimator (WLSE) is proposed. To simplify the calculation, a closed-form solution based on the two-step Least Square approach is also designed. The Cramer-Rao Lower Bound (CRLB) is derived as a performance benchmark. Simulation results exhibit excellent performance of the proposed methods which attain the CRLB in different scenarios. The proposed methods work even if only one TOA measurement (in addition to a GPS location fix) is available and the corresponding accuracy improvement (compared with the initial GPS location fix) can be as much as 30%.

Debiasing crowdsourced quantitative characteristics in local businesses and services

Information about quantitative characteristics in local businesses and services, such as the number of people waiting in line in a cafe and the number of available fitness machines in a gym, is important for informed decision, crowd management and event detection. In this paper, we investigate the potential of leveraging crowds as sensors to report such quantitative characteristics and investigate how to recover the true quantity values from noisy crowdsourced information. Through experiments, we find that crowd sensors have both bias and variance in quantity sensing, and task difficulties impact the sensing accuracy. Based on these findings, we propose an unsupervised probabilistic model to jointly assess task difficulties, ability of crowd sensors and true quantity values. Our model differs from existing categorical truth finding models as ours is specifically designed to tackle quantitative truth. In addition to devising an efficient model inference algorithm in a batch mode, we also design an even faster online version for handling streaming data. Experimental results in various scenarios demonstrate the effectiveness of our model.

Hybrid TOA/AOA-Based Mobile Localization with and without Tracking in CDMA Cellular Networks

This paper proposes a hybrid TOA/AOA (Time of Arrival/Angle of Arrival)-based localization algorithm for Code Division Multiple Access (CDMA) networks. The algorithm extends the Taylor Series Least Square (TS-LS) method originally developed for TOA-based systems to incorporate AOA measurements. In addition, tracking algorithms utilizing velocity and acceleration measurements are investigated. Simulation results illustrate that the proposed TOA/AOA TS-LS can provide better performance than conventional schemes in localization accuracy and in reduced likelihood of encountering non-convergence problem compared with TOA TS-LS. Tracking algorithms using the Extended and Unscented Kalman Filter (EKF and UKF) can track the objects relatively well, further decreasing the positioning error. UKF is found to provide closer tracking of the trajectory than EKF, for it truly captures the statistical mean and variance of the noises.

An Enhanced Toa-Based Wireless Location Estimation Algorithm for Dense NLOS Environments

Non-Line-Of-Sight (NLOS) signal propagation is the major source of error in conventional Time-Of-Arrival (TOA) based wireless location estimation algorithms. Previous research has mainly sought to address this problem in two ways: NLOS identification and NLOS mitigation. This paper focuses on the latter issue. It deals with the problem that even when NLOS measurements can be identified, among all the measurements obtained, there may still be not enough Line-Of-Sight (LOS) measurements for accurate location estimation using traditional TOA-based algorithms. With the assumptions that the total number of the measurements is greater than the minimum required and the NLOS measurements are identifiable, this paper proposes an enhanced TOA-based localization algorithm. It contains two parts: a combination stage and a Maximum Likelihood (ML) estimator. The proposed algorithm has an advantage that it does not require the information of the distribution of the NLOS bias. Simulation results show that the proposed algorithm outperforms all the other algorithms compared in dense NLOS environment.

Energy Efficient Assisted GPS Measurement and Path Reconstruction for People Tracking

In the use of a wearable GPS and cellular tracker for applications such as elderly tracking, device power consumption is an important consideration. To save power, assisted GPS (AGPS) location fixes should not be performed frequently. On the other hand, we also do not want to lose important information about the users mobility patterns and routines. To solve this dilemma, in this paper, we present the design of a system that intelligently schedules on-line AGPS location fixes only when necessary based on information extracted from users historical mobility data, and then reconstruct the user path based on these sparsely taken on-line location fixes. Experimental results show that our on-line algorithm can significantly reduce the number of AGPS fixes needed and the reconstruction method works well without a priori knowledge of a map and streets information.

Indoor Localization via Discriminatively Regularized Least Square Classification

In this paper, we address the received signal strength (RSS)-based indoor localization problem in a wireless local area network (WLAN) environment and formulate it as a multi-class classification problem using survey locations as classes. We present a discriminatively regularized least square classifier (DRLSC)-based localization algorithm that is aimed at making use of the class label information to better distinguish the RSS samples taken from different locations after proper transformation. Besides DRLSC, two other regularized least square classifiers (RLSCs) are also presented for comparison. We show that these RLSCs can be expressed in a unified problem formulation with a closed-form solution and convenient assessment of the convexity of the problem. We then extend the linear RLSCs to their nonlinear counterparts via the kernel trick. Moreover, we address the missing value problem, utilize clustering to reduce the training and online complexity, and introduce kernel alignment for fast kernel parameter tuning. Experimental results show that, compared with other methods, the kernel DRLSC-based algorithm achieves superior performance for indoor localization when only a small fraction of the data samples are used.

Received Signal Strength-Based Wireless Localization via Semidefinite Programming

Wireless localization has drawn significant attention over the past decade and the received signal strength (RSS) based localization scheme provides a low-cost, low-complexity and easy-implementation solution. When the statistics of the RSS measurement error is known, the Maximum Likelihood (ML) estimator is asymptotically optimal. However, due to the nature of the localization problem itself, the formed ML estimator is nonconvex, causing the search for the global minimum very difficult. In addition, its performance highly depends on the initial point provided if a local optimization method is applied to find the solution. To circumvent this problem, we apply the Semidefinite Programming (SDP) relaxation technique to the RSS-based localization problem. After reformulation and relaxation, we finally form a convex SDP estimator. A superior property of a convex estimator is that the solution is not affected by the initial point provided since any local minimum is also its global minimum. The Cramer-Rao Lower Bound (CRLB) is then derived as a benchmark for the performance comparison. Simulation results show that the proposed SDP estimator exhibit excellent performance in the RSS-based localization system and it is very suitable for the case when there are only very limited base stations hearable.

Parallel and Streaming Truth Discovery in Large-Scale Quantitative Crowdsourcing

To enable reliable crowdsourcing applications, it is of great importance to develop algorithms that can automatically discover the truths from possibly noisy and conflicting claims provided by various information sources. In order to handle crowdsourcing applications involving big or streaming data, a desirable truth discovery algorithm should not only be effective, but also be scalable. However, with respect to quantitative crowdsourcing applications such as object counting and percentage annotation, existing truth discovery algorithms are not simultaneously effective and scalable. They either address truth discovery in categorical crowdsourcing or perform batch processing that does not scale. In this paper, we propose new parallel and streaming truth discovery algorithms for quantitative crowdsourcing applications. Through extensive experiments on real-world and synthetic datasets, we demonstrate that 1) both of them are quite effective, 2) the parallel algorithm can efficiently perform truth discovery on large datasets, and 3) the streaming algorithm processes data incrementally, and it can efficiently perform truth discovery both on large datasets and in data streams.