In this work, we propose a set-theoretic approach to collaborative position location for wireless networks. The proposed method borrows the concept from the parallel projection method (PPM), originally developed for signal recovery with inconsistent convex feasibility sets, modifies and extends the technique to an iterative and distributed numerical algorithm to estimate node locations, based on incomplete and noisy internode distance estimates. We demonstrate that in the case of noncollaborative position location, the proposed method is analytically equivalent to the parallel implementation of Kaczmarz Algorithm that is guaranteed to converge to a local minimizer and thus a stationary point. For collaborative position location, the proposed iterative PPM is computationally much more efficient than existing methods such as SDP and MDS-MAP, while achieving comparable or better localization accuracy and robustness to non-line-of-sight (NLOS) bias. Finally, our proposed method can be implemented in a parallel and distributed fashion, and is scalable for large network deployment.

In this work, we propose a set-theoretic approach to collaborative position location for wireless networks. The proposed method borrows the concept from the parallel projection method (PPM), originally developed for signal recovery with inconsistent convex feasibility sets, revises and extends the technique to an iterative and distributed numerical framework to estimate node locations, based on incomplete and noisy inter-node distance estimates. We demonstrate that the proposed iterative PPM is computationally much more efficient than existing methods, while achieving comparable and often better localization accuracy and robustness to non-line-of-sight (NLOS) bias. Our proposed method can be implemented in a parallel and distributed fashion, and is scalable for large network deployment.

Indoor positioning has become a hot research topic due to a plethora of interesting applications ranging from emergency responder tracking to location-based services. In this work we focus on the problem of network localization also sometimes called collaborative localization where a network of nodes is to be localized using both connections to anchors (when they exist) and connections between unlocalized nodes [1]. Although several algorithms have been investigated in the literature (e.g., [2]), most have assumed line-of-sight (LOS) propagation which is uncommon in most indoor environments. Although the impact of non-LOS (NLOS) propagation has been considered in the literature (e.g., [3], [4]), this has typically been limited to traditional single-node localization such as location estimation in cellular systems. Specifically, in this work we propose a technique for collaborative position location designed to effectively handle NLOS propagation and demonstrate the performance improvement possible over existing methods.

In this paper, we investigate the optimal performance of collaborative position location. In particular, we develop a branch-and-bound (BB) solution search strategy, coupled with the reformulation linearization technique (RLT), to solve the maximum likelihood estimation (MLE) problem for collaborative position location, which is in general a nonlinear and nonconvex optimization problem. Compared with existing work which has only approximately solved the MLE problem, our approach is guaranteed to produce the (1 - ¿)-optimal solution to the MLE for arbitrarily small ¿. With a guaranteed optimal solution to the MLE, we show that for some node geometries in noncollaborative position location, which can be viewed as a special case of collaborative position location, the Cramer-Rao lower bound (CRLB) for an unbiased estimator is no longer a meaningful performance benchmark. We demonstrate that the timeof- arrival (TOA) based MLE is in general a biased estimator and it sometimes has a mean square error (MSE) smaller than the CRLB, and thus can serve as a more practical performance benchmark. Finally, we compare the MLE with some existing position location schemes and demonstrate that it also serves as a good performance benchmark for collaborative position location.

We present a face recognition method based on sparse representation for recognizing 3D face meshes under expressions using low-level geometric features. First, to enable the application of the sparse representation framework, we develop a uniform remeshing scheme to establish a consistent sampling pattern across 3D faces. To handle facial expressions, we design a feature pooling and ranking scheme to collect various types of low-level geometric features and rank them according to their sensitivities to facial expressions. By simply applying the sparse representation framework to the collected low-level features, our proposed method already achieves satisfactory recognition rates, which demonstrates the efficacy of the framework for 3D face recognition. To further improve results in the presence of severe facial expressions, we show that by choosing higher-ranked, i.e., expression-insensitive, features, the recognition rates approach those for neutral faces, without requiring an extensive set of reference faces for each individual to cover possible variations caused by expressions as proposed in previous work. We apply our face recognition method to the GavabDB and FRGC 2.0 databases and demonstrate encouraging results.

In ad hoc position-location networks, location information is obtained through the sequential estimation of node locations. An unlocalized node can estimate its location based on range and location estimates of nearby localized (";anchor";) nodes, and subsequently provide range and location information to other unlocalized nodes in its vicinity. In such distributed location-estimation scenarios, the accuracy of node location estimates is degraded due to the propagation of localization errors, particularly in NLOS propagation environments. This paper discusses the use of a novel method of mitigating the propagation of localization error using the linear programming framework proposed by the authors in [1]. This method utilizes multi-hop distance estimates to create additional constraints on the feasible region for a node's location, thereby limiting the propagation of error even in NLOS propagation environments.

In this paper, we propose a collaborative localization scheme which utilizes, in addition to the range estimates to nodes with known locations (anchors), the range estimates between un- localized nodes to estimate their positions. The proposed collaborative method is incorporated into the linear programming (LP) framework proposed in Venkatesh, S. and Buehrer, R. M., (2006) and is implemented in a distributed fashion. We also adopt a constrained nonlinear optimization method to deal with the degenerate cases in the LP approach, i.e., when an unlocalized node has less than three line-of-sight (LOS) range estimates to localized nodes. Simulation results suggest that this collaborative quasi-LP framework can improve the localization accuracy as well as increase the percentage of nodes that are able to be localized, as compared to the standard sequential least-squares (LS) estimator.

A new M-ary orthogonal coded signaling is introduced to avoid the inter-frame interference that is especially detrimental to realizing high rate ultra-wideband (UWB) transmitted-reference (TR) systems. To further increase the information rate, the inter-pulse interference by an overlap of multipath-delayed pulses is controlled by integrating the signaling and a pair of balanced matched Alters in a joint manner, so as to permit a shorter time delay between the reference and data pulses in TR systems. To evaluate an achievable information rate increase relative to conventional TR, the symbol error probability (SEP) is theoretically derived for the proposed M-ary orthogonal coded/balanced TR system, considering the realistic IEEE standard UWB channel models. In addition, we consider the issue of receiver complexity and present two alternative low- complexity receiver implementations for the proposed TR system.

Recently, a novel balanced transmitted-reference (TR) system has been proposed for UWB communications. This TR system is capable of properly eliminating the inter-pulse interference (IPI) between the reference and data pulses in a single user multipath environment. In this paper, we investigate its multiple access (MA) performance by evaluating the second- order moments of noise and interference terms. The analytical framework developed here can also be used to accurately evaluate the MA performance of conventional TR system. It is shown that the proposed balanced TR system can achieve comparable MA capacity as conventional TR, while operating at higher information rate. On the other hand, given a target information rate, the balanced TR system in conjunction withM-ary signaling offers higher MA capacity than conventional TR system. The performance results suggest there exists a trade-off between the inter-pulse distance, thus the achievable information rate, and the MA capacity.

In this paper, we analyze the multiuser performance of a modified version of a newly proposed balanced transmitted-reference (TR) system for ultra-wideband (UWB) communications. The proposed TR system is capable of properly eliminating the inter-pulse interference (IPI) between the reference and data pulses in multipath environments, with a low system complexity. We further investigate the effect of the inter-pulse distance on the multiuser capacity and demonstrate the trade-off between the system's achievable information rate and the number of coexisting users.

A new M-ary orthogonal coded/balanced signaling is proposed to realize high rate UWB transmitted-reference (TR) systems. Especially, the inter-pulse interference by an overlap of multipath-delayed pulses is mitigated by combining the signaling and a pair of balanced matched filters in a joint manner, so as to permit a shorter time delay between the reference and data pulses in TR systems. To evaluate an achievable information rate increase relative to conventional TR, the symbol error probability (SEP) is theoretically derived for the M-ary balanced TR system in dense multipath.