Zixi Jia

Grid-Based Improved Maximum Likelihood Estimation for Dynamic Localization of Mobile Robots

The dynamic localization is a kind of technology by which the mobile robot tries to localize the position by itself. According to the dynamic localization failure of mobile robots in indoor network blind areas, an autonomous-dynamic localization system which dynamically chooses beacon node and establishes grids is proposed in this paper. This method applies received signal strength indication (RSSI) for distance measurement. Furthermore, the proposed grid-based improved maximum likelihood estimation (GIMLE) fulfills the localization. Finally, the localization error correction is implemented by Kalman filter. The approach combines the classical Kalman filter with the other localization algorithms. The purpose is to smooth and optimize the results of the algorithms, in order to improve the localization accuracy. In particular, in network blind spots, the Kalman filter provides better performance than the other algorithms listed in the paper. Experimental results show the accuracy, adaptivity, and robustness of the dynamic self-localization of mobile robots.

Compound Event Barrier Coverage in Wireless Sensor Networks under Multi-Constraint Conditions

It is important to monitor compound event by barrier coverage issues in wireless sensor networks (WSNs). Compound event barrier coverage (CEBC) is a novel coverage problem. Unlike traditional ones, the data of compound event barrier coverage comes from different types of sensors. It will be subject to multiple constraints under complex conditions in real-world applications. The main objective of this paper is to design an efficient algorithm for complex conditions that can combine the compound event confidence. Moreover, a multiplier method based on an active-set strategy (ASMP) is proposed to optimize the multiple constraints in compound event barrier coverage. The algorithm can calculate the coverage ratio efficiently and allocate the sensor resources reasonably in compound event barrier coverage. The proposed algorithm can simplify complex problems to reduce the computational load of the network and improve the network efficiency. The simulation results demonstrate that the proposed algorithm is more effective and efficient than existing methods, especially in the allocation of sensor resources.

Research of Directed Spanning Tree Routing Protocol for Wireless Sensor Networks

In wireless sensor networks (WSN), two major challenges are to maximize network lifetime and to enhance the tolerance of networks. Both of them are related to routing protocol. Therefore, a routing protocol that is called directed spanning tree (DST) is proposed. In this paper, the construction of directed tree-based communication path and the data aggregation mechanism based on DST are discussed. For overcoming the frangibility of traditional routing protocols during nodes failure, a directed tree reconstruction scenario is also explored. Furthermore, the proposed protocol is evaluated in comparison with the existing protocols from the effect of energy saving to the failure-tolerance. With the simulation, the performance of DST is verified to be efficient and available.

An Indoor Mobile Location Estimator in Mixed Line of Sight/Non-Line of Sight Environments Using Replacement Modified Hidden Markov Models and an Interacting Multiple Model.

Localization as a technique to solve the complex and challenging problems besetting line-of-sight (LOS) and non-line-of-sight (NLOS) transmissions has recently attracted considerable attention in the wireless sensor network field. This paper proposes a strategy for eliminating NLOS localization errors during calculation of the location of mobile terminals (MTs) in unfamiliar indoor environments. In order to improve the hidden Markov model (HMM), we propose two modified algorithms, namely, modified HMM (M-HMM) and replacement modified HMM (RM-HMM). Further, a hybrid localization algorithm that combines HMM with an interacting multiple model (IMM) is proposed to represent the velocity of mobile nodes. This velocity model is divided into a high-speed and a low-speed model, which means the nodes move at different speeds following the same mobility pattern. Each moving node continually switches its state based on its probability. Consequently, to improve precision, each moving node uses the IMM model to integrate the results from the HMM and its modified forms. Simulation experiments conducted show that our proposed algorithms perform well in both distance estimation and coordinate calculation, with increasing accuracy of localization of the proposed algorithms in the order M-HMM, RM-HMM, and HMM + IMM. The simulations also show that the three algorithms are accurate, stable, and robust.

The Indoor Localization and Tracking Estimation Method of Mobile Targets in Three-Dimensional Wireless Sensor Networks

Indoor localization is a significant research area in wireless sensor networks (WSNs). Generally, the nodes of WSNs are deployed in the same plane, i.e., the floor, as the target to be positioned, which causes the sensing signal to be influenced or even blocked by unpredictable obstacles, like furniture. However, a 3D system, like Cricket, can reduce the negative impact of obstacles to the maximum extent and guarantee the sensing signal transmission by using the line of sight (LOS). However, most of the traditional localization methods are not available for the new deployment mode. In this paper, we propose the self-localization of beacons method based on the Cayley–Menger determinant, which can determine the positions of beacons stuck in the ceiling; and differential sensitivity analysis (DSA) is also applied to eliminate measurement errors in measurement data fusion. Then, the calibration of beacons scheme is proposed to further refine the locations of beacons by the mobile robot. According to the robot’s motion model based on dead reckoning, which is the process of determining one’s current position, we employ the H∞ filter and the strong tracking filter (STF) to calibrate the rough locations, respectively. Lastly, the optimal node selection scheme based on geometric dilution precision (GDOP) is presented here, which is able to pick the group of beacons with the minimum GDOP from all of the beacons. Then, we propose the GDOP-based weighting estimation method (GWEM) to associate redundant information with the position of the target. To verify the proposed methods in the paper, we design and conduct a simulation and an experiment in an indoor setting. Compared to EKF and the H∞ filter, the adopted STF method can more effectively calibrate the locations of beacons; GWEM can provide centimeter-level precision in 3D environments by using the combination of beacons that minimizes GDOP.

Based on Adaptive Variable Rate Particle Filter Tracking Research in Wireless Sensor Networks

Because the state model is too simple to capture the complicated trajectory, the traditional tracking algorithm can not track the moving people precisely. Simon J. Godsil and others improve particle filter to present variable rate particle filter, which employs more complicated model to track moving target, and samples state variables by changing sampling period. In this paper, we propose an adaptive variable rate particle filter algorithm for large-scale building fire rescue system to track trapped people. The algorithm can adaptively adjust the sampling period by comparing the relationship between particle value and measurement. According to our computer simulation, the method can improve the tracking accuracy, especially as the target trajectory has some drastic changes. we consider that the tracking algorithm is effective and practical.

Research of Balanced-Alternative Spanning Tree for Wireless Sensor Networks

In wireless sensor networks (WSNs), lifetime and failure-tolerance seem to be a part of contradiction. The traditional spanning tree mostly focuses on network lifetime, and use data-aggregation to save network energy. In order to keep the whole network available during nodes failure, a routing protocol that is called balanced-alternative spanning tree (BAST) is proposed. This routing protocol has two different strategies in the normal case and nodes failure case. Since the routing protocol use directed construction and reconstruction methods, it can harmonize the energy efficiency and failure-tolerance on a desired balance. In this paper, the construction and working principles of balanced-alternative communication tree are discussed. For overcoming the frangibility of traditional spanning tree, reconstruction scenario of tree-based communication path is emphatically explored. Furthermore, the proposed protocol is evaluated from the effect of energy efficiency to the failure-tolerance. With the simulation, the performance of BAST is verified to be efficient and available.

Barrier Coverage for 3D Camera Sensor Networks

Barrier coverage, an important research area with respect to camera sensor networks, consists of a number of camera sensors to detect intruders that pass through the barrier area. Existing works on barrier coverage such as local face-view barrier coverage and full-view barrier coverage typically assume that each intruder is considered as a point. However, the crucial feature (e.g., size) of the intruder should be taken into account in the real-world applications. In this paper, we propose a realistic resolution criterion based on a three-dimensional (3D) sensing model of a camera sensor for capturing the intruder’s face. Based on the new resolution criterion, we study the barrier coverage of a feasible deployment strategy in camera sensor networks. Performance results demonstrate that our barrier coverage with more practical considerations is capable of providing a desirable surveillance level. Moreover, compared with local face-view barrier coverage and full-view barrier coverage, our barrier coverage is more reasonable and closer to reality. To the best of our knowledge, our work is the first to propose barrier coverage for 3D camera sensor networks.

CSI-based autoencoder classification for Wi-Fi indoor localization

Recently, indoor localization problem has drawn a wide range of attention. However, there are few researches that can keep balance between accuracy and expense, and few plans can achieve both device-free and accuracy. To solve this problem, the scheme of CSI-based autoencoder classification for Wi-Fi indoor localization is proposed. Only one wireless router and one computer are placed as signal emitter and receiver respectively. With so few ordinary devices, expenses have been decreased to a large extent. Device-free is achieved by performing localization based on Wi-Fi signal. Channel State Information (CSI) is measured and calculated to decrease the multipath effect, which reaches a higher accuracy. With the use of CSI, a mass of data are obtained. Machine learning including autoencoder and BP network are utilized owing to their advantage of processing mass data. In our experiment, this plan achieves 2-dim localizing with an accuracy of 50 cm.

Innovation Quality Cultivation of College Students by Scientific Contest

As a high-tech counter-contest, robot contest touches upon advanced research and technological integration of multiple professional fields and will become an important platform of the innovation quality cultivation for college students. According to real experience, the mechanism and exercise of innovation quality cultivation for college students is analyzed based on the robot contest platform. The fundamental cultivation condition that robot contest can provide for college students is explained. It is pointed out that this contest can bring rivalrous study chance and effectively compensate for the defaults of traditional education through the all-round training of students. Combined with acknowledge and experience acquired from real operation, the correct leading necessary for the innovation ability cultivation of college students is declared. At the end, the impact and effect on college student education created by robot contest is discussed.