Anchor Density Minimization for Localization in Wireless Sensor Network (WSN)

In wireless sensor networks (WSN) high-accuracy localization is crucial for both of WNS management and many other numerous location-based applications. Only a subset of nodes in a WSN is deployed as anchor nodes with their locations a priori known to localize unknown sensor nodes. The accuracy of the estimated position depends on the number of anchor nodes. Obviously, increasing the number or ratio of anchors will undoubtedly increase the localization accuracy. However, it severely constrains the flexibility of WSN deployment while impacting costs and energy. This paper aims to drastically reduce anchor number or ratio of anchor in WSN deployment and ensures a good trade-off for localization accuracy. Hence, this work presents an approach to decrease the number of anchor nodes without compromising localization accuracy. Assuming a random string WSN topology, the results in terms of anchor rates and localization accuracy are presented and show significant reduction in anchor deployment rates from 32% to 2%.

Localization Algorithm for 3D Sensor Networks: A Recursive Data Fusion Approach

Location-based applications for security and assisted living, such as human location tracking, pet tracking and others, have increased considerably in the last few years, enabled by the fast growth of sensor networks. Sensor location information is essential for several network protocols and applications such as routing and energy harvesting, among others. Therefore, there is a need for developing new alternative localization algorithms suitable for rough, changing environments. In this paper, we formulate the Recursive Localization (RL) algorithm, based on the recursive coordinate data fusion using at least three anchor nodes (ANs), combined with a multiplane location estimation, suitable for 3D ad hoc environments. The novelty of the proposed algorithm is the recursive fusion technique to obtain a reliable location estimation of a node by combining noisy information from several nodes. The feasibility of the RL algorithm under several network environments was examined through analytic formulation and simulation processes. The proposed algorithm improved the location accuracy for all the scenarios analyzed. Comparing with other 3D range-based positioning algorithms, we observe that the proposed RL algorithm presents several advantages, such as a smaller number of required ANs and a better position accuracy for the worst cases analyzed. On the other hand, compared to other 3D range-free positioning algorithms, we can see an improvement by around 15.6% in terms of positioning accuracy.

Wireless Sensor Network Target Localization Algorithm Based on Two- and Three-Dimensional Delaunay Partitions

The target localization algorithm is critical in the field of wireless sensor networks (WSNs) and is widely used in many applications. In the conventional localization method, the location distribution of the anchor nodes is fixed and cannot be adjusted dynamically according to the deployment environment. The resulting localization accuracy is not high, and the localization algorithm is not applicable to three-dimensional (3D) conditions. Therefore, a Delaunay-triangulation-based WSN localization method, which can be adapted to two-dimensional (2D) and 3D conditions, was proposed. Based on the location of the target node, we searched for the triangle or tetrahedron surrounding the target node and designed the localization algorithm in stages to accurately calculate the coordinate value of the target. The relationship between the number of target nodes and the number of generated graphs was analysed through numerous experiments, and the proposed 2D localization algorithm was verified by extending it the 3D coordinate system. Experimental results revealed that the proposed algorithm can effectively improve the flexibility of the anchor node layout and target localization accuracy.

Research on Wireless Sensor Network Positioning Based on Genetic Algorithm

For smart city wireless sensing network construction needs, a network positioning algorithm based on genetic algorithm is proposed. The genetic algorithm uses a real number encoding, and the positioning model is constructed by analyzing the communication constraint between unknown nodes and a small amount of anchor nodes and constructs the positioning model, and the model is solved. The results show that when the ranging error is 50%, the positioning error is only increased by approximately 15% compared to the nonranging error. In a more harsh environment, if the ranging error is equal to the node wireless range, the ranging error is 100%, and the positioning error and the positioning ratio are not significantly changed. The scheme obtained by this algorithm can be well approarded with an ideal limit. In the case where the sensor node is given, the algorithm can obtain the maximum coverage.

Detection of Outdated Sensors in Wireless Network via a New Protocol

A novel method is proposed using the nonlinear mapping with kernel functions to correctly locate the outdated sensors in a wireless sensor network (WSN). Such detection system used Cornell regression and solved via the vector support regression (VSR) plus multi-dimensional backup vector regression (MBVSR). The developed method was simplistic and effective without the need of any additional hardware for any measurement. It required only the vicinity and information of location from the anchor nodes to detect the outdated sensors. It was achieved in three stages including the measurements, kernel regression, and stepping stage. First step measured the proximity information from a given grid. The relationships between the proximity and geographic distance among the sensors’ nodes were generated in the kernel regression stage. For the stepping phase, every sensor node found its location in the distributed way via the kernel regression. Simulation results showed the robustness and high efficiency of the proposed scheme.

Traffic Information Collection Using Wireless Sensor Network Positioning Technology

The wireless sensor network integrates sensor technology, microelectromechanical technology, distributed information processing technology, and wireless communication technology. In order to solve this problem, this paper designs and proposes an anchor node self-location algorithm. Aiming at the positioning accuracy of wireless sensor network nodes, this paper proposes an improved algorithm for sensor network node positioning that uses error correction methods to reduce accumulated distance errors and positioning errors. In this paper, the designed routing algorithm is simulated and implemented, and the performance of the routing algorithm is evaluated based on different network topologies. From the analysis results, compared with the existing typical routing algorithms, the routing algorithms designed in this paper can effectively reduce the energy consumption of the network and prolong the lifetime of the network. The core of the algorithm is to integrate the known and available information of the system to locate unknown anchor nodes. This greatly reduces the number of anchor nodes whose initial position information is required by the system, and under the condition of less impact on the positioning accuracy of the system, the cost of the system is reduced and the scope of application of the system is improved. This paper has deeply studied the positioning and tracking problems in wireless sensor networks, including node positioning, biochemical gas source positioning, and target tracking, and designed and developed a platform for positioning and tracking application research to lay the foundation for further application research. In the study of the above problems, new methods of positioning and tracking with theoretical and practical value are proposed for different practical application scenarios, and the performance is verified and evaluated through computer simulation.

Graph Deformer Network

Convolution learning on graphs draws increasing attention recently due to its potential applications to a large amount of irregular data. Most graph convolution methods leverage the plain summation/average aggregation to avoid the discrepancy of responses from isomorphic graphs. However, such an extreme collapsing way would result in a structural loss and signal entanglement of nodes, which further cause the degradation of the learning ability. In this paper, we propose a simple yet effective Graph Deformer Network (GDN) to fulfill anisotropic convolution filtering on graphs, analogous to the standard convolution operation on images. Local neighborhood subgraphs (acting like receptive fields) with different structures are deformed into a unified virtual space, coordinated by several anchor nodes. In the deformation process, we transfer components of nodes therein into affinitive anchors by learning their correlations, and build a multi-granularity feature space calibrated with anchors. Anisotropic convolutional kernels can be further performed over the anchor-coordinated space to well encode local variations of receptive fields. By parameterizing anchors and stacking coarsening layers, we build a graph deformer network in an end-to-end fashion. Theoretical analysis indicates its connection to previous work and shows the promising property of graph isomorphism testing. Extensive experiments on widely-used datasets validate the effectiveness of GDN in graph and node classifications.

Graph Universal Adversarial Attacks: A Few Bad Actors Ruin Graph Learning Models

Deep neural networks, while generalize well, are known to be sensitive to small adversarial perturbations. This phenomenon poses severe security threat and calls for in-depth investigation of the robustness of deep learning models. With the emergence of neural networks for graph structured data, similar investigations are urged to understand their robustness. It has been found that adversarially perturbing the graph structure and/or node features may result in a significant degradation of the model performance. In this work, we show from a different angle that such fragility similarly occurs if the graph contains a few bad-actor nodes, which compromise a trained graph neural network through flipping the connections to any targeted victim. Worse, the bad actors found for one graph model severely compromise other models as well. We call the bad actors ``anchor nodes'' and propose an algorithm, named GUA, to identify them. Thorough empirical investigations suggest an interesting finding that the anchor nodes often belong to the same class; and they also corroborate the intuitive trade-off between the number of anchor nodes and the attack success rate. For the dataset Cora which contains 2708 nodes, as few as six anchor nodes will result in an attack success rate higher than 80% for GCN and other three models.