The majority of current research on sensor network localization focuses on wireless sensor networks deployed on two-dimensional (2D) plane or in three-dimensional (3D) space, very few on the 3D terrain surface. However, many real-world applications require large-scale sensor networks deployed on the surface of a complex 3D terrain. Compared with planar and 3D network localization, terrain surface network localization generates unique and fundamental hardness. We explore 3D surface network localization with terrain models. A digital terrain model (DTM), available to the public with a variable resolution of up to one meter, is a 3D representation of a terrain's surface. It is commonly built using remote sensing technology or from land surveying and can be easily converted to a triangular mesh. Given a sensor network deployed on the surface of a 3D terrain with one-hop distance information available, we can extract a triangular mesh from the connectivity graph of the network. The constraint that the sensors must be on the known 3D terrain's surface ensures that the triangular meshes of the network and the terrain's surface overlap and approximate the same geometric shape. The basic idea of the localization algorithms is to map the two triangular meshes extracted from the connectivity graph of a sensor network and the DTM of its deployed terrain surface to the plane. The two meshes mapped to the plane can be easily aligned if the location information of anchor nodes is available. We introduce a fully distributed algorithm to construct a well-aligned mapping between the two triangular meshes in the plane based on anchor nodes information. However, accidents may happen on anchor nodes. We then introduce an anchor-free algorithm to extract feature points with geometric properties intrinsic to surface distances and independent of the embedding of the two meshes in 3D. The matched feature points induce transformations to align the two meshes in the plane. With the aligned triangular meshes of a network and its deployed terrain surface, each sensor node of the network can easily locate reference grid points from the DTM of the terrain to calculate its own geographic location. We carry out extensive simulations under various scenarios to evaluate the overall performance of the proposed algorithms with different factors such as the one-hop distance measurement error, the resolution of a DTM, and the performance of the algorithm in the situation of connectivity only.

This paper considers a fundamental problem of designing routing scheme resilient to node or link failures for wireless sensor networks deployed on a surface of a complex-connected three-dimensional (3D) setting. Instead of heuristically detouring around the failed path, we borrow homotopy, an important topological concept, to effectively create and evaluate the diversity of alternative paths. We propose a tessellation-free and GPS-free method to compute paths with different homotopy types on surface networks. A source node greedily forwards a packet to its destination based on the computed nodes’ virtual planar coordinates. When the current path fails, the source node can flexibly choose another greedy path from a different homotopy type to deliver the packet. The proposed algorithms are distributed and scalable to both the size and genus number of a surface network. We evaluate the performance of the proposed routing scheme under three different failure models. Simulation results show that our method achieves the best performance under geographically correlated failure models compared with other resilient routing schemes. We also compare our routing scheme with existing state-of-the-art ones specifically designed for surface networks when a network is failure free. Our method achieves the lowest stretch factor.

The recent advances in sensors, actuators, robots, and mobile wireless communication technologies have acceleratedinterest in autonomous networked robots (ANRs), where theindividual robots coordinate among themselves to complete atask, e.g., to explore or monitor a Field of Interest (FoI). Byteamwork, which is especially important in complex tasks, ANRsystem expresses much more capacity than traditional staticsensor networks. Existing work focuses on improving the coverageperformance of a group of ANRs within a single FoI. In thisresearch, we consider a group of ANRs that are instructed toexplore a number of FoIs. After they complete a task at currentFoI, they move to the next one, which may be far away fromcurrent one and the shape can also vary dramatically. Ourresearch focuses on how to efficiently enable such transition. TheANRs must be able to redeploy themselves to desired positionsin the new FoI based on distributed algorithms. Besides, to avoidunexpected event breaks network's integrity, the ANRs shouldpreserve their local connectivities as much as they can andorganize themselves as a whole network without any isolatednodes during the transition. Furthermore, considering energyconsumption, such relocation algorithm should work at the costof reasonable total moving distance. We study this problemand call it optimal marching of autonomous networked robots. The proposed algorithms guarantee global connectivity, andpreserve local connectivities as much as possible at negligiblecost of moving distance. Additionally, ANRs can automaticallyadjust their deployment density in the new FoI based on therequirements of various tasks or regions.