高亏格三维曲面传感器网络中基于负载均衡的弹性几何路由协议研究

The research proposes a novel resilient and load-balanced routing protocol for high genus 3D surface sensor networks. A high genus 3D sensor network is first triangulated so that it is equivalent to the surface of a 3D high genus manifold, and the genus of the triangular mesh is then effectively identified as well as cut open. Topological surfaces with different number of genuses are mapped to different geometry domains, and then are embedded in the same geometric space by the topological mirrors’ transformation and alignment. As there exists a corresponding mirror virtual image in the topological mirror for every destination node, multiple paths that are homotopic to the primary path can be obtained by connecting the source node and the mirror virtual images of the destination node in the way of greedy routing. As such, resilient routing with multi-path access can be achieved. Meanwhile, in the case of network power imbalance, a route that can balance the network load can be chosen by transforming existing mapping with the parameter of nodes’ existing power.

本项目针对高亏格三维曲面传感器网络，提出一种全新的基于负载均衡的弹性几何路由协议，其研究思路是：首先通过三角化剖分将高亏格三维曲面传感器网络等效于高亏格三维流形的表面，并对该三角网格中亏格进行有效地识别和切割。对于不同亏格数的拓扑曲面分别映射到不同的几何空间，并对其拓扑镜像进行变换和对齐，使它们嵌入同一几何空间内。目的节点在不同的拓扑镜像中均存在对应的虚像，通过贪婪路由的方式连接源节点与目的节点的任一虚像，可获得多条与主路径非同伦的备选路径，从而达到弹性路由多路径访问的目的。同时在网络电量不均衡的情况下，以节点现有电量为参数对现有映射方式进行变换，以重新选择路由进一步平衡网络负载。

本项目针对高亏格三维曲面传感器网络，提出一种全新的基于负载均衡的弹性几何路由协议，首先通过三角化剖分将高亏格三维曲面传感器网络等效于高亏格三维流形的表面，并对该三角网格中亏格进行有效地识别和切割。对于不同亏格数的拓扑曲面分别映射到不同的几何空间，并对其拓扑镜像进行变换和对齐，使它们嵌入同一几何空间内。通过贪婪路由的方式连接源节点与目的节点的任一虚像，可获得多条与主路径非同伦的备选路径，从而达到弹性路由多路径访问的目的。发表高水平学术论文21篇，期刊论文19篇，会议论文2篇；培养博士5人，硕士3人。相关成果在领域内产生较大学术影响，项目负责人因而受邀担任国际顶级期刊IEEE/ACM Trans on Networking、IEEE Transactions on Mobile Computing的编委，国际知名会议IFIP NETWORKING 2017的程序委员会共同主席，入选英国工程技术学会会士和英国计算机学会会士。