地理位置路由算法是指借助节点获得的地理位置信息进行无线传感网络中的路由发现与数据转发工作。本文提出一种基于Hull树的贪婪地理位置路由算法——Greedy Hull Tree Geographic Routing(GHTGR)。通过图形学中凸包的概念,在网络初始阶段分布式地在各节点上建立Hull树以探查网络局部拓扑结构;同时在数据分组的路由转发阶段,通过Hull树内的搜索,寻找下一跳转发节点,完成数据分组的转发传输。通过仿真实验表明,与现有地理位置路由算法相比,该算法能够正确地寻找数据转发路径,有效地减少网络能耗,提高网络传输性能。
节点定位是无线传感器网络的关键支撑技术之一,目前三维定位算法的研究较少。本文在针对地形建模的应用场景上,将APS多跳AOA定位算法的思想移植到三维定位中,提出了基于空间角度传递的多跳AOA三维定位算法MSAT3DAOA(Multi-Hop Three Dimentional AOA with Space-Based Angle Transmission),使得能够利用一跳通信范围外的信标节点信息进行定位,并且与Delaunay三角剖分算法结合起来,应用在部署环境的地形建模。仿真实验数据显示,随着信标节点的比例、通信半径的变化,MSAT3D AOA定位误差率在7%到27%不等,相比DV-Hop降低了将近30%,保证了59%到98%的定位覆盖率,而且在保证定位覆盖率的同时具有较好的地形建模定位精度。
A Two-hop Wireless Network (TWN) is the basic topology structure that provides network coding opportunity for improving throughput. Network coding on a homogeneous TWN, in which all the data flows have the same packet size and all the links have the same transmission rate, has been extensively investigated. In this paper, network coding on more practical heterogeneous TWNs, featured by various packet sizes and transmission rates, is studied. Based on the Markov model, the throughput of the proposed network coding scheme, together with the throughput gain, is derived, which matches the simulation results very well. Numerical analyses indicate that, encoding the packets with close size and close transmission rate and enlarging buffer size at the relay node help in improving the throughput gain.
