基于ICT分布的容迟网络接触模型

在容迟网络中,掌握节点之间的接触间隔时间(ICT)的特性,能够为网络性能分析、路由协议设计以及算法优化等研究提供理论指导和帮助,但目前的ICT模型往往缺乏普适性.通过对节点运动做出一般性假设,基于可靠性数学方法,给出了一个基于ICT分布的接触模型——IDCM.该模型证明了两个移动节点之间的ICT服从指数分布,且指数分布的参数仅与两个节点的历史接触次数和累积ICT有关.在随机方向(RD)移动模型、随机路点(RWP)移动模型、北京市出租车网络、口袋交换网4个数据集上进行了仿真验证,并与基于统计拟合参数的指数分布模型进行对比.仿真实验结果表明,IDCM能够准确反映节点对之间的接触间隔时间分布,且模型准确性优于基于统计拟合参数的指数分布模型.

ICT distribution-based contact model for delay tolerant networks

Characteristics of inter contact time (ICT) are essential for delay tolerant networks (DTNs), which are beneficial to network performance analysis, routing protocol design and algorithm optimization, but existing ICT models lack general applicability. Based on the general assumptions, a reliability mathematical model was proposed, which is called ICT distribution-based contact model (IDCM), to demonstrate that the ICT between two mobile nodes was exponentially distributed and to verify that the exponential parameter of node pairs only related to the historical contact numbers and cumulative ICTs. Based on random direction (RD) mobility model dataset, random way point (RWP) mobility model dataset, Beijing taxi network dataset, pocket switch network dataset, extensive experimental simulations were conducted and comparison to parameter-aggregated exponential model was made. The simulation results show that IDCM characterizes contacts of node pairs accurately and have better accuracy than parameter-aggregated exponential model.