基于多层增量式可扩展扇形兴趣区域的大规模DVE场景对等渐进式传输机制

发布时间：2019-05-20 15:21

【摘要】：为了解决在有限的互联网带宽上高效下载海量虚拟场景的"瓶颈"问题,文中针对基于对等网络(P2P)的虚拟场景渐进式传输的增量判断和场景下载源节点最优选择这两个关键点,首先将传统的兴趣域(Area of Interests,AOI)细化为兴趣扇区域(Sector of Interest,SOI),又将其提升为多层增量式可扩展兴趣扇形区域(Multi-LayeredIncrementally Scalable SOI,MISSOI),给出了MISSOI的分划扩展算法,并基于MISSOI提出了一个高效的整数级SOI场景下载的拾取算法;然后设计了一个基于对等网的海量DVE场景渐进式下载框架,并给出了在P2P-DVE中快速搜索场景下载源节点以及选择最优场景下载源节点的高效算法.仿真实验结果表明作者提出的方法在可视场景下载域的拾取、下载源查询成功率和下载请求失败率等性能指标上明显好于目前主流海量虚拟场景P2PDVE传输机制FLoD,更是远优于传统的Client/Server虚拟场景传输模式.
[Abstract]:in order to solve the "Bottleneck" problem that the mass virtual scene is efficiently downloaded over the limited internet bandwidth, the two key points are optimally selected for the incremental judgment of the virtual scene progressive transmission based on the peer-to-peer network (P2P) and the scene download source node, Firstly, the traditional region of interest (AOI) is refined into a sector of interest (SOI), and then the region of interest (AOI) is promoted to a multi-layer incremental scalable region of interest (MISSOI), and a division extension algorithm for MISSOI is given. Based on MISSOI, a high-efficiency pick-up algorithm for integer-level SOI scene download is proposed, and a massive DVE scene progressive download framework based on peer-to-peer network is designed, and a high-efficiency algorithm for fast searching scene download source node and selecting optimal scene download source node in P2P-DVE is given. The simulation results show that the method proposed by the author is better than that of the current mainstream mass virtual scene P2PVE transmission mechanism, such as the pick-up of the visual scene download domain, the success rate of the download source query and the download request failure rate, and is far superior to the traditional Client/ Server virtual scene transmission mode.
【作者单位】：同济大学软件学院;中国科学院云计算产业技术创新与育成中心;
【基金】：国家自然科学基金面上项目(61272276) 国家“九七三”重点基础研究发展规划项目基金(2010CB328106) 国家十二五计划重大科技支撑项目(2012BAC11B00-04-03) 高等学校博士学科点专项科研基金项目(20130072110035) 吉林省重点科技攻关课题(20140204088GX) 长春高新区长白慧谷英才计划(3-2013006)资助~~
【分类号】：TP393.01

【参考文献】