In the NEtwork MObility （NEMO） environment, mobile networks can form a nested structure. In nested mobile networks that use the NEMO Basic Support （NBS） protocol, pinball routing problems occur because packets are...In the NEtwork MObility （NEMO） environment, mobile networks can form a nested structure. In nested mobile networks that use the NEMO Basic Support （NBS） protocol, pinball routing problems occur because packets are routed to all the home agents of the mobile routers using nested tunneling. In addition, the nodes in the same mobile networks can communicate with each other regardless of Internet connectivity. However, the nodes in some mobile networks that are based on NBS cannot communicate when the network is disconnected from the Internet. In this paper, we propose a route optimization scheme to solve these problems. We introduce a new IPv6 routing header named ＂destination-information header＂ （DH）, which uses DH instead of routing header type 2 to optimize the route in the nested mobile network. The proposed scheme shows at least 30% better performance than ROTIO and similar performance improvement as DBU in inter-route optimization. With respect to intra-route optimization, the proposed scheme always uses the optimal routing path. In addition, the handover mechanism in ROAD＋ outperforms existing schemes and is less sensitive to network size than other existing schemes.展开更多
In this paper, we consider the regularized learning schemes based on l1-regularizer and pinball loss in a data dependent hypothesis space. The target is the error analysis for the quantile regression learning. There i...In this paper, we consider the regularized learning schemes based on l1-regularizer and pinball loss in a data dependent hypothesis space. The target is the error analysis for the quantile regression learning. There is no regularized condition with the kernel function, excepting continuity and boundness. The graph-based semi-supervised algorithm leads to an extra error term called manifold error. Part of new error bounds and convergence rates are exactly derived with the techniques consisting of l1-empirical covering number and boundness decomposition.展开更多
基金supported by MKE,Korea,under ITRC NIPA-2009-(C1090-0902-0046)by MEST,Korea under WCU Program supervised by the KOSEF(No.R31-2008-000-10062-0).
文摘In the NEtwork MObility （NEMO） environment, mobile networks can form a nested structure. In nested mobile networks that use the NEMO Basic Support （NBS） protocol, pinball routing problems occur because packets are routed to all the home agents of the mobile routers using nested tunneling. In addition, the nodes in the same mobile networks can communicate with each other regardless of Internet connectivity. However, the nodes in some mobile networks that are based on NBS cannot communicate when the network is disconnected from the Internet. In this paper, we propose a route optimization scheme to solve these problems. We introduce a new IPv6 routing header named ＂destination-information header＂ （DH）, which uses DH instead of routing header type 2 to optimize the route in the nested mobile network. The proposed scheme shows at least 30% better performance than ROTIO and similar performance improvement as DBU in inter-route optimization. With respect to intra-route optimization, the proposed scheme always uses the optimal routing path. In addition, the handover mechanism in ROAD＋ outperforms existing schemes and is less sensitive to network size than other existing schemes.
文摘In this paper, we consider the regularized learning schemes based on l1-regularizer and pinball loss in a data dependent hypothesis space. The target is the error analysis for the quantile regression learning. There is no regularized condition with the kernel function, excepting continuity and boundness. The graph-based semi-supervised algorithm leads to an extra error term called manifold error. Part of new error bounds and convergence rates are exactly derived with the techniques consisting of l1-empirical covering number and boundness decomposition.
文摘现有的面向大规模数据分类的支持向量机(support vector machine,SVM)对噪声样本敏感,针对这一问题,通过定义软性核凸包和引入pinball损失函数,提出了一种新的软性核凸包支持向量机(soft kernel convex hull support vector machine for large scale noisy datasets,SCH-SVM).SCH-SVM首先定义了软性核凸包的概念,然后选择出能代表样本在核空间几何轮廓的软性核凸包向量,再将其对应的原始空间样本作为训练样本并基于pinball损失函数来寻找两类软性核凸包之间的最大分位数距离.相关理论和实验结果亦证明了所提分类器在训练时间,抗噪能力和支持向量数上的有效性.
基金Supported by by the Special Fund of Basic Scientific Research of Central Colleges(CZQ13015)the Teaching Research Fund of South-Central University for Nationalities(JYX13023)
