A Quorum-Based Management Scheme for Ad-Hoc Networking

随着规模的增加和配置变化的速度，ad-hoc网络中的移动性管理预计将成为这类网络设计中的基本挑战之一。传统的基于HLR/VLR的移动性管理方案的基本缺点源于两个问题：作为HLR的节点可能经常无法到达（由于此类网络中的间歇性无线连接），以及移动设备与特定HLR的关联需要将该信息传播到所有网络节点，这些节点在网络中的存在可能经常变化。在提出的方案中，网络中的节点子集动态地承担“移动报告中心”（MRCs）的功能。移动设备的“位置”被报告给mrc的一组（quorum）。要定位移动设备，需要查询一个集合（quorum），可能不同于报告的集合。当任意两个群体相交时，查询操作确保生成移动设备的位置。由于法定人数可能在多个mrc中重叠，因此与有限数量的mrc失去连接可能对方案定位移动设备的能力没有影响。最后，由于quorum创建的“全网已知”方案，不需要传播移动设备与特定MRC的关联信息——任何向两个quorum报告和查询的信息都将产生所寻求的答案。在为期一年的可行性研究中，我们将研究以下主题：(i)研究基于群体的移动管理方案的复杂性和相关开销；（ii）在有限的网络负载和统一的MRCs处理负载下确保高度可靠性的群体构建策略；（iii）减少在重新配置期间出现错误操作和/或查询结果不一致的可能性；

With increasing size and the rate at which their configuration change, mobility management in ad-hoc networks is expected to be one of the basic challenges in the design of this breed of networks. The basic shortcoming of the traditional HLR/VLR based mobility management schemes stems from two problems: nodes that serve as HLRs may be frequently unreachable (due to the intermittent wireless connectivity in such networks) and the association of a mobile with a particular HLR requires dissemination of this information to all the network nodes, presence of which in the network may frequently change. In the proposed scheme, a subset of nodes in the network dynamically assumes the function of the "Mobility Reporting Centers" (MRCs). The "location" of a mobile is reported to a set (quorum) of the MRCs. To locate a mobile, a set (quorum), possibly different from the reported-to one, is queried. As any two quorums intersect, the querying operation is ensured to produce the location of the mobile. As the quorums may overlap in multiple MRCs, loss of connectivity with a limited number of MRCs may have no effect on the ability of the scheme to locate a mobile. Finally, due to the "network-wide known" scheme for quorum creation, there is no need to disseminate the information of association of a mobile with a particular MRC - any reported-to and queried-from two quorums will produce the sought answer. In the proposed one-year feasibility study, we will tackle the following topics: (i) study of the complexity and associated overhead of the quorum-based mobility management scheme' (ii) quorum construction strategies that ensure high degree of reliability with limited network load and uniform MRCs' processing load; (iii) reduction of the probability of erroneous operations and/or inconsistent querying results during the reconfiguration periods;