在移动IP网络中,层次移动IP协议可降低移动IP位置管理的信令消耗。该文研究支持寻呼的HMIPv6,提出了时延限制下的多步寻呼策略以及基于遗传算法的多步寻呼方案——根据移动节点的位置概率分布,采用遗传算法对寻呼区域内的所有子网进行分组,每步寻呼一个组,使平均寻呼信令开销达到或接近最优。此外,对HMIPv6、单步寻呼的HMIPv6以及基于遗传算法的多步HMIPv6寻呼方案的信令开销进行了比较分析,得出寻呼节省信令开销的必要条件。最后,给出一组数值结果以说明所提出的寻呼策略的有效性。
Hierarchical Mobile IP (HMIP) is able to reduce signaling cost for location management in mobile IP (MIP) networks. This paper focuses on paging-supported HMIPv6. Sequential paging schemes under limited paging delay are studied. In addition,a genetic algorithm (GA) based sequential paging scheme (GABSP) is proposed,in which all the subnets in a paging area are partitioned into groups by a GA making use of location probabilities of MNs, and then, each group is paged in one paging step. The GABSP can achieve the optimal or near-optimal signaling cost. Moreover, the signaling costs of HMIPv6, the HMIPv6 with single-step paging, and the GABSP are compared, and besides, the prerequisite conditions for single or sequential paging to reduce signaling cost are given. Further, numeric results show that the GABSP occupies the lowest signaling cost.