基于后缀摘要的可选下一跳转发信息表聚合方法

NDN（named data networking）网络直接依据层次化内容标志进行路由, 内容条目数量激增使路由表规模呈现爆炸式增长。传统的单径路由方式对于转发信息表（forwarding information base, FIB）的聚合和压缩的作用已不大, 为此提出一种基于后缀摘要的可选下一跳转发信息表聚合方法。一方面, 将多可选下一跳的路由方式引入到转发信息表聚合过程, 使得具有共同下一跳的内容条目进一步聚合, 缩减了路由表项数量; 另一方面, 为解决前缀过度聚合带来的后缀空洞问题, 利用布鲁姆过滤器提取后缀摘要, 提高了路由的成功率。理论分析和仿真实验表明：将后缀摘要和可选下一跳FIB聚合相结合, 可使路由表项缩减到原来的20%以下, 同时随着布鲁姆过滤器哈希函数的增多, 可使内容路由的成功率接近100%。

Nexthop-selectable aggregation based on suffix digest for forwarding information base

NDN is a recently proposed general-purpose network architecture whose routing based on hierarchical content name. The explosion of the number of routing entries have driven the growth of routing table size at an alarming rate. The traditional methods of single-hop routing have little effects on FIB aggregation and reduction. This paper proposed a nexthop-selectable aggregation method based on suffix digest for FIB. On one hand, it put the nexthop-selectable method into the FIB aggregation to achieve a better aggregation with the same prefix and reduced the number of routing table size. On the other hand, it also provided the method based on suffix hole to make sure the success of routing. Theoretical analysis and results show that nexthop-selectable FIB aggregation method based on suffix digest can reduce the FIB sizes to below 20% and the success probability can be close to 100% with the growing number of hash functions of Bloom filter.