论文部分内容阅读
网络地址翻译与协议翻译(NAT-PT),是IPv6(In-ternet P rotoco l vers ion 6)过渡协议中非常重要的一种。地址池是NAT-PT中的重要概念,其大小对于NAT-PT的性能起着关键的影响。该文的主要研究目的就是获得性能和地址数量上的平衡点。该文使用随机Petri网对NAT-PT进行形式化描述,建立了其广义随机高级Petri网模型和广义随机Petri网模型。并且通过工具包SPNP(S tochastic Petri N etPackage)计算分析得到了一定主机数量下地址池大小的近似最优化结果,同时对NAT-PT地址翻译的系统吞吐量进行了评价。当地址数量为近似最优值时,不仅可以节省地址分配,而且系统性能几乎不受影响。
Network Address Translation and Protocol Translation (NAT-PT) is a very important transitional protocol in IPv6 (In-ternet P rotoco l ion 6). Address pooling is an important concept in NAT-PT and its size has a key impact on the performance of NAT-PT. The main purpose of this paper is to find the balance of performance and address quantity. This paper formally describes NAT-PT using stochastic Petri nets, and establishes its generalized stochastic advanced Petri net model and generalized stochastic Petri net model. The approximate optimization results of the size of the address pool for a certain number of hosts are calculated and analyzed through the toolkit SPNP (S tochastic Petri Nackackage), and the system throughput of NAT-PT address translation is also evaluated. When the number of addresses is an approximate optimal value, not only the address allocation can be saved, but also the system performance is hardly affected.