Method for realizing network address conversion

Abstract

The invention relates to a method for realizing network address translation. This method is to reasonably divide the various tasks required for network address translation, and complete them respectively by CPRC and SDP in the C5 network processor, that is, to determine the address information that needs to be converted in the message and the calculation needs in the CPRC module. The difference value of the converted address information, and look up the table to determine the converted address information. In the SDP module, the calculation of the checksum value of the converted address information of the message, the conversion of the address information of the message, and the report The text is forwarded. The invention can improve the efficiency of network address translation, and provides a basis for wire-speed forwarding of messages based on network address translation in a network communication system.

Description

technical field [0001] The invention relates to the technical field of network communication, in particular to a method for realizing network address translation. Background technique [0002] NAT (Network Address Translation) is divided into two types: one is pure IP (Internet Protocol) address translation, private network IP addresses and public network IP addresses have a one-to-one correspondence, and the main purpose is to isolate private network and external public network users . The other is NAPT (Network Address Port Transformation) technology based on the binding of IP addresses and Layer 4 protocol port numbers, which is mainly used to solve the problem of insufficient IP addresses. NAPT can be used to realize many-to-one connections from private networks to public networks. Address translation. [0003] The existing NAT application is mainly the application of NAPT technology. NAPT is proposed to solve the problem of insufficient public network IP addresses. By...

AI Technical Summary

Problems solved by technology

[0004] At present, there are mainly two ways to realize NAPT: one is a pure software algorithm implemented on the CPU of the network device, which is very slow, and when the number of entries in the NAT (Network Address Translation) table is large, the software will The table lookup performance will be greatly reduced, resulting in a sharp drop in packet forwarding performance; the other is a software algorithm based on the network processor, which uses the powerful computing power and table lookup ability of the network processor to implement NAT with software. , the network processor used is the IXP1200 network processor. IXP1200 is the first generation of network processors, so it has some inherent defects that lead to low efficiency when processing NAPT. The defects of IXP1200 include: On the one hand, the network processor There is no hardware table lookup unit inside, and all table lookup operations are completed by software, which is also powerless for NAPT transformations that require a large number of table lookup operations; on the other hand, there is no dedicated hardware for byte stream processing inside the network processor, so for the needs The operation efficiency of a large number of byte conversion processing is low; and the network processor internally shares the bus, so the data access capability is limited. Therefore, the NAPT based on IXP1200 also has low performance and cannot meet the performance requirements of the growing network communication system

[0005] To sum up, the two network address translation methods described in the prior art are due to the defects of the technology itself, so that the forwarding performance is generally not high in the application, and the forwarding performance decreases when the number of users increases

Images