千兆以太网中CRC-32的并行实现  

Parallel implementation of CRC-32 in Gigabit Ethernet

作　　者：田野[1] 佟皓萌 

机构地区：[1]天津理工大学中环信息学院,天津300380 [2]浙江工业大学,浙江杭州310014

出　　处：《电子设计工程》2016年第15期112-114,118,共4页Electronic Design Engineering

摘　　要：为了保证数据通信的可靠性,要使用一定的检错和纠错方式。循环冗余校验码(CRC)作为一种分组码,具有一定的检错功能。以太网传输帧中使用的是CRC-32校验码作为以太帧的最后4个字节,同以太帧一起传输。CRC的实现方式分为串行方式和并行方式,由于并行方式一个时钟周期内可以处理8个bit,与千兆以太网的GMII接口协议相符合,故千兆以太网的CRC校验码的生成和校验中常使用并行算法实现。本文研究了CRC校验码的串行实现算法和并行实现算法,并且用modelsim进行了二者的仿真,比较二者结果的一致性和实现效率,说明了CRC-32校验码的并行实现算法更适合使用于千兆以太网中。In order to ensure the reliability of data communication, the certain error detection and error correction mode must be used. Cyclic redundancy check code（CRC） as a kind of block code, has a certain function of error detection. The CRC-32 check code is used in the last 4 bytes of frame along with the Ethernet frame transmission. CRC realization methods are classified into serial and parallel, due to the parallel mode can handle 8 bits in a clock cycle, and it is in line with Gigabit Ethernet GMII interface protocol, so the generation and validation of Gigabit Ethernet CRC check code often use parallel algorithm. This paper studied CRC serial algorithm and parallel algorithm, and conducted simulation of the two using Modelsim. The consistency and efficiency of the two are compared, and the parallel implementation of the CRC-32 check code is more suitable for Gigabit Ethernet.

关 键 词：循环冗余校验(CRC) 千兆以太网 并行实现 CRC-32 

分 类 号：TN911[电子电信—通信与信息系统]