摘要
过去对CAN的漏检错帧概率的研究很有限,数据的获得主要依靠大量的仿真测试。由于要仿真的量太大,实际上仿真的仍然是极小的样本,所以得到的漏检错帧概率可信性不足。本文介绍了漏检实例的构造方法,从而进行漏检错帧概率下限的分析计算。得到的CAN协议的漏检错帧概率远大于以前的结论,因此对CAN的应用有巨大的冲击。由于已有大量应用必须加以改进,提出了改进的软件补救措施,它从根本上解决了填充规则对CAN错帧漏检率的影响。
Limited literatures were published about the research of undetected frame error rate of CAN protocol. They were based on the software fault injection. Though the simulation was exhausting, but in comparing with the possible quantity of error cases it was a very small sampling. Hence the conclusion based on that simulation is less persuasive. Present paper gives a reconstruct method of error un- detected frame. Based on this method the lower boundary of undetected error rate obtained is several orders higher than that claimed by Bosch specification 2.0. This has an impaction on the user. As the applications are so widely in use, it is an urgent need to fix this problem. Present paper provides a software patch. It radically eliminates the stuffing rule's disturbs on CRC check.
出处
《单片机与嵌入式系统应用》
2011年第9期8-11,共4页
Microcontrollers & Embedded Systems