面向接触网状态监测的无线传感网络循环冗余校验多项式性能研究

Research on Performance of Cyclic Redundancy Check Polynomial for OCS State Monitoring in Wireless Sensor Network

为满足接触网状态在线监测中无线传感模块传输数据的校验要求,保障无线传感网络节点通信的高可靠性,研究循环冗余校验(CRC)多项式的校验性能。首先,对含x+1因式的CRC多项式进行残余误差概率计算,比较不同CRC多项式之间的残余误差概率,确定影响残余误差概率的主要因素;然后,提出1种可定量评估CRC多项式性能的简单计算方法,适用于任意数据位长校验时性能更优多项式的精确选取;最后,以无线传感模块中数据位长为256 bit的待校验数据为验证对象,针对无法通过简化计算直接比较性能优劣的CRCANSI多项式和CRC-CCITT多项式,采用近似残余误差概率与位出错漏检率精确计算的方式,实现对2种多项式性能的定量评估。结果表明:多项式残余误差概率与待校验数据位长、最小汉明距离、CRC校验码位数和字节出错概率直接相关;待校验数据位长为256 bit时,16位CRC多项式更为适用;总数据位长为272 bit(含16 bit校验位)、BSC信道下位出错概率小于5.422×10^(-3)时,CRC-CCITT多项式总体检错性能优于CRC-ANSI多项式。

In order to meet the calibration requirements of data transmitted by the wireless sensor module in on-line monitoring of OCS state and ensure the high reliability of wireless sensor network node communication,the calibration performance of cyclic redundancy check(CRC)polynomial is studied.Firstly,the residual error probability of CRC polynomials with x+1 factor is calculated,and the residual error probabilities of different CRC polynomials are compared to determine the main factors affecting the residual error probability.Then,a simple calculation method that can quantitatively evaluate the performance of CRC polynomials is proposed,which is suitable for the accurate selection of polynomials with better performance in the calibration of arbitrary data bit length.Finally,taking the 256 bit length data to be checked in the wireless sensing module as the verification object,for the CRC-ANSI polynomial and the CRC-CCITT polynomial whose performance can not be directly compared through simplified calculation,the approximate residual error probability and the miss check rate of bit error are accurately calculated to realize the quantitative evaluation of the performance of the two polynomials.The results show that the residual error probability of the polynomial is directly related to the data bit length to be checked,the minimum Hamming distance,the CRC check code number and the byte error probability.The 16-bit CRC polynomial is more suitable when the data bit length to be checked is 256 bit.When the total data bit length is 272 bit(including 16 bit check bit)and the error probability of the lower bit of the BSC channel is less than 5.422×10^(-3),the overall error detection performance of the CRC-CCITT polynomial is better than that of the CRC-ANSI polynomial.