With the rapid development of computer technology and related applications

more and more information systems have been put into people’s daily lives. At the same time

the popularity of IPv6 technology has also led to an explosive growth of more and more Internet of things devices. However

attacks on various information systems and Internet of things devices are also emerging in an endless stream

which has seriously threatened the safe operation of daily information systems. Therefore

security detection technology for malicious traffic plays a vital role in network security. This paper proposes a malicious traffic detection algorithm based on multi-granularity scanning and BP neural network. Through calculation and simulation of experimental data

this algorithm is used to obtain a better accuracy rate

which proves the effectiveness of this algorithm.