The paper discusses the application of Honeypot technology with adaptive behavior for tracking and analyzing attacks on the Internet of Things networks. The analysis of existing adaptive systems is carried out and the optimal one for building a honeypot is determined. It is proposed to use the Markov decision process as a mathematical apparatus for the adaptive Honeypot system. The resulting honeypot can be used to track XMPP and SSH attacks