The Internet of Things is a network structure that utilizes sensors or radio frequency identification tag technology to connect things. Through the construction of the Internet of things can realize the information perception and data transmission between objects, the Internet of things technology in the smart home control wireless sensor network design and logistics supply chain construction and other application areas show high application prospects. The node positioning algorithm is studied from the aspects of node distance, sending and receiving time, obstacle blocking range, various interference factors, etc. By analyzing the distance of sensor nodes, sending and receiving time, obstacle blocking range, and various interference factors, the effective data are calculated, the invalid data are eliminated, and the node-based positioning algorithm is established to realize the secure transmission of information based on the nodes, which will have a profound impact on the user's location determination. In this regard, this paper proposes a big data algorithm based IoT security detection model to improve IoT security performance based on ABC (Artificial Bee Colony) algorithm for IoT node localization model research, and experimentally verified that in the node localization experiments, the SFLA algorithm converges around 425 iterations, and ABC algorithm starts to converge around 300 iterations. The algorithm in this paper is better than SFLA (Shuffled Frog Leaping Algorithm) algorithm.