Edge cloud-enabled data deviation recognition in the internet of things.

Internet of Things (IoT) networks inter-connect to user's devices and also virtual networks and devices deployed in different situations, for example, intelligent housing, smart healthcare, and intelligent transportation based on emerging technologies of computing, networking, and integrated in the network. The amount of IoT-devices involved increases expressively with rapid-growth. Dependent on the no. of devices and data, IoT face critical challenges. To improve the general efficiency of IoT data processing and to discover physical topology on self-organized IoT devices, this paper seeks to resolve related technical issues. The following topics will be addressed in external detection and data aggregation by designing the recursive key component research method. In the final section of this article, a novel Auto-Encoder (AE) is proposed for the NN-based outlining detection algorithm for edge devices using an AE encoder and decoder. Via wide variations in square error, which occurred via the encoder and decoder, data outliers can be accurately detected. [ABSTRACT FROM AUTHOR]