Smart home IoT technologies have provided a new level of overall degrees of control freedom over modern homes. The core of such a system is an edge device. In this paper, a CNN-based LSTM-Autoencoder method is presented to detect anomaly points in five critical operating parameters of an edge device while managing its perpetual operation. This proposed method is based on a hybrid model using 1D-CNN layers in the encoder layer and LSTM layers in the decoder layer. Experiments were conducted using real data from Raspberry Pi devices. Compared to other state-of-the-art methods, the proposed approach had a remarkable accuracy close to 0.996 and an execution time of 312 msec. |
*** Title, author list and abstract as seen in the Camera-Ready version of the paper that was provided to Conference Committee. Small changes that may have occurred during processing by Springer may not appear in this window.