Lora-Based Data Transmission System for Smart Buildings

The necessity of ease and convenience for users, a Smart Building with complex applications needs to be operated through a centrally automated system such that the data can be transmitted over a long range. Hence, a Long-Range Technology is needed to extend the low-power communication standard of an effective range of 13–15 Km with coverage across the entire city using a single Long Range (LoRa) gateway. Presently, developing countries use cellular technologies such as 3G, LTE, and 5G, but these technologies have their limitations because of their less coverage and high power consumption. Therefore, this paper aims to compress LoRa data before transmission to the LoRa gateway using deep learning method. Also, Channel-Net is trained to represent the noise in the channel. A Deep Neural Network (DNN) architecture is required to compress and decompress the LoRa data. Moreover, in the spectrogram analysis of the chirps, the LoRa technology is used to differentiate different device signals. The results show an error in the compression-decompression Auto Encoder and the classification of devices.