An Effective Approach on Clustering and Data Aggregation for the application of Internet of Things in Long Range Networks

Abstract

IoT technologies let non-computer devices act smartly and collaborate. They present many challenges and necessitate specialised standards and communication protocols. This includes Zigbee, a low-power, short-range, energy-efficient network, and LoRa, a long-range network with comparable energy efficiency. ZigBee and LoRa parameters are thoroughly analysed. The thesis suggests clustering star-based LoRa networks to conserve energy. The thesis has three goals. Analyze IoT communication parameters and factors to meet application QoS requirements. Real-time data transmission requires more complex, scalable IoT architecture. Conventional IoT architectures are unsuitable for real-time applications that require timely data transmission to avoid casualties. Appropriate and cost-effective technologies are needed as IoT applications grow. IoT applications' two most significant aspects are network efficiency and energy conservation. Since many parameters affect communications, Canonical Correlation Analysis (CCA) has been adopted to identify primary and secondary significant factors in the LoRa network. Bayesian Belief Nets (BBNs) are applied to determine the cause of low network efficiency and predict values of independent variables for the suitable scenario. This motivates us to introduce the multi-hop clustering concept in LoRa technology to reduce the overall energy consumption in LoRa networks. One of the big problems is handling the deluge of data generated by IoT devices. Hence, a proposal to develop algorithms for pre-collection data packet aggregation in LoRa technology. We propose a data packet aggregation scheme for LoRa networks and compare it with existing equivalent algorithms. We also present a scalable energy-conserving architecture for Wireless Body Area Network (WBAN) based on the MANET-LoRa framework by applying a data aggregation scheme.