Review of artificial intelligence techniques used in IoT networks

Artificial intelligence (AI) is an effective and efficient solution to manage and analyse data flow in any network. Internet of things (IoT) has quickly attracted significant global attention as an innovative, progressively growing technology. It has shown a rapid and successful involvement in many fields. Thus, IoT applications evolve exorbitantly and produce vast amounts of data required for intelligent data processing. It is approximately calculated that by 2025, IoT could make significant traffic of 79 zettabytes, and by 2030 around 25 billion active smart gadgets would be linked and woven through a single massive information network. It creates hurdles for the end-user to effectively evaluate and analyse the collected information. Therefore, IoT networks utilise robust and effective AI techniques such as machine learning (ML) and data analytics (DA), which examine large amounts of data and generate meaningful information promptly. ML is a self-learning process, and DA is another effective method for predicting the future behaviour of object or activities, using past data to improve productivity in different industries such as agriculture, transportation, online gaming, eHealth, etc. This paper discussed AI techniques such as ML and DA used in IoT networks and their impacts on productivity. Furthermore, we have discussed the future trends and challenges of IoT networks.