Multi-RAT Network Slicing for Heterogeneous MEC-Enabled Internet of Things

Heterogeneous Internet of Things (IoT) networks operate under various radio access technologies (RATs) such as WiFi, Cellular, Zigbee, and LoRa to serve IoT applications (e.g., industrial IoT, smart city, eHealth) with different requirements. These applications rely on secure data collection from heterogeneous IoT devices using different wireless protocols, mobile edge computing (MEC) for local data processing, and data offloading and aggregation for better decision-making. Network slicing (NS) is a promising solution to serve applications with different requirements by creating logically isolated slices sharing the same physical network. However, creating multi-RAT slices is challenging due to the diverse coverage, protocols, and spectrum bands, as well as more complex network orchestration and security management. This paper pioneers multi-RAT network slicing and presents a new NS design in which service providers (SPs) operating on different wireless protocols collaborate by sharing resources and infrastructure to create multi-RAT network slices. A multi-RAT orchestrator allocates the necessary resources from each RAT to serve slices for different IoT applications to maximize the secrecy capacity and minimize cost. Multi-RAT slices are created by jointly optimizing transmission power, resource allocation, and selection of the offloading access points (APs) per RAT in the presence of eavesdroppers. We discuss centralized and distributed implementations of our scheme, collaboration mechanisms, and pricing. Then, we assess the performance of our scheme for common IoT use cases and show that our approach achieves up to seven times higher utility and offloads three times more secure data than existing schemes. Lastly, we discuss ongoing challenges and promising future research directions.