The growing interest in Internet of Things (IoT) applications has resulted in deployment of a large number of heterogeneous end devices and services. The increasing need for the interaction between the IoT device and cloud computing systems has resulted in formation of the edge computing paradigm, enabling the processing and execution of services at the edge of the network. In the efficient distribution and management of these services to the edge devices, orchestration plays an important role.
The critical communication systems, such as Industrial IoT applications, demand high performance coupled with ultra-reliability. The emerging 5G technology focuses on low latency and high bandwidth URLLC (Ultra-Reliable Low-Latency Communication) services. Taking the advantage of this fast and reliable communication, dynamic orchestration is possible in the cloud edge to perform critical IoT applications effortlessly.
This thesis aims to adopt a novel component to the network architecture, for reporting run-time data of the edge devices to the edge-cloud network management system. The run-time data includes real-time measurement of edge device parameters such as sensors data, workload parameters, hardware parameters, and network link parameters. This telemetry solution enables advanced orchestration and management (application roll-out deployment, relocation algorithms ,optimized load balancing, configurable cost trade-offs etc.) along with improving the fault tolerant capabilities of the network edge.