Abstract
Industrial automation and management designed using the Internet of Things (IoT) paradigm leverage the functional and reliable operations of the production industry. Different network slices for independent and collaborative functioning connect the internal operations between different industrial IoT (IIoT) layers. This article proposes an independently-tailored network-slicing architecture for improving the mutual operations of different layers of IIoT functions. The proposed architecture helps improve the processing of scheduled jobs in different layers in an associated manner. The association process between the scheduled processes is independently analyzed for improving the swiftness in IIoT production outcomes. The scheduling and network-slicing features are recommended based on the processing and production outcomes of the analyzed industry using regression learning, which helps to assign associated processes in a queue depending on the time and production unit availability. Therefore, the architecture's network-slicing feature is modified based on the recommendations from the learning in providing seamless processing and interconnection support for IIoT production enhancements. The proposed architecture's performance is assessed using the metrics production responses, processing rate, processing time, process lag, and resource allocations. The proposed architecture achieves 10.35%, 15.32%, and 9.05% high response ratio, processing rate, and resource allocation. It reduces processing time and lag factor by 9.25% and 9.19% respectively. This observation is with respect to the processing units.