Performance Evaluation of Message Routing Strategies in the Internet of Robotic Things Using the D/M/c/K/FCFS Queuing Network

Abstract

The Internet of Robotic Things (IoRT) has emerged as a promising computing paradigm integrating the cloud/fog/edge computing continuum in the Internet of Things (IoT) to optimize the operations of intelligent robotic agents in factories. A single robot agent at the edge of the network can comprise hundreds of sensors and actuators; thus, the tasks performed by multiple agents can be computationally expensive, which are often possible by offloading the computing tasks to the distant computing resources in the cloud or fog computing layers. In this context, it is of paramount importance to assimilate the performance impact of different system components and parameters in an IoRT infrastructure to provide IoRT system designers with tools to assess the performance of their manufacturing projects at different stages of development. Therefore, we propose in this article a performance evaluation methodology based on the D/M/c/K/FCFS queuing network pattern and present a queuing-network-based performance model for the performance assessment of compatible IoRT systems associated with the edge, fog, and cloud computing paradigms. To find the factors that expose the highest impact on the system performance in practical scenarios, a sensitivity analysis using the Design of Experiments (DoE) was performed on the proposed performance model. On the outputs obtained by the DoE, comprehensive performance analyses were conducted to assimilate the impact of different routing strategies and the variation in the capacity of the system components. The analysis results indicated that the proposed model enables the evaluation of how different configurations of the components of the IoRT architecture impact the system performance through different performance metrics of interest including the (i) mean response time, (ii) utilization of components, (iii) number of messages, and (iv) drop rate. This study can help improve the operation and management of IoRT infrastructures associated with the cloud/fog/edge computing continuum in practice.