Abstract
Internet of Things (IoT) assisted robotic applications are becoming prominent in smart cities for granting autonomous services for end-users. The IoT-enabled edge-computing paradigm provides ubiquitous computing and resource access for the robots for robust services. The interactive sessions and partial computations of the robots result in unnecessary exploitation of real-time resources. This paper introduces a swarm intelligence-based modular interactive computation scheme (MICS) for IoT-assisted robots to address the incompleteness in resource utilization and reduce computational complexity. In this scheme, particle swarm optimization (PSO) is incorporated with intelligent decision-making. The role of swarm intelligence is to improve interactive computations using different agents that provide offloading support. The flexible sharing of swarm agents provides shared resource utilization and comprehensive calculations for robot-based assistance. The proposed scheme performs computing and interaction decisions based on the robot inputs and its parallel processing feature. The proposed scheme's performance is verified using the metrics interaction span, decision complexity, offloading ratio, computing time, and outages. The proposed scheme achieves a 12.74% high interaction span, 6.68% less decision complexity, 11.69% less offloading ratio, and 19.41% less computing time under different intervals. Based on the resource availability analysis, MICS achieves 62.1% compared to IoT-IF 25.6%, SF-EM 32.9, and 25.61%. Further MICS offloading ratio achieves up to 8.5% following IoT-IF 11.9%, SF-EM 10.1%, and SDF 9.1%.