A QoS-aware resource management scheme over fog computing infrastructures in IoT systems

Fog computing has been introduced in recent years to extend cloud services to the network edge in order to minimize network delay and network congestion and overcome cloud computing limitations. However, several challenges are yet to be addressed in order to achieve the full benefits of the fog-IoT paradigm for low latency applications. The foremost crucial challenge is designing efficient resource management schemes capable of maximizing the throughput and minimizing the delay of IoT application. In this paper, we propose a QoS-aware greedy edge placement scheme that defines the way in which application modules are distributed across fog devices in order to minimize the end-to-end latency of real-time IoT applications. The proposed scheme is composed of two stages: a greedy-delay minimizing application module selection stage and a greedy-delay minimizing application module placement stage. The first stage aims to reduce the end-to-end latency by determining the order in which the application modules are placed across fog devices to maintain high QoS of real-time applications. The second stage uses the depth first search algorithm to select the fog node that meets application modules’ processing, storage and bandwidth requirements. To evaluate the performance of the proposed scheme intelligent surveillance through distributed camera networks application was used. The application modules are mapped onto fog nodes connected in different configurations. The experimental results demonstrate that the proposed scheme provides high QoS by reducing the end-to-end latency, the network usage and the energy consumption.