Delay-aware application offloading in fog environment using multi-class Brownian model

In the era of technology, compact and low-powered IoT devices have become an integral part of the daily routines of various sectors such as healthcare, education, industries, defense, and agriculture, etc. The IoT has increased the flow of enormous data from one network to another and extended the dependability of the public networks. In health applications, this dependency can cause excessive packets to be lost and prolonged waiting, which can cause sensitive applications to suffer. The authors suggested fog computing as a potential middle layer in the delay-aware computational system in existing works. There are many issues with the cloud-based computation model, such as location awareness, long-distance communication path, and unmanageable network bandwidth. These parameters can be solved using a fog computing model, which helps achieve ultra-low latency and minimum service delivery time. However, task offloading and scheduling are significant open research issues concerning resource management and allocation in a cloud-fog system. This paper presents a delay-aware application offloading focusing on reducing the response time and maximizing the system performance. Our proposed model developed a multiclass open queueing model that maintains the traffic on different queues, aiming to achieve maximum system performance. The simulation results demonstrate that our proposed model DAAO (Delay-aware application offloading) yields a performance improvement of 14.30% service rate and minimized failure rate by 2.0%. The simulation results supported the robustness of the proposed model in terms of minimum response time and maximum success rate. Our proposed model achieves a better tradeoff than existing work in terms of offloading and minimum failure rate.