A qos aware framework for resource and service provisioning in fog computing

Abstract

The world of internet has created an era where any device or thing has newlinethe ability to interconnect with each other. This has introduced a new paradigm newlineof working called the Internet of Things which has started to generate huge newlinevolumes of data. Gathering intelligence from streaming data is a challenge newlinebut when acquired can create wonders in innovation useful for humanity. An newlineinfrastructure to store, analyze and predict the data is of utmost importance. newlineThe infrastructure development is yet another area of research which has newlineevolved from centralized servers to distributed cloud architectures and still newlineunder research. The limitations of the cloud platform to handle the streaming newlinedata and provide a near real-time decision are an area with broad scope of newlineresearch.The shortcomings of connectivity due to the remote location of the newlinecloud from the edge devices induce latency and performance issues in the newlineapplication. Further sending a large stream of data to a cloud also increases the newlinebandwidth utility. In case of applications of healthcare, sending sensitive data newlineto a third party server also leads to security issues. Thus, the use of a traditional newlinecloud infrastructure may not be suitable for all applications and a need for a newlinemore secure, low latent, low bandwidth infrastructure which was under research newlineled to Fog Computing. As of any infrastructure, creating an infrastructure is newlinea challenge and has to be highly robust to handle the data. The emergence of newlineSoftware Defined Network (SDN) has paved way for a stable network creation newlineto support IoT applications.In case of healthcare applications, immediate response to the data and accurate decision making is highly recommended. Gathering of data from the edge devices, analyzing them for actuation has to be done meticulously with least latency and optimal bandwidth utility. Use of Deep learning algorithms should not exploit the available resources which may lead to improper resource utilization in a large network. Proper design of a fog network to handle real-time health care a