Secure IoT Network Structure Based on Distributed Fog Computing, with SDN/Blockchain

Abstract

IoT is a new communication paradigm that gains a very high importance in the past few 18 years. This communication paradigm supports various heterogeneous applications in many fields 19 and with the dramatic increase of the number of sensor devices, it becomes a demand. Designing 20IoT networks faces many challenges that include security, massive traffic, high availability, high 21 reliability and energy constraints. Thus, new communication technologies and paradigms should 22 be deployed for IoT networks to overcome these challenges and achieve high system performance. 23Distributed computing techniques (e.g. fog and MEC), software defined networking (SDN), 24 network virtualization and blockchain are common recent paradigms that should be deployed for 25IoT networks, either combined or individually, to achieve the main requirements of the IoT 26 networks at a high system performance. Fog computing is a form of edge computing that has been 27 developed to provide the computing capabilities (e.g. storage and processing) at the edge of the 28 access network. Employing Fog computing in IoT networks, as an intermediate layer between IoT 29 devices and the remote cloud, becomes a demand to make use of the edge computing benefits. In 30 this work, we provide a framework for the IoT system structure that employs an edge computing 31 layer of Fog nodes controlled and managed by SDN network with the blockchain technology to 32 achieve a high level of security for latency sensitive IoT applications. The proposed system 33 employs SDN network with distributed controllers and distributed OpenFlow switches; these 34 switches are enabled with limited computing and processing capabilities. Furthermore, a data 35 offloading algorithm is developed to allocate different processing and computing tasks to the 36 distributed OpenFlow switches with available resources. Moreover, a traffic model is proposed to 37 model and analyze the traffic among different parts of the network. The proposed work achieves 38 various benefits to the IoT network, such as the latency reduction, security improvement and high 39 efficiency of resources utilization. The proposed algorithm is simulated and also the proposed 40 system is experimentally tested over a developed testbed to validate the proposed structure. 41Experimental results show that the proposed system achieves higher efficiency in terms of latency, 42 security and resource utilization. 43network that enables the communication and interaction between physical objects; this transforms 49 these objects from being blind to be smart [2]. Recently, IoT gains a very high significance because of 50 the great impact of all life fields [3]. IoT is expected to completely change our life by introducing 51 wide range of applications in various fields [4]. These applications include smart home, smart 52 cities, health care, smart vehicle and remote monitoring [5,6]. The IoT technology has a high market 53 impact as it comes with big market opportunities for various sectors such as hardware 54 manufactu...

Distributed Intelligence in the Internet of Things: Challenges and Opportunities

“…A hybrid distributed data flow is introduced in [72,73]. All levels of the architecture comprises of fog nodes and these fog nodes works based on their computing resources.…”

A Review on Data protection and privacy in Fog Computing Network

“…A distributed model of CCAF would be more suitable for Fog computing. In [17] authors proposed a secure framework for IOT. It uses SDN based management of Fog Nodes at edge computing layer and blockchain for key management.…”

Towards a Scalable IOTA Tangle-Based Distributed Intelligence Approach for the Internet of Things

“…More advanced approaches are proposed in [20,21,23,26] in which they rely on fog computing to enable DI, for example the work presented in [20], in which the authors applied two techniques, device-driven and human driven intelligence to reduce energy consumption and latency. The approach relies on machine learning (ML) to detect user behaviors and adaptively adjusts the sampling rate of sensors and resource schedules (timeslots in the MAC layer) between sensor nodes.…”

“…An architecture that is composed of three layers is proposed in [26]. The approach employs several technologies to achieve DI at different layers.…”