IoT requires data processing, which is provided by the cloud and fog computing. Fog computing shifts centralized data processing from the cloud data center to the edge, thereby supporting faster response due to reduced communication latencies. Its distributed architecture raises security and privacy issues; some are inherited from the cloud, IoT, and network whereas others are unique. Securing fog computing is equally important as securing cloud computing and IoT infrastructure. Security solutions used for cloud computing and IoT are similar but are not directly applicable in fog scenarios. Machine learning techniques are useful in security such as anomaly detection, intrusion detection, etc. So, to provide a systematic study, the chapter will cover fog computing architecture, parallel technologies, security requirements attacks, and security solutions with a special focus on machine learning techniques.
TopIntroduction
Cybersecurity has become the highest priority and focus for any industry, organization, or government agency that adapted automation or smart infrastructure. With increased internet connectivity (including human to human as well as machine to machine) and dependency on the machine.), the risk has been increased manifold. Internet of things(IoT)(Tan and Wang 2010)-based smart infrastructure has changed the way of living, manufacturing, or the service industry. IoT is not only about micro-controllers, sensors, actuators, and other circuitry but also related to data processing. The data center processes the sensor and device-generated data, to get insights from it so that it can be used for control or automation tasks, performance enhancement, management, and monitoring of the system.
Data transmitted to the cloud data center for processing usually consists of sensitive data that should be protected from unauthorized access. IoT as a business enabler, backbone of smart cities, industry 4.0 and smart home creates ease but in turn, exposes infrastructure (such as home, manufacturing unit, or power plant) control to the internet. Therefore, IoT security has become a very crucial component of cybersecurity, which includes the protection of data and control to unauthorized access.
The Number of IoT devices is growing exponentially which generates a large amount of data that is following the 3V’s (Volume, Velocity, and Variety) can be classified as big data (Soia, Konnikova, and Konnikov 2019). Fog computing(Cisco White Paper: Fog Computing and the Internet of Things: Extend the Cloud to Where the Things Are What You Will Learn 2015) (Vaquero and Rodero-Merino 2014) came into the picture to make more efficient IoT infrastructure by reducing the dependency on a data center for data processing. Fog Computing provides midway data processing thus thwarts unnecessary data movement to the cloud. So Fog computing provides faster response and more secure IoT infrastructure to support localized processing of data. However, the distributed architecture makes it vulnerable to new types of attacks since maintaining trust and authentication is a challenging task. So, along with providing more secure IoT infrastructure, Fog computing adds new challenges related to security which consists of a mixed bag of old and new security threats. Therefore, it is obligatory to secure the fog nodes and data processed on the fog nodes from the adversary.
The traditional methods will not be sufficient for securing the Fog computing environment due to the dynamic nature, a large number of nodes, dependencies on the 3rd party vendors, distributed control, and a huge amount of generated data. The Fog computing threats can be classified as security and privacy threats. The requirements of a secure Fog computing environment are trust, authentication, protected data storage, reliable communication, data privacy, and confidentiality. To satisfy these requirements techniques such as trust management, encryption, and authentication are required along with intrusion detection.
Machine learning-based security solutions are very efficient in the scenario where a huge amount of data is present to process as well as attack patterns may be unknown (Xin et al. 2018). Since in Fog environment, the data is large so Machine learning techniques can be used to identify the anomaly, traffic classification, or nodes authenticity. Both supervised and unsupervised techniques are capable to provide security solutions. Along with machine learning, deep learning-based solutions such as LSTM are used to create intrusion detection systems.