Читать книгу Machine Learning Techniques and Analytics for Cloud Security - Группа авторов - Страница 71

The rapid evolution of smart industry [1] in the first part of 21st century is in synchronous with growth of semiconductor industry, and as per the latest available data, it is close to a 50 billion industry. This exponential advancement becomes realizable due to the growth of information and communication technology which ensures the transform of computer-aided industry to embedded system-based industry [2, 3], where decision-making becomes totally data-driven, augmented with emotional intelligence, and sentiment analysis. The role of Internet of Things (IoT) here is to make sustainable connection between corporeal industrial environments [4, 5] with cyberspace of calculating systems. The proper augmentation generates a new system, termed as Cyber-Physical System (CPS), which makes a large paradigm shift in the industrial progress. The rapid multidisciplinary industrial wings as obtained after inclusion of IoT, in terms of living, manufacturing, food industry, travel, and utilities, improve the efficacy of production by enhancing work flow [6, 7], whereas machine effort reduces the physical manpower requirement, and therefore, quality of output is enhanced. The major challenges that the emerging industry face can effectively subdued through [i] proper decentralization of systems, [ii] assortment of systems and associated devices, [iii] data complexity and heterogeneity, and [iv] complexity of existing network. Major security issues are nullified through incorporation of these features [8, 9]. In particular, IoT-enabled system contains the following assets:

 Across various IoT-based devices, the ability to exchange information between the systems/devices applied in households/industrial sectors plays the key role in making it suitable for all classes of candidates. This unique property can be achieved through network-composite layer which enables to provide uniform access to all systems. This is made on the top of an overlay P2P (peer-to-peer) network.

 IoT data can be traceable, i.e., its temporal and spatial information can be detected and verified/authenticated/validated when saved inside the memory device. This memory device is always associated to the timestamp which efficiently makes the data tracing at real-time. Also historical information can easily be retrieved.

 Another combined features associated closely with IoT data are reliability and flexibility. It is the data which is considered as asset. Novel cryptographic mechanisms are incorporated in order to ensure the data integrity. This can be achieved through inclusion of hash functions, asymmetric encryption algorithms, and digital signature.

 Interactions in automatic mode are another important property which ensures the capability of the system under consideration for mutual interaction with similar kind of systems, and the major advantage is that any trusted third party intervention is not at all required.

In today’s developing world, living in a house has become an important part of our day-to-day life. A substantial amount of resources as well as expenditure that we would carry might not always be pleasurable as there are various problems such as accidents caused in house with the lonely child or old-aged person. These accidents happened due to various human errors. Also, the problem that haunts all the people is the possibility of theft of the household appliances by unauthorized people. To overcome these challenges, there is a pressing need to develop automated smart home system [10] or smart homes. This can be reached using the IoT-based platforms and services.

Many applications have been developed using IoT methodologies. The platform for developing such technologies is Raspberry Pi, Arduino, or other microcontrollers [11, 12] which are responsible for controlling various operations. These systems can be controlled and monitored from a registered device which is present with the user and install with the household devices.