No abstract

Architectural changes are happening in the modern industries due to the adaption and the deployment of ‘Internet of Things (IoT)’ for monitoring and controlling various devices remotely from the external world. The most predominant place where the IoT technology makes the most sense is the industrial automation processes in smart industries (Industry 4.0). In this paper, a reliable ‘Next Generation Cyber Security Architecture (NCSA)’ is presented for Industrial IoT (IIoT) environment that detects and thwarts cybersecurity threats and vulnerabilities. It helps to automate the processes of exchanging real-time critical information between devices without any human intervention. It proposes an analytical framework that can be used to protect entities and network traffics involved in the IIoT wireless communication. It incorporates an automated cyber-defense authentication mechanism that detects and prevents security attacks when a network session has been established. The defense mechanism accomplishes the required level of security protection in the network by generating an identity token which is cryptographically encrypted and verified by a virtual gateway system. The proposed NCSA improves security in the IIoT environment and reduces operational management cost.

No abstract

Cloud computing (CC) is vulnerable for the attacks of current information technology, because it prolongs and uses the information technology infrastructure, traditional operating system, and applications. Several security issues in detecting irregular network performances are confronted by the CC environment. With the intention of solving these security issues, dual-channel capsule generation adversarial network optimized with red fox optimization algorithm fostered intrusion detection framework (IDS-CC-DCCGAN-RFOA) is proposed in this article, for securing the privacy attacks in CC environment. Initially, the data are taken from NSL-KDD benchmark dataset. Then the data are given to the preprocessing segment. When using the developed random forest and local least squares in preprocessing segment, the redundancy, and missing values are purged. Then the output of preprocessing is given to the feature selection stage. In feature selection, the optimal features are extracted using univariate ensemble feature selection (UEFS) technique. Based on the optimal features, the data are classified into secured data and privacy attack data based on dual-channel capsule generative adversarial network. Here, the weight of dual-channel capsule generative adversarial network is optimally tuned using the red fox optimization algorithm for generating an effective and optimal solution to detect the intruders. The proposed IDS-CC-DCCGAN-RFOA method is implemented in MATLAB. Here, the performance metrics, like accuracy, specificity, sensitivity, F-score, precision, false positive rate, AUC are analyzed.Then, the proposed IDS-CC-CCGAN-WSOA method provides 13.9367%, 11.12%, 13.268%, and 13.739% higher accuracy and 24.54%, 29.76%, 12.87%, and 34.21% higher sensitivity compared with the existing methods, like intrusion detection framework in cloud computing using deep belief network with salp swarm algorithm (IDS-CC-DBN-CSSA), intrusion detection framework in cloud computing using deep neural network with improved genetic algorithm and simulated annealing algorithm (IDS-CC-DNN-IGASAA), intrusion detection for cloud computing using multilayer perceptron neural network and artificial bee colony optimization algorithm (IDS-CC-MLPNN-ABC), and intrusion detection for cloud computing using recurrent convolutional neural network with ant lion optimization algorithm (IDS-CC-RCNN-ALO).

A data collection via secure routing in wireless sensor networks (WSNs) has given attention to one of security issues. WSNs pose unique security challenges due to their inherent limitations in communication and computing, which makes vulnerable to various attacks. Thus, how to gather data securely and efficiently based on routing protocol is an important issue of WSNs. In this paper, we propose a secure hybrid routing protocol, denoted by SHRP, which combines the geographic based scheme and hierarchical scheme. First of all, SHRP differentiates sensor nodes into two categories, nodes with GPS (NG) and nodes with antennas (NA), to put different roles. After proposing a new clustering scheme, which uses a new weight factor to select cluster head efficiently by using energy level, center weight and mobility after forming cluster, we propose routing scheme based on greedy forwarding. The packets in SHRP are protected based on symmetric and asymmetric cryptosystem, which provides confidentiality, integrity and authenticity. The performance analyses are done by using NS2 and show that SHRP could get better results of packet loss rate, delivery ratio, end to end delay and network lifetime compared to the well known previous schemes.