With the rapid development of Internet of Things (IoT) technology, researches in IoT and Wireless Sensor Networks (WSNs) has led to a very heavy reliance on Software Defined Networking (SDN). WSNs have some resource scarcities that affect their performance, such as power supply, processing units, memory, and communication capacity. This paper suggests an optimized routing protocol, named Future Search Algorithm-Temperature Routing protocol (FSA-TR) that is based on the temperature and recent technologies to find the optimal set of clusters. The SDN and Cloud technologies have been used in the architecture of the proposed protocol. A Multi-objective FSA that considers resource restrictions of the sensors, such as energy and communication capability, has been presented to identify the best solution. The temperature of the IoT sensors according to their distribution in the geographical has been considered. Accordingly, a Cluster Head (CH) is selected depending on the remaining energy, the distance between nodes, and environmental temperature. Finally, the experimental results showed an enhanced system performance of many features, such as energy dissipation, network lifetime and the number of packets sent to the sink. For instance, the network lifetime of the proposed protocol has been increased by approximately 36%, 27% and 20% compared to LEACH, SEP and optimized protocol PSO-C, consecutively.

Machine learning is now being used for applications ranging from healthcare to network security. However, machine learning models can be easily fooled into making mistakes using adversarial machine learning attacks. In this article, we focus on the evasion attacks against Network Intrusion Detection System (NIDS) and specifically on designing novel adversarial attacks and defenses using adversarial training. We propose white box attacks against intrusion detection systems. Under these attacks, the detection accuracy of model suffered significantly. Also, we propose a defense mechanism against adversarial attacks using adversarial sample augmented training. The biggest advantage of proposed defense is that it doesn’t require any modification to deep neural network architecture or any additional hyperparameter tuning. The gain in accuracy using very small adversarial samples for training deep neural network was however found to be significant.

The huge increase amount of Cyber Attacks in computer networks emerge essential requirements of intrusion detection system, IDS to monitors the cybercriminals. The inefficient or unreliable IDS can decrease the performance of security services and today world applications and make the ongoing challenges on the Cyber Security and Data mining fields. This paper proposed a new detection system for the cyber-attacks with the ensemble classification of efficient cost sensitive decision trees, CSForest classifier and the least numbers of most relevant features are selected as the additional mechanism to reduce the cost. The standard dataset, NSL-KDD, IDS is used to appraise the results and compare the previous existing systems and state-of-the-art methods. The proposed system outperforms the other existing systems and can be public a new benchmark record for the NSL-KDD datasets of intrusion detection system. The proposed combination of choosing the appropriate classifier and the selection of perfect features mechanism can produce the cost-efficient IDS system for the security world.

Wireless Sensor Networks (WSNs) have included one of the major challenges as energy efficiency. The optimal routing is the better solution for tackling the energy-efficiency problem as the energy is consumed with massive amount by the communication of a network. The clustering technique is the reliable data gathering algorithm for achieving energy-efficiency. The data transmit to the cluster head (CH) by each node that belonging to the cluster for clustered networks. The data transmission towards the BS or SINK is occurred once all data is received by CH from all member nodes. In multi-hop environments, the data transmission happens via other cluster-heads. It leads to the earlier death of CHs that are nearer to the SINK because of the heavy inter-cluster relay, i.e. called as hot-spot problem. The existing methods of unequal clustering approaches have been used to resolve this issue. The algorithms generate the clusters with smaller sizes while approaching towards the sink for reducing the intra-cluster relay. These algorithms didn’t consider the hotspot problem very effectively. Therefore, in this work, we introduce a multi-objective fuzzy logic based multi-path routing solution (MOFL-MPR) to address the above said problem. The popular clustering algorithms have been evaluated the performance results and the MOFL-MPR is outperformed the existing algorithms based on the obtained experimental results in terms of stable CH selection, energy efficiency and better data delivery.

Single Sign-On (SSO) allows the client to access multiple partner e-services through a single login session. SSO is convenient for the users as the user neither needs to set multiple login credentials nor login separately for individual services every time. SSO (single sign-on) authentication is a password-authentication approach that permits end users to login into multiple systems and websites with a single set of login credentials. SSO authentication is mainly useful for IT organizations that consist of many different commercial applications. The outstanding feature of SSO is that it gives organizations centralized control of their systems by giving different levels of access to each individual. It reduces password fatigue and increases security because users only need to remember a single username/password that grants them access to multiple systems. However, the Single Sign-on poses risks related to a single point of attack which may lead to a path for cybercrimes. This paper proposes a trust model to increase the security of Single Sign-on systems against the vulnerabilities discussed in the subsequent sections. The proposed Trust model is named as DANE-based Trust Plugin (DTP) which acts as an added security layer over DNS Based Authentication of Named entities(DANE). The DTP proposes the modified SAML XML schema which enables the DTP to counter the attacks.