Internet of Things (IoT) represent a network of resource-constrained sensor devices connected through the open Internet, susceptible to misuse by intruders. Traditional standalone intrusion detection systems (IDS) are tasked with monitoring device behaviours to identify malicious activities. These systems not only require extensive network and system resources but also cause delays in detecting a malicious actor due to unavailability of a comprehensive view of the intruder's activities. Collaboration among IoT devices enables considering knowledge from a collection of host and network devices to achieve improved detection accuracy in a timely manner. However, collaboration introduces the challenge of energy efficiency and event processing which is particularly significant for resource-constrained devices. In this paper, we present a collaborative intrusion detection framework (COLIDE) for IoT leveraging collaboration among resource-constrained sensor and border nodes for effective and timely detection of intruders. The paper presents a detailed formal description of the proposed framework along with analysis to assess its effectiveness for a typical IoT system. We implemented the COLIDE framework with Contiki OS and conducted thorough experimentation to evaluate its performance. The evaluation demonstrates efficiency of COLIDE framework with respect to energy and processing overheads achieving effectiveness within an IoT system.

Optimising the overall power in a cluster-assisted internet of things (IoT) network is a challenging problem for emerging IoT applications. In this study, the authors propose a mathematical model for the cluster-assisted IoT network. The cluster-assisted IoT network consists of three types of nodes: IoT nodes, core cluster nodes (CCNs) and base stations (BSs). The objective is to minimise transmission, between IoT nodes (IoTs)–CCNs and CCNs–BSs, and computational power (at CCNs), while satisfying the requirements of communicating nodes. The formulated mathematical model is a integer programming problem. They propose three swarm intelligence-based evolutionary algorithms: (i) a discrete fireworks algorithm (DFWA), (ii) a load-aware DFWA (L-DFWA), and (iii) a hybrid of the L-DFWA and the low-complexity biogeography-based optimisation algorithm to solve the optimisation problem. The proposed algorithms are population-based metaheuristic algorithms. They perform extensive simulations and statistical tests to show the performance of the proposed algorithms when compared with the existing ones.

In the era of the Internet of Things (IoT), smart connected devices have the ability to generate data that could be of interest to the public. This paves the way for an emerging market for monetized data exchanges, where IoT device owners can sell access to live data generated by their connected devices to interested users. Implementing a trusted, cost-efficient, automatic monetization solution of IoT data can be a challenging problem and usually involves intermediaries and centralised governance and management. Blockchain and smart contracts introduce a secure and trusted platform to carry out transactions in a highly trusted, secure, decentralised manner. In this study, they present a blockchain-solution and implementation using Ethereum smart contracts for monetizing IoT data with automated payment involving no intermediary. The authors discuss key aspects related to architectural design, entity relations, interactions among participants, logic flow, implementation and testing of the overall system functionality.

A new model for malicious file propagation and transmission in a computer network is formulated. This model introduced a cure rate (E-cure rate) which is calculated after the regular running of anti-virus in the systems of the computer network. This will result in monitoring the anti-virus efficiency in fighting not only a local malicious file but also the global spreading of such a file. The file has to go through different scenarios including many other states. Therefore, a vertical transmission concept is introduced in the proposed model. The threshold conditions for the stability of the proposed model are determined. Numerous methods are used to prove that the viral equilibrium points of our model are globally and locally stable. The behaviour of the compartment nodes in the computer network has been analysed. In addition, the optimal control theory is used to study immunisation strategies against Malware in order to control the economic losses of security investments and keep infection loss at the lowest possible level. The existence and uniqueness of results related to the system are confirmed. Finally, the simulation shows that the spread of malware can be effectively controlled by choosing the appropriate control strategy of specific parameters.

In a typical healthcare monitoring system, the cloud is the preferred platform to aggregate, store and analyse data collected from Medical Internet of Things (or MIoT) devices. However, remote cloud servers and storage can be a source of substantial delay. To overcome such delays, an intermediate layer of fog or edge nodes is used for localised processing and storage of MIoT data. To this end, an integrated architecture consisting of MIoT devices, fog and cloud computing has now become the most preferred solution for a healthcare monitoring system. In this study, we propose an analytical model of such a system and use it to show how to reduce the cost for computing resources while guaranteeing performance constraints. The proposed analytical model is based on network of queues and has the ability to estimate the minimum required number of computing resources to meet the service level agreement. The authors verify and cross-validate the analytical model through Java modelling tools discrete event simulator. Results obtained from analysis and simulation show that the proposed model, under different workload conditions, can predict the system response time, and can accurately determine the number of computing resources needed for health data services to achieve desired performance.

In this study, a modified fluid model is proposed to understand the design of controlled delay (CoDel) algorithm for active queue management (AQM) and analyse its sensitivity to parameter settings. CoDel significantly differs from other AQM algorithms because it operates at the head of the queue and adopts a deterministic packet drop strategy, unlike other algorithms that operate at the tail and adopt a probabilistic packet drop strategy. The correctness of the proposed fluid model is verified by comparing its results with those obtained from ns-2. Subsequently, using the model developed in this study, the authors analyse the performance of CoDel algorithm by changing its internal parameters and modifying its control law. They highlight the role of the internal parameters and control law on the ability of the CoDel algorithm to control queue delay. Their analysis shows that the CoDel algorithm is sensitive to its parameter settings and that its control law requires minor modifications to gain a better control over the queue delay.

Address resolution protocol (ARP) cache poisoning is mostly used to perform man-in-the-middle (MITM) and denial of service (DoS) attacks for sniffing and network services disruption, respectively, in switch LAN networks. The former attack affects the privacy of a network user, whereas the latter attack causes huge damage to networks in terms of unavailability of services. There are many security solutions that provide protection against such type of attacks at firewall level with the assumption that an insider attack does not exist. However, the most damaging and severe attacks on universities database servers to well-known banking system have been caused by insider malicious users. In this study, the authors propose an agent-based ARP cache poisoning detection (AACPD) scheme that detects the MITM and DoS attacks in a switch LAN environment. The objective of this scheme is to provide protection from insider malicious users. The proposed AACPD is compared with similar existing schemes in a real-time environment. The experimental results show that AACPD efficiently and accurately detects attackers with minimum effect on network performance as compared to similar existing schemes.

Traffic signs have come a long way since the first automobile was invented. They have long served the purpose of warning and guiding drivers and also enforcing the traffic laws governing speed, parking, turns, and stopping. In this study, the authors discuss the issues and challenges facing current traffic signs, and how it will evolve into a next-generation traffic sign architecture using advanced wireless communications technologies. With technological advances in the areas of wireless communications and embedded electronics and software, we foresee that, in the future, digital traffic sign posts will be capable of transmitting the traffic sign information wirelessly to road users, and this will transform our roads into intelligent roads, where signs will appear promptly and automatically on in-vehicle displays to alert the driver. There is no longer the need to watch out for traffic signs since the detection will be automatic and performed wirelessly. This transformation will lessen burden on the drivers, so that they can then focus more on the traffic ahead while driving. Also, this evolution into wireless digital sign posts will fit well with the vision of future smart cities, where smart transportation technologies will be present to transform how we drive and commute, yielding greater safety, ease, and assistance to drivers.