Biblio

The latest generation of IoT systems incorporate machine learning (ML) technologies on edge devices. This introduces new engineering challenges to bring ML onto resource-constrained hardware, and complications for ensuring system security and privacy. Existing research prescribes iterative processes for machine learning enabled IoT products to ease development and increase product success. However, these processes mostly focus on existing practices used in other generic software development areas and are not specialized for the purpose of machine learning or IoT devices. This research seeks to characterize engineering processes and security practices for ML-enabled IoT systems through the lens of the engineering lifecycle. We collected data from practitioners through a survey (N=25) and interviews (N=4). We found that security processes and engineering methods vary by company. Respondents emphasized the engineering cost of security analysis and threat modeling, and trade-offs with business needs. Engineers reduce their security investment if it is not an explicit requirement. The threats of IP theft and reverse engineering were a consistent concern among practitioners when deploying ML for IoT devices. Based on our findings, we recommend further research into understanding engineering cost, compliance, and security trade-offs.

With the global transition to the IPv6 (Internet Protocol version 6), IP (Internet Protocol) validation efficiency and IPv6 support from the aspect of network programming are gaining more importance. As global computer networks grow in the era of IoT (Internet of Things), IP address validation is an inevitable process for assuring strong network privacy and security. The complexity of IP validation has been increased due to the rather drastic change in the memory architecture needed for storing IPv6 addresses. Low-level programming languages like C/C++ are a great choice for handling memory spaces and working with simple devices connected in an IoT (Internet of Things) network. This paper analyzes some user-defined and open-source implementations of IP validation codes in Boost. Asio and POCO C++ networking libraries, as well as the IP security support provided for general networking purposes and IoT. Considering a couple of sample codes, the paper gives a conclusion on whether these C++ implementations answer the needs for flexibility and security of the upcoming era of IPv6 addressed computers.

Military networks consist of heterogeneous devices that provide soldiers with real-time terrain and mission intel-ligence. The development of next-generation Software Defined Networks (SDN)-enabled devices is enabling the modernization of traditional military networks. Commonly, traditional military networks take the trustworthiness of devices for granted. How-ever, the recent modernization of military networks introduces cyber attacks such as data and identity spoofing attacks. Hence, it is crucial to ensure the trustworthiness of network traffic to ensure the mission's outcome. This work proposes a Continuous Behavior-based Authentication (CBA) protocol that integrates network traffic analysis techniques to provide robust and efficient network management flow by separating data and control planes in SDN-enabled military networks. The evaluation of the CBA protocol aimed to measure the efficiency of the proposed protocol in realistic military networks. Furthermore, we analyze the overall network overhead of the CBA protocol and its accuracy to detect rogue network traffic data from field devices.

The increase of autonomy in autonomous surface vehicles development brings along modified and new risks and potential hazards, this in turn, introduces the need for processes and methods for ensuring that systems are acceptable for their intended use with respect to dependability and safety concerns. One approach for evaluating software requirements for claims of safety is to employ an assurance case. Much like a legal case, the assurance case lays out an argument and supporting evidence to provide assurance on the software requirements. This paper analyses safety and security requirements relating to autonomous vessels, and regulations in the automotive industry and the marine industry before proposing a generic cybersecurity and safety assurance case that takes a general graphical approach of Goal Structuring Notation (GSN).

The new architecture of transformer networks proposed in the work can be used to create an intelligent chat bot that can learn the process of communication and immediately model responses based on what has been said. The essence of the new mechanism is to divide the information flow into two branches containing the history of the dialogue with different levels of granularity. Such a mechanism makes it possible to build and develop the personality of a dialogue agent in the process of dialogue, that is, to accurately imitate the natural behavior of a person. This gives the interlocutor (client) the feeling of talking to a real person. In addition, making modifications to the structure of such a network makes it possible to identify a likely attack using social engineering methods. The results obtained after training the created system showed the fundamental possibility of using a neural network of a new architecture to generate responses close to natural ones. Possible options for using such neural network dialogue agents in various fields, and, in particular, in information security systems, are considered. Possible options for using such neural network dialogue agents in various fields, and, in particular, in information security systems, are considered. The new technology can be used in social engineering attack detection systems, which is a big problem at present. The novelty and prospects of the proposed architecture of the neural network also lies in the possibility of creating on its basis dialogue systems with a high level of biological plausibility.

Software-Defined Networking (SDN) can be a good option to support Industry 4.0 (4IR) and 5G wireless networks. SDN can also be a secure networking solution that improves the security, capability, and programmability in the networks. In this paper, we present and analyze an SDN-based security architecture for 4IR with 5G. SDN is used for increasing the level of security and reliability of the network by suitably dividing the whole network into data, control, and applications planes. The SDN control layer plays a beneficial role in 4IR with 5G scenarios by managing the data flow properly. We also evaluate the performance of the proposed architecture in terms of key parameters such as data transmission rate and response time.

Efficiency is essential to support responsiveness w.r.t. ever-growing datasets, especially for Deep Learning (DL) systems. DL frameworks have traditionally embraced deferred execution-style DL code that supports symbolic, graph-based Deep Neural Network (DNN) computation. While scalable, such development tends to produce DL code that is error-prone, non-intuitive, and difficult to debug. Consequently, more natural, less error-prone imperative DL frameworks encouraging eager execution have emerged at the expense of run-time performance. While hybrid approaches aim for the “best of both worlds,” the challenges in applying them in the real world are largely unknown. We conduct a data-driven analysis of challenges-and resultant bugs-involved in writing reliable yet performant imperative DL code by studying 250 open-source projects, consisting of 19.7 MLOC, along with 470 and 446 manually examined code patches and bug reports, respectively. The results indicate that hybridization: (i) is prone to API misuse, (ii) can result in performance degradation-the opposite of its intention, and (iii) has limited application due to execution mode incompatibility. We put forth several recommendations, best practices, and anti-patterns for effectively hybridizing imperative DL code, potentially benefiting DL practitioners, API designers, tool developers, and educators.

Supervisory control and data acquisition (SCADA) systems play pivotal role in the operation of modern critical infrastructures (CIs). Technological advancements, innovations, economic trends, etc. have continued to improve SCADA systems effectiveness and overall CIs’ throughput. However, the trends have also continued to expose SCADA systems to security menaces. Intrusions and attacks on SCADA systems can cause service disruptions, equipment damage or/and even fatalities. The use of conventional intrusion detection models have shown trends of ineffectiveness due to the complexity and sophistication of modern day SCADA attacks and intrusions. Also, SCADA characteristics and requirement necessitate exceptional security considerations with regards to intrusive events’ mitigations. This paper explores the viability of supervised learning algorithms in detecting intrusions specific to SCADA systems and their communication protocols. Specifically, we examine four supervised learning algorithms: Random Forest, Naïve Bayes, J48 Decision Tree and Sequential Minimal Optimization-Support Vector Machines (SMO-SVM) for evaluating SCADA datasets. Two SCADA datasets were used for evaluating the performances of our approach. To improve the classification performances, feature selection using principal component analysis was used to preprocess the datasets. Using prominent classification metrics, the SVM-SMO presented the best overall results with regards to the two datasets. In summary, results showed that supervised learning algorithms were able to classify intrusions targeted against SCADA systems with satisfactory performances.

Malicious software (malware) poses a significant threat to the security of our networks and users. In the ever-evolving malware landscape, Excel 4.0 Office macros (XL4) have recently become an important attack vector. These macros are often hidden within apparently legitimate documents and under several layers of obfuscation. As such, they are difficult to analyze using static analysis techniques. Moreover, the analysis in a dynamic analysis environment (a sandbox) is challenging because the macros execute correctly only under specific environmental conditions that are not always easy to create. This paper presents SYMBEXCEL, a novel solution that leverages symbolic execution to deobfuscate and analyze Excel 4.0 macros automatically. Our approach proceeds in three stages: (1) The malicious document is parsed and loaded in memory; (2) Our symbolic execution engine executes the XL4 formulas; and (3) Our Engine concretizes any symbolic values encountered during the symbolic exploration, therefore evaluating the execution of each macro under a broad range of (meaningful) environment configurations. SYMBEXCEL significantly outperforms existing deobfuscation tools, allowing us to reliably extract Indicators of Compromise (IoCs) and other critical forensics information. Our experiments demonstrate the effectiveness of our approach, especially in deobfuscating novel malicious documents that make heavy use of environment variables and are often not identified by commercial anti-virus software.

Emails are widely used as a form of communication and sharing files in an organization. However, email is widely used by cybercriminals to spread malware and carrying out cyber-attacks. We implemented an open-source email gateway in conjunction with a security sandbox for securing emails against malicious attachments. The email gateway scans all incoming and outgoing emails and stops emails containing suspicious files. An automated python script would then send the suspected email to the sandboxing element through sandbox API for further analysis, while the script is used also for the prevention of duplicate results. Moreover, the mail server administrator receives notifications from the email gateway about suspicious attachments. If detected attachment is a true positive based on the sandbox analysis result, email is deleted, otherwise, the email is delivered to the recipient. The paper describes in an empirical way the steps followed during the implementation, results, and conclusions of our research.

To train new staff to be efficient and ready for the tasks assigned is vital. They must be equipped with knowledge and skills so that they can carry out their responsibility to ensure smooth daily working activities. As transitioning to IPv6 has taken place for more than a decade, it is understood that having a dual-stack network is common in any organization or enterprise. However, many Internet users may not realize the importance of IPv6 security due to a lack of awareness and knowledge of cyber and computer security. Therefore, this paper presents an approach to educating people by introducing a security mechanisms model that can be applied in handling security challenges via network sandboxing by setting up an isolated dual stack network testbed using GNS3 to perform network security analysis. The finding shows that applying security mechanisms such as access control lists (ACLs) and host-based firewalls can help counter the attacks. This proves that knowledge and skills to handle dual-stack security are crucial. In future, more kinds of attacks should be tested and also more types of security mechanisms can be applied on a dual-stack network to provide more information and to provide network engineers insights on how they can benefit from network sandboxing to sharpen their knowledge and skills.

Mobile small cells that are enabled with Network Coding (NC) are seen as a potentially useful technique for Fifth Generation (5G) networks, since they can cover an entire city and can be put up on demand anywhere, any time, and on any device. Despite numerous advantages, significant security issues arise as a result of the fact that the NC-enabled mobile small cells are vulnerable to attacks. Intrusions are a severe security threat that exploits the inherent vulnerabilities of NC. In order to make NC-enabled mobile small cells to realize their full potential, it is essential to implement intrusion detection systems. When compared to homomorphic signature or hashing systems, homomorphic message authentication codes (MACs) provide safe network coding techniques with relatively smaller overheads. A number of research studies have been conducted with the goal of developing mobile small cells that are enabled with secure network coding and coming up with integrity protocols that are appropriate for such crowded situations. However, the intermediate nodes alter packets while they are in transit and hence the integrity of the data cannot be confirmed by using MACs and checksums. This research study has analyzed numerous intrusion detection models for NC enabled small cells. This research helps the scholars to get a brief idea about various intrusion detection models.

Sometimes we have the need to inject new services in an operational satellite, but as the injection of new codes in equipment that has communication link is a critical process due to the possibility of injection of broke or malicious codes, this document proposes a protocol for the safe injection of code in satellite microcontrollers of the CubeSat’ type. This protocol is based on the use of HMAC with SHA-3 to guarantee integrity and authenticity and is enhanced by the same security measures to mitigate communication link problems and satellite attacks, such as the guarantee of delivery and displacement between communication windows and periods of high processing.

Smart Security Solutions are in high demand with the ever-increasing vulnerabilities within the IT domain. Adjusting to a Work-From-Home (WFH) culture has become mandatory by maintaining required core security principles. Therefore, implementing and maintaining a secure Smart Home System has become even more challenging. ARGUS provides an overall network security coverage for both incoming and outgoing traffic, a firewall and an adaptive bandwidth management system and a sophisticated CCTV surveillance capability. ARGUS is such a system that is implemented into an existing router incorporating cloud and Machine Learning (ML) technology to ensure seamless connectivity across multiple devices, including IoT devices at a low migration cost for the customer. The aggregation of the above features makes ARGUS an ideal solution for existing Smart Home System service providers and users where hardware and infrastructure is also allocated. ARGUS was tested on a small-scale smart home environment with a Raspberry Pi 4 Model B controller. Its intrusion detection system identified an intrusion with 96% accuracy while the physical surveillance system predicts the user with 81% accuracy.

The Network-on-Chip (NoC) is the communication heart in Multiprocessors System-on-Chip (MPSoC). It offers an efficient and scalable interconnection platform, which makes it a focal point of potential security threats. Due to outsourcing design, the NoC can be infected with a malicious circuit, known as Hardware Trojan (HT), to leak sensitive information or degrade the system’s performance and function. An HT can form a security threat by consciously dropping packets from the NoC, structuring a Black Hole Router (BHR) attack. This paper presents an end-to-end secure interconnection network against the BHR attack. The proposed scheme is energy-efficient to detect the BHR in runtime with 1% and 2% average throughput and energy consumption overheads, respectively.

Soft real-time applications, including multimedia, gaming, and smart appliances, rely on specific architectural characteristics to deliver output in a time-constrained fashion. Any violation of application deadlines can lower the Quality-of-Service (QoS). The data sets associated with these applications are distributed over cores that communicate via Network-on-Chip (NoC) in multi-core systems. Accordingly, the response time of such applications depends on the worst-case latency of request/reply packets. A malicious implant such as Hardware Trojan (HT) that initiates a delay-of-service attack can tamper with the system performance. We model an HT that mounts a time-delay attack in the system by violating the path selection strategy used by the adaptive NoC router. Our analysis shows that once activated, the proposed HT increases the packet latency by 17% and degrades the system performance (IPC) by 18% over the Baseline. Furthermore, we propose an HT detection framework that uses packet traffic analysis and path monitoring to localise the HT. Experiment results show that the proposed detection framework exhibits 4.8% less power consumption and 6.4% less area than the existing technique.

The spread of the Internet of Things (IoT) and the use of smart control systems in many mission-critical or safety-critical applications domains, like automotive or aeronautical, make devices attractive targets for attackers. Nowadays, several of these are mixed-criticality systems, i.e., they run both high-criticality tasks (e.g., a car control system) and low-criticality ones (e.g., infotainment). High-criticality routines often employ Real-Time Operating Systems (RTOS) to enforce hard real-time requirements, while the tasks with lower constraints can be delegated to more generic-purpose operating systems (GPOS).Much of the control code for these devices is written in memory-unsafe languages such as C and C++. This makes them susceptible to powerful binary attacks, such as the famous Return-Oriented Programming (ROP). Control-Flow Integrity (CFI) is the most investigated security technique to protect against such threats. At now, CFI solutions for real-time embedded systems are not as mature as the ones for general-purpose systems, and even more, there is a lack of in-depth studies on how different operating systems with different security requirements and timing constraints can coexist on a single multicore platform.This paper aims at drawing attention to the subject, discussing the current scientific proposal, and in turn proposing a solution for an optimized asymmetric verification system for execution integrity. By using an embedded hypervisor, predefined cores could be dedicated to only high or low-criticality tasks, with the high-priority core being monitored by the lower-criticality core, relying on offline binary instrumentation and a light exchange of information and signals at runtime. The work also presents preliminary results about a possible implementation for multicore ARM platforms, running both RTOS and GPOS, both in terms of security and performance penalties.

Side-channel attacks have been a constant threat to computing systems. In recent times, vulnerabilities in the architecture were discovered and exploited to mount and execute a state-of-the-art attack such as Spectre. The Spectre attack exploits a vulnerability in the Intel-based processors to leak confidential data through the covert channel. There exist some defenses to mitigate the Spectre attack. Among multiple defenses, hardware-assisted attack/intrusion detection (HID) systems have received overwhelming response due to its low overhead and efficient attack detection. The HID systems deploy machine learning (ML) classifiers to perform anomaly detection to determine whether the system is under attack. For this purpose, a performance monitoring tool profiles the applications to record hardware performance counters (HPC), utilized for anomaly detection. Previous HID systems assume that the Spectre is executed as a standalone application. In contrast, we propose an attack that dynamically generates variations in the injected code to evade detection. The attack is injected into a benign application. In this manner, the attack conceals itself as a benign application and gen-erates perturbations to avoid detection. For the attack injection, we exploit a return-oriented programming (ROP)-based code-injection technique that reuses the code, called gadgets, present in the exploited victim's (host) memory to execute the attack, which, in our case, is the CR-Spectre attack to steal sensitive data from a target victim (target) application. Our work focuses on proposing a dynamic attack that can evade HID detection by injecting perturbations, and its dynamically generated variations thereof, under the cloak of a benign application. We evaluate the proposed attack on the MiBench suite as the host. From our experiments, the HID performance degrades from 90% to 16%, indicating our Spectre-CR attack avoids detection successfully.

Robot co-workers, like human co-workers, make mistakes that undermine trust. Yet, trust is just as important in promoting human-robot collaboration as it is in promoting human-human collaboration. In addition, individuals can signif-icantly differ in their attitudes toward robots, which can also impact or hinder their trust in robots. To better understand how individual attitude can influence trust repair strategies, we propose a theoretical model that draws from the theory of cognitive dissonance. To empirically verify this model, we conducted a between-subjects experiment with 100 participants assigned to one of four repair strategies (apologies, denials, explanations, or promises) over three trust violations. Individual attitudes did moderate the efficacy of repair strategies and this effect differed over successive trust violations. Specifically, repair strategies were most effective relative to individual attitude during the second of the three trust violations, and promises were the trust repair strategy most impacted by an individual's attitude.

The success of human-robot interaction is strongly affected by the people’s ability to infer others’ intentions and behaviours, and the level of people’s trust that others will abide by their same principles and social conventions to achieve a common goal. The ability of understanding and reasoning about other agents’ mental states is known as Theory of Mind (ToM). ToM and trust, therefore, are key factors in the positive outcome of human-robot interaction. We believe that a robot endowed with a ToM is able to gain people’s trust, even when this may occasionally make errors.In this work, we present a user study in the field in which participants (N=123) interacted with a robot that may or may not have a ToM, and may or may not exhibit erroneous behaviour. Our findings indicate that a robot with ToM is perceived as more reliable, and they trusted it more than a robot without a ToM even when the robot made errors. Finally, ToM results to be a key driver for tuning people’s trust in the robot even when the initial condition of the interaction changed (i.e., loss and regain of trust in a longer relationship).

For designing the interaction with robots in healthcare scenarios, understanding how trust develops in such situations characterized by vulnerability and uncertainty is important. The goal of this study was to investigate how technology-related user dispositions, anxiety, and robot characteristics influence trust. A second goal was to substantiate the association between hospital patients' trust and their intention to use a transport robot. In an online study, patients, who were currently treated in hospitals, were introduced to the concept of a transport robot with both written and video-based material. Participants evaluated the robot several times. Technology-related user dispositions were found to be essentially associated with trust and the intention to use. Furthermore, hospital patients' anxiety was negatively associated with the intention to use. This relationship was mediated by trust. Moreover, no effects of the manipulated robot characteristics were found. In conclusion, for a successful implementation of robots in hospital settings patients' individual prior learning history - e.g., in terms of existing robot attitudes - and anxiety levels should be considered during the introduction and implementation phase.

Cloud computing provides customers with enormous compute power and storage capacity, allowing them to deploy their computation and data-intensive applications without having to invest in infrastructure. Many firms use cloud computing as a means of relocating and maintaining resources outside of their enterprise, regardless of the cloud server's location. However, preserving the data in cloud leads to a number of issues related to data loss, accountability, security etc. Such fears become a great barrier to the adoption of the cloud services by users. Cloud computing offers a high scale storage facility for internet users with reference to the cost based on the usage of facilities provided. Privacy protection of a user's data is considered as a challenge as the internal operations offered by the service providers cannot be accessed by the users. Hence, it becomes necessary for monitoring the usage of the client's data in cloud. In this research, we suggest an effective cloud storage solution for accessing patient medical records across hospitals in different countries while maintaining data security and integrity. In the suggested system, multifactor authentication for user login to the cloud, homomorphic encryption for data storage with integrity verification, and integrity verification have all been implemented effectively. To illustrate the efficacy of the proposed strategy, an experimental investigation was conducted.

We present the new class of non-uniform Rowhammer access patterns that bypass undocumented, proprietary in-DRAM Target Row Refresh (TRR) while operating in a production setting. We show that these patterns trigger bit flips on all 40 DDR4 DRAM devices in our test pool. We make a key observation that all published Rowhammer access patterns always hammer “aggressor” rows uniformly. While uniform accesses maximize the number of aggressor activations, we find that in-DRAM TRR exploits this behavior to catch aggressor rows and refresh neighboring “victims” before they fail. There is no reason, however, to limit Rowhammer attacks to uniform access patterns: smaller technology nodes make underlying DRAM technologies more vulnerable, and significantly fewer accesses are nowadays required to trigger bit flips, making it interesting to investigate less predictable access patterns. The search space for non-uniform access patterns, however, is tremendous. We design experiments to explore this space with respect to the deployed mitigations, highlighting the importance of the order, regularity, and intensity of accessing aggressor rows in non-uniform access patterns. We show how randomizing parameters in the frequency domain captures these aspects and use this insight in the design of Blacksmith, a scalable Rowhammer fuzzer that generates access patterns that hammer aggressor rows with different phases, frequencies, and amplitudes. Blacksmith finds complex patterns that trigger Rowhammer bit flips on all 40 of our recently purchased DDR4 DIMMs, \$2.6 \textbackslashtimes\$ more than state of the art, and generating on average \$87 \textbackslashtimes\$ more bit flips. We also demonstrate the effectiveness of these patterns on Low Power DDR4X devices. Our extensive analysis using Blacksmith further provides new insights on the properties of currently deployed TRR mitigations. We conclude that after almost a decade of research and deployed in-DRAM mitigations, we are perhaps in a worse situation than when Rowhammer was first discovered.

Public Key Infrastructure (PKI) as a techno-policy ecosystem for establishing electronic trust has survived for several decades and evolved as the de-facto model for centralized trust in electronic transactions. In this paper, we study the PKI ecosystem that are prevailing in the South Asian and Oceanic countries and brief them. We also look at how PKI has coped up with the rapid technological changes and how policies have been realigned or formulated to strengthen the PKI ecosystem in these countries.

Phishing has become a prominent method of data theft among hackers, and it continues to develop. In recent years, many strategies have been developed to identify phishing website attempts using machine learning particularly. However, the algorithms and classification criteria that have been used are highly different from the real issues and need to be compared. This paper provides a detailed comparison and evaluation of the performance of several machine learning algorithms across multiple datasets. Two phishing website datasets were used for the experiments: the Phishing Websites Dataset from UCI (2016) and the Phishing Websites Dataset from Mendeley (2018). Because these datasets include different types of class labels, the comparison algorithms can be applied in a variety of situations. The tests showed that Random Forest was better than other classification methods, with an accuracy of 88.92% for the UCI dataset and 97.50% for the Mendeley dataset.