Complex machine learning (ML) algorithms consistently outperform traditional ML models, yielding significantly improved results. Furthermore, these complex system uses large amounts of data with many features to train, increasing their predictive power but reducing our ability to explain them. Thus, complex ML models primarily operate like black boxes. Explaining their inference processes and predictions poses a challenge due to the necessity to account for millions of interacting weights in a complicated way. Still, we need to clarify their inner workings and procedures, as understanding a model can improve transparency and promote trustworthiness. Therefore, ‘Explainability’ assumes pivotal importance, as it enhances accuracy and broadens the scope of applicability for complex ML methods within critical domains. Thus, in this review paper, we have investigated the complete ML system development lifecycle, compiled recommendations to cover explainability gaps from diverse literature sources, listed different methods of interpretability and explainability, and improved our explainability-augmented ML system development life cycle for all stakeholders to enhance trustworthiness.

Proper security solutions in the cyber world are crucial for enforcing network security by providing real-time network protection against network vulnerabilities and data exploitation. An effective intrusion detection strategy is capable of taking a holistic approach for protecting critical systems against unauthorized access or attack. In this paper, we describe a machine learning (ML) based comprehensive security solution for network intrusion detection using ensemble supervised ML framework and ensemble feature selection methods. In addition, we provide a comparative analysis of several ML models and feature selection methods. The goal of this research is to design a generic detection mechanism and achieve higher accuracy with minimal false positive rates (FPR). NSL-KDD, UNSW-NB15, and CICIDS2017 datasets are used in the experiment, and results show that our detection model can identify 99.3% of intrusions successfully with the lowest 0.5% of false alarms, which depicts better performance metrics compared to existing solutions.

As the Internet of things continues to mature, there has been a wide and rapid adoption of Internet of things solutions in many domains. However, the security and privacy of these solutions are often underrated. The lack of security and privacy in the solutions can lead to catastrophic results, especially in sensitive domains like healthcare. Therefore, it is imperative to ensure their security and privacy. Stakeholders often find themselves unguided when assessing the security and privacy of the solutions. This paper proposes a framework to quantitatively assess and compare the security and privacy in Internet of things solutions using Analytic Hierarchy Process. It also provides in-depth security and privacy assessment criteria to enhance security and privacy in the Internet of things solutions. This work aims to fill the gap between consumers and providers by promoting transparency and increasing the security awareness.

Moving data and applications to the cloud implies shifting their control from cloud consumers to the cloud service provider (CSP) indefinitely. Hence, the security and privacy (S&P) of the consumers' assets becomes an important issue. Assessing and comparing potential cloud computing (CC) services, poses an issue for CC adopters to choose S&P options that are sufficient and robust at the same time. In this paper, we describe CC adopters' S&P concerns. We identify a set of attributes that reflect various aspects of cloud S&P in an attempt to alleviate those concerns. We classify the attributes based on their tangibility and their relevance to S&P threats. Every attribute in our list is represented by a set of considerations (i.e. polar questions) to assess its quality. We present a tool that embodies our set of attributes. We aim to enable consumers to assess and compare CC services, in terms of S&P, so that they can make well-educated choices. CSPs can also make use of these attributes to offer better cloud S&P.

The emerging paradigm of cloud computing (CC) presents many security risks that can potentially and adversely impact any one of the plethora of stakeholders. The widespread deployment and service models of CC in addition to the wide variety of stakeholders make it difficult to comprehend security and privacy (S&P). In this paper, we present CSSR1, a Cloud Services Security Recommender tool. CSSR codifies a stakeholder-oriented taxonomy. The goal for CSSR is to identify the various S&P risks for the kaleidoscope of different CC models from the stakeholder's perspective. CSSR will recommend a comprehensive list of S&P attributes that must be considered as controls necessary to minimize the CC attack surface. By identifying the S&P concerns that are unique to the particular usage scenarios (again from a stakeholder perspective), CSSR provides a comprehensive basis from which to choose alternative security solutions. This model then provides a structured and well-informed process of mitigating risk as envisioned by each and every stakeholder based on their needs.

Cloud computing is a promising technology and its adoption is increasing rapidly. Still, many potential cloud customers are concerned about the security, privacy, and availability of their data in the cloud as they have less control over it. Cloud certification authorities can perform rigorous and thorough testing based on the well-known security standards procedures and make the audit reports public. Cloud customers can gain confidence by viewing those reports. However, the certification process for cloud services is more difficult than that of traditional IT services. In this paper, we have addressed the importance and challenges of the cloud service certification process and made a comparison among the cloud certification authorities which will enable potential cloud customers to evaluate the certification authorities.

There have been attempts to model the interaction between users, both malicious and benign, and network administrators as games. Building on such works, we here present a game model which is generic enough to capture various modes of such interactions. The model facilitates stochastic games with imperfect information. The imperfect information is due to erroneous sensors leading to incorrect perception of the current state by the players. To model this error in perception distributed over other multiple states, we use Euclidean distances between inputs from the sensors.

ABSTRACT Game theory offers a promising approach toward modeling cyber attacks and countermeasures as games played among attackers and system defenders. The widely accepted concept of Nash equilibrium can be used to determine the optimal strategy for all players playing the game. In this work, we demonstrate the applicability of game theory in modeling the defense against distributed denial of service attacks. In particular, we focus on attack scenarios that attempt to create congestion in bottleneck network links. We design a game model to determine the best defense strategy for the defender in such attack scenarios. In our model, the attacker uses a botnet consisting of multiple nodes to send rogue traffic in order to flood one or more links in the target network. The defender's challenge is to determine the best firewall settings to block rogue traffic while allowing legitimate traffic. We validate our game model using simulation. Furthermore, we build a comprehensive defense architecture called game inspired defense architecture and test its performance via emulation on one publicly available testbed called DeterLab. Open source software, such as Bro intrusion detection system, Dummynet network emulator, IP firewall, and Iperf network testing tool are used to build game inspired defense architecture. Our proposed defense solution exhibits promising results during both simulation and emulation. Copyright © 2014 John Wiley & Sons, Ltd.

The emerging paradigm of cloud computing (CC) arises security risks that adversely impact its different stakeholders. The widespread deployment and service models of CC in addition to the wide variety of stakeholders make it difficult to guarantee privacy and security. This work-in-progress paper proposes a stakeholder-oriented taxonomical approach that determines the security and privacy issues for various CC models from a stakeholder's perspective. It recommends a comprehensive list of security and privacy attributes that are related to these issues. The goal is to provide stakeholders with the security issues associated with their interaction with the cloud. This will assist every stakeholder of a CC model in identifying the security and privacy concerns that are pertinent to them, based on the service and deployment model they use. This will help promote security and privacy among the stakeholders and refute their fears from utilizing this emerging technology.

Denial of service (DoS) attacks are currently one of the biggest risks any organization connected to the Internet can face. Hence, the congestion handling techniques at its edge router(s), such as active queue management (AQM) schemes must consider possibilities of such attacks. Ideally, an AQM scheme should (a) ensure that each network flow gets its fair share of bandwidth, and (b) identify attack flows so that corrective actions (e.g. drop flooding traffic) can be explicitly taken against them to further mitigate the DoS attacks. This paper presents a proof-of-concept work on devising such an AQM scheme, which we name Deterministic Fair Sharing (DFS). Most of the existing AQM schemes do not achieve the above goals or have a significant room for improvement. DFS uses the concept of weighted fair share (wfs) which allows it to dynamically self-adjust the router buffer usage based on the current level of congestion, while assuring fairness among flows and aiding in identifying the malicious ones. We demonstrate the performance of DFS via extensive simulation and compare against other existing AQM techniques.

While there are significant advances in information technology and infrastructure which offer new opportunities, cyberspace is still far from completely secured. In many cases, the employed security solutions are ad hoc and lack a quantitative decision framework. To this end, game theory poses huge potential in building a defense architecture based on a solid analytical setting. In this paper, we explore the applicability of game theoretic approaches to the cyber security problem while keeping the focus on active bandwidth depletion attacks on TCP/TCP-friendly flows. We model the interaction between the attacker and the defender as a game in two attack scenarios: (i) one single attacking node for Denial of Service (DoS) and (ii) multiple attacking nodes for Distributed DoS (DDoS). The defender's challenge is to determine optimal firewall settings to block rogue traffic while allowing legitimate ones. Our analysis considers the worst-case scenario where the attacker also attempts to find the most effective sending rate or botnet size. In either case, we build a static game model to compute the Nash equilibrium that represents the best strategy for the defender. We validate the effectiveness of our game theoretic defense mechanisms via extensive simulation.

It is often assumed that to pursue knowledge management, a knowledge management system must be independently developed or purchased. The intent of this paper is to illustrate that a knowledge management framework can be designed by combining existing technologies and techniques. While some development work may be required, it should be largely limited to the integration required to produce and fill gaps that are critical to a usable knowledge framework. This recommended approach will be demonstrated in a globally diverse corporate environment through the implementation of a proof-of-concept project. Promising results appropriate for the maturity level of the proof-of-concept implementation are documented. With management support gained, future plans for maturing the knowledge sharing project and evaluating the impact on the organization will be presented.

There are many stakeholders in the information technology (IT) domain, including the business community and software development and support staff. A high degree of collaboration and knowledge sharing is required to take a system from business concept to production implementation in complex organizations where numerous teams are involved in the software engineering process. Much of historical knowledge management strategy has a technology focus with knowledge collected, organized, and disseminated with a push or pull approach. An approach that considers common technology implementations, such as intranets, document management systems, and email, and how those can be exploited with little or no introduction of new processes or tools to promote knowledge sharing is suggested. Human and social aspects are considered fundamental to the success of a solution that leverages current technology implementations and capitalizes on the relationships that exist and are created among individuals in the context of software engineering work.

This paper discusses two related challenges faced by software engineering instructors. First, assuming that projects are necessary to produce successful computer science majors, what should be the role of projects and how best do we integrate theory and application? Second, what life cycle models and associated processes should students have the opportunity to experience, and where in the curriculum should a disciplined process first appear? We review several curriculum plans that have been employed to address these problems. We also offer recommendations based on our experiences both with undergraduate computer science majors and with high school students in project Tri-P-LETS, where beginning programming students are taught to develop games and software simulations following a process.

With the increased complexity of integrated circuits, a true top-down methodology is mandatory in their design. A construct that enables a modular description of a system design is added to DDL. The Translator, Simulator and Synthesis Software has been modified to retain this modularity throughout the design cycle. These enhancements allow a multi-level simulation and multi-technology synthesis of an integrated circuit.

Hardware Description Languages are used to input the details of a digital system into an automatic design system. An algorithm to synthesize combinational logic from the description in one such language (DDL) is discussed. A sample implementation and the cost comparison are provided.

A family of multiple-valued (MV) electronic memory elements, referred to herein as flip-flops, is presented along with a system of MV algebra upon which they are based. These MV flip-flops are compared to binary flip-flops. MV asynchronous set-clear flip-flops and synchronous set-clear, D-type, JK, and modulo N counter flip-flops are presented, their next-state equations are derived, and they are shown to have desirable properties for use in MV sequential circuits. Experimental results and schematic diagrams are presented for a level restoring three-valued logic gate, the clocked set-clear flip-flop, and an example synchronous sequential circuit.