Biblio

Autonomy in cybersystems depends on their ability to be self-aware by understanding the intent of services and applications that are running on those systems. In case of mission-critical cybersystems that are deployed in dynamic and unpredictable environments, the newly integrated unknown applications or services can either be benign and essential for the mission or they can be cyberattacks. In some cases, these cyberattacks are evasive Advanced Persistent Threats (APTs) where the attackers remain undetected for reconnaissance in order to ascertain system features for an attack e.g. Trojan Laziok. In other cases, the attackers can use the system only for computing e.g. cryptomining malware. APTs such as cryptomining malware neither disrupt normal system functionalities nor trigger any warning signs because they simply perform bitwise and cryptographic operations as any other benign compression or encoding application. Thus, it is difficult for defense mechanisms such as antivirus applications to detect these attacks. In this paper, we propose an Operating Context profiling system based on deep neural networks-Long Short-Term Memory (LSTM) networks-using Windows Performance Counters data for detecting these evasive cryptomining applications. In addition, we propose Deep Cryptomining Profiler (DeCrypto Pro), a detection system with a novel model selection framework containing a utility function that can select a classification model for behavior profiling from both the light-weight machine learning models (Random Forest and k-Nearest Neighbors) and a deep learning model (LSTM), depending on available computing resources. Given data from performance counters, we show that individual models perform with high accuracy and can be trained with limited training data. We also show that the DeCrypto Profiler framework reduces the use of computational resources and accurately detects cryptomining applications by selecting an appropriate model, given the constraints such as data sample size and system configuration.

5G mobile networks promise universal communication environment and aims at providing higher bandwidth, increased communication and networking capabilities, and extensive signal coverage by using multiple communication technologies including Device-to-Device (D-to-D). This paradigm, will allow scalable and ubiquitous connectivity for large-scale mobile networks where a huge number of heterogeneous devices with limited resources will cooperate to enhance communication efficiency in terms of link reliability, spectral efficiency, system capacity, and transmission range. However, owing to its decentralized nature, cooperative D-to-D communication could be vulnerable to attacks initiated on relay nodes. Consequently, a source node has the interest to select the more protected relay to ensure the security of its traffic. Nevertheless, an improvement in the protection level has a counterpart cost that must be sustained by the device. To address this trade-off as well as the interaction between the attacker and the source device, we propose a dynamic game theoretic based approach to model and analyze this problem as a cost model. The utility function of the proposed non-cooperative game is based on the concepts of return on protection and return on attack which illustrate the gain of selecting a relay for transmitting a data packet by a source node and the reward of the attacker to perform an attack to compromise the transmitted data. Moreover, we discuss and analyze Nash equilibrium convergence of this attack-defense model and we propose an heuristic algorithm that can determine the equilibrium state in a limited number of running stages. Finally, we perform simulation work to show the effectiveness of the game model in assessing the behavior of the source node and the attacker and its ability to reach equilibrium within a finite number of steps.

Radio frequency identification wireless sensing technology widely adopted and developed from last decade and has been utilized for monitoring and autonomous identification of objects. However, wider utilization of RFID technologies has introduced challenges such as preserving security and privacy of sensitive data while maintaining the high quality of service. Thus, in this work, we will deliberately build up a RFID system by utilizing learning automata based multi agent intelligent system to greatly enhance and secure message transactions and to improve operational efficiency. The incorporation of these two advancements and technological developments will provide maximum benefit in terms of expertly and securely handle data in RFID scenario. In proposed work, learning automata inbuilt RFID tags or assumed players choose their optimal strategy via enlarging its own utility function to achieve long term benefit. This is possible if they transmit their utility securely to back end server and then correspondingly safely get new utility function from server to behave optimally in its environment. Hence, our proposed authentication protocol, expertly transfer utility from learning automata inbuilt tags to reader and then to server. Moreover, we verify the security and privacy of our proposed system by utilizing automatic formal prover Scyther tool.

With the ever so growing boundaries for security in the cloud, it is necessary to develop ways to prevent from total cloud server failure. In this paper, we try to design a Game Strategy Block that sets up rules for security based on a tower defence game to secure the hypervisor from potential threats. We also try to define a utility function named the Virtual Machine Vitality Measure (VMVM) that could enlighten on the status of the virtual machines on the virtual environment.

Gaussian random attacks that jointly minimize the amount of information obtained by the operator from the grid and the probability of attack detection are presented. The construction of the attack is posed as an optimization problem with a utility function that captures two effects: firstly, minimizing the mutual information between the measurements and the state variables; secondly, minimizing the probability of attack detection via the Kullback-Leibler (KL) divergence between the distribution of the measurements with an attack and the distribution of the measurements without an attack. Additionally, a lower bound on the utility function achieved by the attacks constructed with imperfect knowledge of the second order statistics of the state variables is obtained. The performance of the attack construction using the sample covariance matrix of the state variables is numerically evaluated. The above results are tested in the IEEE 30-Bus test system.