Trojan

About: Trojan is a research topic. Over the lifetime, 2028 publications have been published within this topic receiving 33209 citations.

Computer viruses

Abstract: This paper presents an overview of the main categories of malicious programs known as Trojan horses, viruses, bacteria, worms, and logic bombs. The focus is on their general behavior and the properties seen in their implementations rather than the ultimate effects or their intended destructive behavior. Possible preventive measures are also discussed. >

Golden IC free methodology for hardware Trojan detection using symmetric path delays

Abstract: Hardware Trojans can be inserted by an adversary at any phase of IC manufacturing. In this paper, a methodology is proposed to detect Trojans inserted after design sign-off i.e the Trojan insertion occurs at layout level. In such attack models, golden IC are not always available in all cases, thus requiring golden IC free detection methodologies. This work exploits the concept of symmetric path delays to detect Trojans, considering the change in delays of symmetric pairs due to Trojan insertion. We propose detection metric (DM) of a suspect IC and compare the same with a detection threshold (DT) to decide whether IC under purview is Trojan free. Moreover, this method does not require any golden IC. Additionally, this method is robust enough against process variation effects. Simulation results establish that, a detection rate of 100% is achievable with maximum of 8% intra-die and 10% inter-die variation in both threshold voltage (V th ) and length (L), respectively.

Application of Machine Learning in Hardware Trojan Detection

Abstract: Hardware Trojans (HTs), maliciously inserted in an integrated circuit during untrusted design or fabrication process pose critical threat to the system security. With the ever increasing capabilities of an adversary to subvert the system during run-time, it is imperative to detect the manifested Trojans in order to reinforce the trust in hardware. In this regard, Machine Learning (ML) algorithms, with their intrinsic capability to execute feature engineering at high learning rates, are emerging as promising candidates to be utilized by system defenders. In this paper, we explore Trojan detection mechanisms that are based on ML, and thereby investigate the prowess of the ML algorithms in bolstering system security. Furthermore, we analyze the efficiency of each proposed Trojan detection strategy based on the underlying ML algorithm. Finally, we underline some problems with existing Trojan detection approaches and discuss future research in the interest of improved performance of the employed ML algorithms, thus aiding in enhancing the intended hardware security.

On the dynamics of Trojan planets in extrasolar planetary systems

Abstract: Abstract In this article we examine the motion of fictitious Trojan planets close to the equilateral Lagrangean equilibrium points in extrasolar planetary systems. Whether there exist stable motion in this area or not depends on the massratio of the primariy bodies in the restricted three body problem, namely the host star and the gasgiant. Taking into account also the eccentricity of the primaries we show via results of extensive numerical integrations that Trojan planets may survive only for e < 0.25. We also show first results of a mapping in the 1:1 resonance with a gas giant on an eccentric orbit which is applied to the extrasolar planetary systems HD 17051. We furthermore study the influence of an additional outer planet which perturbs the motion of the gasgiant as well as the Trojan cloud around its L4 Lagrangean point.