Trojan

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

Automated Design of Propellant-Optimal, Low-Thrust Trajectories for Trojan Asteroid Tours

Abstract: The sun–Jupiter Trojan asteroid swarms are targets of interest for robotic spacecraft missions, where low-thrust propulsion systems offer a viable approach for realizing tours of these asteroids. This investigation introduces a novel scheme for the automated creation of prospective tours under the natural dynamics of a multibody regime with thrust supplied by a variable-specific-impulse engine. The procedure approximates tours by combining independently generated fuel-optimal rendezvous arcs between asteroid pairs into a series of trajectory legs. Propellant costs as well as departure and arrival times are estimated from the performance of the individual thrust arcs. Tours of interest are readily constructed in higher-fidelity models, and options for the end-to-end trajectories are easily assessed. In this investigation, scenarios where constant and varying-power sources are available to the low-thrust engine are explored. In general, the automation procedure rapidly generates a large number of potential ...

Identification and break of positive feedback loops in Trojan States Vulnerable Circuits

Abstract: A systematic method is proposed for automatically identifying and breaking positive feedback loops (PFLs) in Trojan States Vulnerable Circuit. The method first converts the netlist of a circuit into a directed dependency graph (DDG) and then partitions the DDG into strongly connected components (SCCs). It then employs graph theory techniques to detect all PFLs and locate the break-points for every SCC. The proposed method could identify the circuit's vulnerability to Trojan States only by its structure without the computation of DC solutions and it also provides insights on how and where to break the PFLs such that break-loop continuation methods can be applied. With Sub-Bandgap reference and widlar-Banba examples, it is demonstrated that the proposed approach can effectively identify all the PFLs and break-points.

CLEANN: Accelerated Trojan Shield for Embedded Neural Networks

Abstract: We propose CLEANN, the first end-to-end framework that enables online mitigation of Trojans for embedded Deep Neural Network (DNN) applications. A Trojan attack works by injecting a backdoor in the DNN while training; during inference, the Trojan can be activated by the specific backdoor trigger. What differentiates CLEANN from the prior work is its lightweight methodology which recovers the ground-truth class of Trojan samples without the need for labeled data, model retraining, or prior assumptions on the trigger or the attack. We leverage dictionary learning and sparse approximation to characterize the statistical behavior of benign data and identify Trojan triggers. CLEANN is devised based on algorithm/hardware co-design and is equipped with specialized hardware to enable efficient real-time execution on resource-constrained embedded platforms. Proof of concept evaluations on CLEANN for the state-of-the-art Neural Trojan attacks on visual benchmarks demonstrate its competitive advantage in terms of attack resiliency and execution overhead.