scispace - formally typeset
Search or ask a question
Topic

Trojan

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


Papers
More filters
Proceedings ArticleDOI
01 Feb 2017
TL;DR: This proposed method is suitable to detect Trojans which gets activated by aging or Trojan that gets activated after a specific time period and Hamming distance Pattern optimization technique improves the toggling inside the regions which makes the Trojan effect more observable.
Abstract: In today's scenario, any design is vulnerable to third party attackers like Hardware Trojan depends on their accessibility to the design. Hardware Trojans are the pernicious circuitry which are added to an original design by an attacker. These Trojans may lead to erroneous output of a system or even can incapacitate the design. Hence there is a demand for detecting the Trojans. The proposed method segments the whole design into small regions and generates fingerprint out of those. The number of fingerprints again reduced by selection of only regions containing low activity nodes in the first iteration due to the probability of inserting a Trojan in the low activity node is more. Segmentation helps to improve to locate where exactly the Trojan is inserted. This proposed method is suitable to detect Trojans which gets activated by aging or Trojan that gets activated after a specific time period. Hamming distance Pattern optimization technique improves the toggling inside the regions which makes the Trojan effect more observable. Proposed algorithm is implemented and validated using ISCAS'85 benchmark circuits.

6 citations

Journal ArticleDOI
TL;DR: STRIP-ViTA as discussed by the authors detects trigger inputs with small false acceptance rate (FAR) with an acceptable preset false rejection rate (FRR) by setting FRR to be 3 percent, average FAR of 1.1 and 3.55 percent are achieved for text and audio tasks, respectively.
Abstract: Trojan attacks on deep neural networks (DNNs) exploit a backdoor embedded in a DNN model that can hijack any input with an attacker’s chosen signature trigger. Emerging defence mechanisms are mainly designed and validated on vision domain tasks (e.g., image classification) on 2D Convolutional Neural Network (CNN) model architectures; a defence mechanism that is general across vision, text, and audio domain tasks is demanded. This work designs and evaluates a run-time Trojan detection method exploiting STR ong I ntentional P erturbation of inputs that is a multi-domain input-agnostic Trojan detection defence across Vi sion, T ext and A udio domains—thus termed as STRIP-ViTA. Specifically, STRIP-ViTA is demonstratively independent of not only task domain but also model architectures. Most importantly, unlike other detection mechanisms, it requires neither machine learning expertise nor expensive computational resource, which are the reason behind DNN model outsourcing scenario—one main attack surface of Trojan attack. We have extensively evaluated the performance of STRIP-ViTA over: i) CIFAR10 and GTSRB datasets using 2D CNNs for vision tasks; ii) IMDB and consumer complaint datasets using both LSTM and 1D CNNs for text tasks; and iii) speech command dataset using both 1D CNNs and 2D CNNs for audio tasks. Experimental results based on more than 30 tested Trojaned models (including publicly Trojaned model) corroborate that STRIP-ViTA performs well across all nine architectures and five datasets. Overall, STRIP-ViTA can effectively detect trigger inputs with small false acceptance rate (FAR) with an acceptable preset false rejection rate (FRR). In particular, for vision tasks, we can always achieve a 0 percent FRR and FAR given strong attack success rate always preferred by the attacker. By setting FRR to be 3 percent, average FAR of 1.1 and 3.55 percent are achieved for text and audio tasks, respectively. Moreover, we have evaluated STRIP-ViTA against a number of advanced backdoor attacks and compare its effectiveness with other recent state-of-the-arts.

6 citations

Proceedings Article
01 Jan 2015
TL;DR: This research examines the systemic threat of hardware trojans with an actual hardware trojan implementation to evaluate the impact and can degrade network services inside a corporate network, controllable from outside the network.
Abstract: Hardware trojans are a systemic threat that can impact the operations and infrastructure of corporations and government organisations. In this paper, we evaluate a credible and organisation-wide hardware trojan threat from compromised network cards. Our research examines the systemic threat of hardware trojans with an actual hardware trojan implementation to evaluate the impact. Our hardware trojan can degrade network services inside a corporate network, controllable from outside the network. An external activation mechanism is used to activate the trojan; the implementation bypasses data encryption, firewall packet inspection, and is agnostic to software protection and the operating system.

6 citations

29 Jan 2013
TL;DR: An architecture that fragments and replicates computation over a pool of Commercial-Off-The-Shelf processors with widely heterogeneous architectures that provides integrity, data confidentiality, and availability for executing applications is presented.
Abstract: Hardware Trojans pose a credible and increasing threat to computer security, with the potential to compromise the very electronics that ostensibly provide the security primitives underpinning various computer architectures. The discovery of stealthy Hardware Trojans within Integrated Circuits by current state-of-the-art pre-and post-manufacturing test and verification techniques cannot be guaranteed. Therefore electronic systems, especially those controlling safety or security critical systems should be designed to operate with integrity in the presence of any Hardware Trojans, and regardless of any Trojan activity. We present an architecture that fragments and replicates computation over a pool of Commercial-Off-The-Shelf processors with widely heterogeneous architectures. Processors are loosely synchronised through their use of a voted, architecture-independent message box mechanism to access a common memory space. A minimal Trusted Computing Base abstracts the processors as a single computational entity that can tolerate the effects of arbitrary Hardware Trojans within individual processors. The architecture provides integrity, data confidentiality, and availability for executing applications.

6 citations

Patent
13 May 2009
TL;DR: In this paper, a control terminal (240) is connected with server (210) through Internet (230) so that control terminal(240) transmits instruction to server (20) and /or server (10) transmittes data back to control node (240).
Abstract: The invention relates to network safety system and method, more concretely a network safety system and method for preventing Trojan. A control terminal (240) is connected with server (210) through Internet (230) so that control terminal (240) transmits instruction to server (210) and /or server (210) transmits data back to control terminal (240). The network safety system of the invention identifies Trojan virus through virus identification module (221), performs protection operation for server (210) through strategy control module (222). In one embodiment, Trojan virus is identified based on data of application layer, while in another embodiment based on data of application layer and TCP head or IP head information, and based on relation between data packet during connection. The network safety system and method for preventing Trojan of the invention are widely used among LAN and single PC.

6 citations


Network Information
Related Topics (5)
Cloud computing
156.4K papers, 1.9M citations
70% related
Cache
59.1K papers, 976.6K citations
70% related
Planet
27K papers, 980.6K citations
68% related
Compiler
26.3K papers, 578.5K citations
66% related
Key (cryptography)
60.1K papers, 659.3K citations
66% related
Performance
Metrics
No. of papers in the topic in previous years
YearPapers
2023136
2022282
2021111
2020139
2019144
2018168