1.4CVJul 13, 2021
Dynamic Distribution of Edge Intelligence at the Node Level for Internet of ThingsHawzhin Mohammed, Tolulope A. Odetola, Nan Guo et al.
In this paper, dynamic deployment of Convolutional Neural Network (CNN) architecture is proposed utilizing only IoT-level devices. By partitioning and pipelining the CNN, it horizontally distributes the computation load among resource-constrained devices (called horizontal collaboration), which in turn increases the throughput. Through partitioning, we can decrease the computation and energy consumption on individual IoT devices and increase the throughput without sacrificing accuracy. Also, by processing the data at the generation point, data privacy can be achieved. The results show that throughput can be increased by 1.55x to 1.75x for sharing the CNN into two and three resource-constrained devices, respectively.
6.6CRJun 13, 2021
FeSHI: Feature Map Based Stealthy Hardware Intrinsic AttackTolulope Odetola, Faiq Khalid, Travis Sandefur et al.
To reduce the time-to-market and access to state-of-the-art techniques, CNN hardware mapping and deployment on embedded accelerators are often outsourced to untrusted third parties, which is going to be more prevalent in futuristic artificial intelligence of things (AIoT) systems. These AIoT systems anticipate horizontal collaboration among different resource-constrained AIoT node devices, where CNN layers are partitioned and these devices collaboratively compute complex CNN tasks. This horizontal collaboration opens another attack surface to the CNN-based application, like inserting the hardware Trojans (HT) into the embedded accelerators designed for the CNN. Therefore, there is a dire need to explore this attack surface for designing secure embedded hardware accelerators for CNNs. Towards this goal, in this paper, we exploited this attack surface to propose an HT-based attack called FeSHI. Since in horizontal collaboration of RC AIoT devices different sections of CNN architectures are outsourced to different untrusted third parties, the attacker may not know the input image, but it has access to the layer-by-layer output feature maps information for the assigned sections of the CNN architecture. This attack exploits the statistical distribution, i.e., Gaussian distribution, of the layer-by-layer feature maps of the CNN to design two triggers for stealthy HT with a very low probability of triggering. Also, three different novel, stealthy and effective trigger designs are proposed.
2.3CVJun 16, 2020
How Secure is Distributed Convolutional Neural Network on IoT Edge Devices?Hawzhin Mohammed, Tolulope A. Odetola, Syed Rafay Hasan
Convolutional Neural Networks (CNN) has found successful adoption in many applications. The deployment of CNN on resource-constrained edge devices have proved challenging. CNN distributed deployment across different edge devices has been adopted. In this paper, we propose Trojan attacks on CNN deployed across a distributed edge network across different nodes. We propose five stealthy attack scenarios for distributed CNN inference. These attacks are divided into trigger and payload circuitry. These attacks are tested on deep learning models (LeNet, AlexNet). The results show how the degree of vulnerability of individual layers and how critical they are to the final classification.
9.7CRNov 2, 2019
A Stealthy Hardware Trojan Exploiting the Architectural Vulnerability of Deep Learning Architectures: Input Interception Attack (IIA)Tolulope A. Odetola, Hawzhin Raoof Mohammed, Syed Rafay Hasan
Deep learning architectures (DLA) have shown impressive performance in computer vision, natural language processing and so on. Many DLA make use of cloud computing to achieve classification due to the high computation and memory requirements. Privacy and latency concerns resulting from cloud computing has inspired the deployment of DLA on embedded hardware accelerators. To achieve short time-to-market and have access to global experts, state-of-the-art techniques of DLA deployment on hardware accelerators are outsourced to untrusted third parties. This outsourcing raises security concerns as hardware Trojans can be inserted into the hardware design of the mapped DLA of the hardware accelerator. We argue that existing hardware Trojan attacks highlighted in literature have no qualitative means how definite they are of the triggering of the Trojan. Also, most inserted Trojans show a obvious spike in the number of hardware resources utilized on the accelerator at the time of triggering the Trojan or when the payload is active. In this paper, we introduce a hardware Trojan attack called Input Interception Attack (IIA). In this attack, we make use of the statistical properties of layer-by-layer output to ensure that asides from being stealthy. Our IIA is able to trigger with some measure of definiteness. Moreover, this IIA attack is tested on DLA used to classify MNIST and Cifar-10 data sets. The attacked design utilizes approximately up to 2% more LUTs respectively compared to the un-compromised designs. Finally, this paper discusses potential defensive mechanisms that could be used to combat such hardware Trojans based attack in hardware accelerators for DLA.
4.2CROct 3, 2018
EPIC: Efficient Privacy-Preserving Scheme with E2E Data Integrity and Authenticity for AMI NetworksAhmad Alsharif, Mahmoud Nabil, Samet Tonyali et al.
In Advanced Metering Infrastructure (AMI) networks, smart meters should send fine-grained power consumption readings to electric utilities to perform real-time monitoring and energy management. However, these readings can leak sensitive information about consumers' activities. Various privacy-preserving schemes for collecting fine-grained readings have been proposed for AMI networks. These schemes aggregate individual readings and send an aggregated reading to the utility, but they extensively use asymmetric-key cryptography which involves large computation/communication overhead. Furthermore, they do not address End-to-End (E2E) data integrity, authenticity, and computing electricity bills based on dynamic prices. In this paper, we propose EPIC, an efficient and privacy-preserving data collection scheme with E2E data integrity verification for AMI networks. Using efficient cryptographic operations, each meter should send a masked reading to the utility such that all the masks are canceled after aggregating all meters' masked readings, and thus the utility can only obtain an aggregated reading to preserve consumers' privacy. The utility can verify the aggregated reading integrity without accessing the individual readings to preserve privacy. It can also identify the attackers and compute electricity bills efficiently by using the fine-grained readings without violating privacy. Furthermore, EPIC can resist collusion attacks in which the utility colludes with a relay node to extract the meters' readings. A formal proof, probabilistic analysis are used to evaluate the security of EPIC, and ns-3 is used to implement EPIC and evaluate the network performance. In addition, we compare EPIC to existing data collection schemes in terms of overhead and security/privacy features.
2.3CRJul 30, 2018
Load Control and Privacy-Preserving Scheme for Data Collection in AMI NetworksHawzhin Mohammed, Syed Rafay Hasan, Mohammad Ashiqur Rahman
In Advanced Metering Infrastructure (AMI) systems, smart meters (SM) send fine-grained power consumption information to the utility company, yet this power consumption information can uncover sensitive information about the consumers' lifestyle. To allow the utility company to gather the power consumption information while safeguarding the consumers' privacy, different methods that broadly utilize symmetric key and asymmetric key cryptography operation have been generally utilized. In this paper, we propose an effective method that uses symmetric key cryptography and hashing operation to gather power consumption information. Moreover, provide the utility company with an overview of the type of the appliances used by its power consumer and range of power use. The idea is based on sending cover power consumption information from the smart meters and removes these covers by including every one of the smart meters' messages, with the goal that the utility can take in the accumulated power consumption information, yet cannot take in the individual readings. Our assessments show that the cryptographic operations required in our scheme are substantially more effective than the operations required in other schemes.