An official website of the United States government
Smart-home technology is now pervasive, demanding increased attention to the security of the devices and the privacy of the home's residents. To assist residents in making security and privacy decisions - e.g., whether to allow a new device to connect to the network, or whether to be alarmed when an unknown device is discovered - it helps to know whether the device is inside the home, or outside. In this paper we present MOAT, a system that leverages Wi-Fi sniffers to analyze the physical properties of a device's wireless transmissions to infer whether that device is located inside or outside of a home. MOAT can adaptively self-update to accommodate changes in the home indoor environment to ensure robust long-term performance. Notably, MOAT does not require prior knowledge of the home's layout or cooperation from target devices, and is easy to install and configure. We evaluated MOAT in four different homes with 21 diverse commercial smart devices and achieved an overall balanced accuracy rate of up to 95.6%. Our novel periodic adaptation technique allowed our approach to maintain high accuracy even after rearranging furniture in the home. MOAT is a practical and efficient first step for monitoring and managing devices in a smart home.
more »
« less
SPLICEcube Architecture: An Extensible Wi-Fi Monitoring Architecture for Smart-Home Networks
The vision of smart homes is rapidly becoming a reality, as the Internet of Things and other smart devices are deployed widely. Although smart devices offer convenience, they also create a significant management problem for home residents. With a large number and variety of devices in the home, residents may find it difficult to monitor, or even locate, devices. A central controller that brings all the home’s smart devices under secure management and a unified interface would help homeowners and residents track and manage their devices. We envision a solution called the SPLICEcube whose goal is to detect smart devices, locate them in three dimensions within the home, securely monitor their network traffic, and keep an inventory of devices and important device information throughout the device’s lifecycle. The SPLICEcube system consists of the following components: 1) a main cube, which is a centralized hub that incorporates and expands on the functionality of the home router, 2) a database that holds network data, and 3) a set of support cubelets that can be used to extend the range of the network and assist in gathering network data. To deliver this vision of identifying, securing, and managing smart devices, we introduce an architecture that facilitates intelligent research applications (such as network anomaly detection, intrusion detection, device localization, and device firmware updates) to be integrated into the SPLICEcube. In this thesis, we design a general-purpose Wi-Fi architecture that underpins the SPLICEcube. The architecture specifically showcases the functionality of the cubelets (Wi-Fi frame detection, Wi-Fi frame parsing, and transmission to cube), the functionality of the cube (routing, reception from cubelets, information storage, data disposal, and research application integration), and the functionality of the database (network data storage). We build and evaluate a prototype implementation to demonstrate our approach is scalable to accommodate new devices and extensible to support different applications. Specifically, we demonstrate a successful proof-of-concept use of the SPLICEcube architecture by integrating a security research application: an "Inside-Outside detection" system that classifies an observed Wi-Fi device as being inside or outside the home.
more »
« less
- Award ID(s):
- 1955805
- PAR ID:
- 10343242
- Date Published:
- Journal Name:
- Dartmouth Digital Commons
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Project Connected Home over IP, known as Matter, a unifying standard for the smart home, will begin formal device certification in late 2022. The standard will prioritize connectivity using short-range wireless communication protocols such as Wi-Fi, Thread, and Ethernet. The standard will also include emerging technologies such as Blockchain for device certification and security. In this paper, we rely on the Matter protocol to solve the long-standing heterogeneity problem in smart homes. This work presents a hardware Testbed built using development kits, as there is currently very few devices supporting Matter protocol. In addition, it presents a network architecture that automates smart homes to cloud services. The work is a simple and cheap way of developing a Testbed for automating smart homes that uses Matter protocol. The architecture lays the foundation for exploring security and privacy issues, data collection analysis, and data provenance in a smart home ecosystem built on Matter protocol.more » « less
-
Smart-home devices have become integral to daily routines, but their onboarding procedures - setting up a newly acquired smart device into operational mode - remain understudied. The heterogeneity of smart-home devices and their onboarding procedure can easily overwhelm users when they scale up their smart-home system. While Matter, the new IoT standard, aims to unify the smart-home ecosystem, it is still evolving, resulting in mixed compliance among devices. In this paper, we study the complexity of device onboarding from users' perspectives. We thus performed cognitive walkthroughs on 12 commercially available smart-home devices, documenting the commonality and distinctions of the onboarding process across these devices. We found that onboarding smart home devices can often be tedious and confusing. Users must devote significant time to creating an account, searching for the target device, and providing Wi-Fi credentials for each device they install. Matter-compatible devices are supposedly easier to manage, as they can be registered through one single hub independent of the vendor. Unfortunately, we found such a statement is not always true. Some devices still need their own companion apps and accounts to fully function. Based on our observations, we give recommendations about how to support a more user-friendly onboarding process.more » « less
-
Tracking subjects in videos is one of the most widely used functions in camera-based IoT applications such as security surveillance, smart city traffic safety enhancement, vehicle to pedestrian communication and so on. In computer vision domain, tracking is usually achieved by first detecting subjects, then associating detected bounding boxes across video frames. Typically, frames are transmitted to a remote site for processing, incurring high latency and network costs. To address this, we propose ViFiT, a transformerbased model that reconstructs vision bounding box trajectories from phone data (IMU and Fine Time Measurements). It leverages a transformer’s ability of better modeling long-term time series data. ViFiT is evaluated on Vi-Fi Dataset, a large-scale multimodal dataset in 5 diverse real world scenes, including indoor and outdoor environments. Results demonstrate that ViFiT outperforms the state-of-the-art approach for cross-modal reconstruction in LSTM Encoder-Decoder architecture X-Translator and achieves a high frame reduction rate as 97.76% with IMU and Wi-Fi data.more » « less
-
New capabilities in wireless network security have been enabled by deep learning that leverages and exploits signal patterns and characteristics in Radio Frequency (RF) data captured by radio receivers to identify and authenticate radio transmitters. Open-set detection is an area of deep learning that aims to identify RF data samples captured from new devices during deployment (aka inference) that were not part of the training set; i.e. devices that were unseen during training. Past work in open-set detection has mostly been applied to independent and identically distributed data such as images. In contrast, RF signal data present a unique set of challenges as the data forms a time series with non-linear time dependencies among the samples. In this paper, we introduce a novel open-set detection approach for RF data-driven device identification that extracts its neural network features from patterns of the hidden state values within a Convolutional Neural Network Long Short-Term Memory (CNN+LSTM) model. Experimental results obtained using real datasets collected from 15 IoT devices, each enabled with LoRa, wireless-Wi-Fi, and wired-Wi-Fi communication protocols, show that our new approach greatly improves the area under the precision-recall curve, and hence, can be used successfully to monitor and control unauthorized network access of wireless devices.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- The DOI is not currently available.
