Machine Learning Vulnerabilities in IoT-Driven Human Activity Recognition Systems
Human Activity Recognition (HAR) is a problem of interpreting sensor data to human movement using an efficient machine learning (ML) approach. The HAR systems rely on data from untrusted users, making them susceptible to data poisoning attacks. In a poisoning attack, attackers manipulate the sensor readings to contaminate the training set, misleading the HAR to produce erroneous outcomes. This research aims to investigate the vulnerabilities of HAR-ML systems, design and develop efficient and effective real-time lightweight solutions.
Abdur R. Shahid, Ahmed Imteaj, Peter Y. Wu, Diane A. Igoche, Tauhidul Alam, "Label Flipping Data Poisoning Attack Against Wearable Human Activity Recognition System". In 2022 IEEE Symposium Series on Computational Intelligence (SSCI), 2022.
Differentially-Private Machine Learning for Healthcare
The field of Machine Learning (ML) has been engaged in intensive research for a while to build an efficient and effective intelligent system for the early identification of chronic diseases such as cancer and diabetes and has recently seen some promising findings. The bulk of the initiatives are aimed at classifying illness onset and minimizing cases of maltreatment. As a supervised learning problem, its accuracy is mostly determined by the training data, which is labeled data on actual patients that is highly privacy-sensitive. Privacy leakage can occur at any point in the machine learning lifecycle, from model training through model deployment, and can lead to a membership inference attack, model inversion attack, and reconstruction attack. As a result, safeguarding users' privacy is critical in healthcare issues, but little has been done to satisfy this demand. This collaborative project aims to develop practical differentially private ML models for various healthcare problems.
Abdur R. Shahid, Sajedul Talukder, and Tasnimun Faika, "Defending Against Personality Prediction Attack in Handwriting Recognition-based Systems Using a Local Differentially Private Machine Learning (ML) Framework". In Journal of Cybersecurity and Privacy (JCP) (In submission).
Abdur R. Shahid, Sajedul Talukder, "A Study of Differentially Private Machine Learning in Healthcare". In 2021 Innovations in Intelligent Systems and Applications Conference (ASYU). IEEE, 2021.
Abdur R. Shahid, Sajedul Talukder, "Evaluating Machine Learning Models for Handwriting Recognition-based Systems under Local Differential Privacy". In 2021 Innovations in Intelligent Systems and Applications Conference (ASYU). IEEE, 2021.
Melinda Goda, Abdur R. Shahid, "Differentially Private Machine Learning for Breast Cancer Classification", Consortium for Computing Sciences in Colleges — Northeastern Region, 2021 (Runner Up in Best Undergraduate Poster Award)
Blockchain Framework for Resource-Constrained Mobile Internet of Things (IoT)
This research aimed to address the blockchain management problem by designing a lightweight blockchain framework, for mobility-centric IoT without relying on a fixed infrastructure of edge devices. We show that breaking down a traditional global blockchain into smaller local blockchains in the spatial domain and limiting their size through a temporal constraint will allow us to design scalable blockchain for mobile IoT systems.
Hussein Zangoti, Alex P. Makki, Niki Pissinou, Abdur R. Shahid, Omar J. Guerra, Joel Rodriguez, "A Multidimensional Blockchain Framework For Mobile Internet of Things". In 2022 International Conferences on Trust, Security and Privacy in Computing and Communications (TrustCom), IEEE.
T Alam, J Talylor, S Badsha, Abdur R. Shahid, A.S.M Kayes, "Leveraging Blockchain for Spoof-Resilient Robot Networks". In International Conference on Intelligent Robotics and Applications 2020 Nov 5 (pp. 207-216). Springer.
Abdur R. Shahid, Niki Pissinou, Corey Staier, and Rain Kwan. "Sensor-Chain: A Lightweight Scalable Blockchain Framework for Internet of Things". In 2019 International Conference on Internet of Things (iThings) and IEEE Green Computing and Communications (GreenCom) and IEEE Cyber, Physical and Social Computing (CPSCom) and IEEE Smart Data (SmartData) (pp. 1154-1161). IEEE.
Abdur R. Shahid, Niki Pissinou, laurent Njilla, Edwin Aguilar, and Eric Perez. "Demo: Towards the Development of a Differentially Private Lightweight and Scalable Blockchain for IoT." In 16th IEEE International Conference on Mobile Ad-Hoc and Smart Systems (MASS), IEEE.
Decentralized Location Privacy in Smart City Blockchain
This research aimed to address the blockchain management problem by designing a lightweight blockchain framework, coined as Sensor-Chain, for mobility-centric IoT without relying on a fixed infrastructure of edge devices. We show that breaking down a traditional global blockchain into smaller local blockchains in the spatial domain and limiting their size through a temporal constraint will allow us to design scalable blockchain for mobile IoT systems.
Abdur R. Shahid, Niki Pissinou, and Sajedul Talukder, "Protecting Location Privacy in Blockchain-Based Mobile Internet of Things". In Principles and Practice of Blockchains, DOI: https://doi.org/10.1007/978-3-031-10507-4_5
Abdur R. Shahid, Niki Pissinou, Sheila Alemany, Ahmed Imteaj, Kia Makki, and Laurent Njilla. "Quantifying Location Privacy in Permissioned Blockchain-based Internet of Things (IoT)". In 2019 EAI International Conference on Mobile and Ubiquitous Systems: Computing, Networking and Services (MobiQuitous).
Privacy Preserving Mechanism for Continuous Location Sharing In Centralized Internet Of Things (IOT) Systems
In this project, we propose a novel location privacy-preserving approach, called KLAP, based on three fundamental obfuscation requirements: minimum k-locations, l-diversity, and privacy area preservation. KLAP models a user's preference for different locations based on the historical data available on the LBS and can attain personalized privacy-preservation by utilizing such a model for sporadic, frequent, and continuous LBS use cases.
Abdur R. Shahid, Niki Pissinou, S.S. Iyengar, and Kia Makki, "Delay‐aware privacy‐preserving location‐based services under spatiotemporal constraints." International Journal of Communication Systems 34, no. 1 (2021): e4656.
Zeng, Wei, Abdur R. Shahid, Keyan Zolfaghari, Aditya Shetty, Niki Pissinou, and Sitharama S. Iyengar. "n-VDD: Location Privacy Protection Based on Voronoi-Delaunay Duality." arXiv preprint arXiv:1906.09158 (2019).
Abdur R. Shahid, Niki Pissinou, S. S. Iyengar, Jerry Miller, Ziqian Ding, and Teresita Lemus. "KLAP for Real-World Protection of Location Privacy." In 2018 IEEE World Congress on Services (SERVICES), pp. 17-18. IEEE, 2018.
Abdur R. Shahid, Liz Jeukeng, Wei Zeng, Niki Pissinou, S. S. Iyengar, Sartaj Sahni, and Maite Varela-Conover. "PPVC: Privacy-Preserving Voronoi Cell for location-based services." In 2017 International Conference on Computing, Networking and Communications (ICNC), pp. 351-355. IEEE, 2017.
Usable Location Privacy for Location-Based Social Networks
In this project, we study the impact of a combination of spatiotemporally correlated geo-tagged and non-geo-tagged contents on existing trajectory obfuscation mechanisms. To do so, we propose a content oblivious inference attack model from the perspective of a malicious authority who has access to historical data of all the users. we designed a probabilistic inference model by considering both checkin and photo probabilities for each location. With Foursquare's New York City and Tokyo datasets, we first implement the inference model and apply it on three variations of dummy-based obfuscation mechanisms, and show that a straightforward application of existing dummy approaches can leak location privacy in the presence of heterogeneous data (in our case, it is geo-tagged and non-geo-tagged photos and checkins). To the best of our knowledge, this is the first work to investigate the impact of historical shared photos on location privacy.
Abdur R. Shahid, Niki Pissinou, S. S. Iyengar, and Kia Makki. "Check-ins and Photos: Spatiotemporal Correlation-Based Location Inference Attack and Defense in Location-Based Social Networks." In 2018 17th IEEE International Conference On Trust, Security And Privacy In Computing And Communications/12th IEEE International Conference On Big Data Science And Engineering (TrustCom/BigDataSE), pp. 1852-1857. IEEE, 2018.