Patents & Publications
Current Research
A central thread in Dr. Martin's current research is software supply-chain security: quantifying, and then reducing, the gap between a software component's nominal version and its true vulnerability exposure — from firmware, through operating-system distributions, to language package managers.
This work grows directly out of Firmware IQ, which identifies the components inside a firmware image and maps them to known CVEs. Dr. Martin's research applies the same lens one layer up, at the operating-system distribution: because enterprise and “stable” Linux distributions maintain soft forks of upstream packages, with incomplete backports and out-of-sync LTS trees, two packages reporting the same upstream version can carry very different real-world exposure.
He quantifies this divergence, and the patch lead time between rolling and enterprise distributions, to ask whether rolling releases are in fact more secure and how distributions might share backporting effort rather than duplicate it — questions that in turn motivate the Linux distribution he is now building.
This research spans several related threads:
Software supply-chain security
- Automated binary version extraction for NVD cross-reference based on fuzzy matching
- Automated analysis of vulnerabilities in containers and virtual appliances
- Large-scale comparison of the nature and kind of firmware vulnerabilities across and within product classes
- Comparison of musl and glibc attack surface normalized by code size and severity
AI systems and applications
- Performance analysis and optimization of local large language model inference, including forecasting of architecture and parameter-scaling trends across frontier, open-weight, and agentic models
- Use of large language models for technical research and analysis
- Exploratory use of multimodal large language models for diagnosis of disease progression
U.S. Patents
US 10,762,214 B1 — System and Method for Automatically Extracting Information from Binary Files for Use in Database Queries
US 10,230,599 B2 — System and Method for Network Traffic Profiling and Visualization
US 9,667,521 B2 — System and Method for Network Traffic Profiling and Visualization
US 8,984,583 B2 — Healthcare Privacy Breach Prevention Through Integrated Audit and Access Control
US 9,438,632 B2 — Healthcare Privacy Breach Prevention Through Integrated Audit and Access Control
Whitepapers
P. Martin. "Reliable Software: Industry Best Practices." Harbor Experts Whitepaper. [PDF]
P. Martin. "Guidelines for Source Code Comparison in Litigation." Harbor Labs Litigation Support Whitepaper, 2022. [PDF]
P. Martin. "Guidelines for Assessing Source Code Quality." Harbor Labs Litigation Support Whitepaper, 2022. [PDF]
P. Martin et al. "Guidelines for Source Code Review in Hi-Tech Litigation." Harbor Labs Litigation Support Whitepaper, 2022. [PDF]
Conference Papers
Published in Proc. ACM/IEEE International Conference on Internet-of-Things Design and Implementation (IoTDI 18), 2018
Sentinel is a hardware-based security system designed to be soldered directly to the CPU of IoT-class embedded devices. It monitors control-flow transitions to build runtime profiles of normal device behavior and enforces those profiles to detect anomalous execution, providing control-flow integrity for resource-constrained embedded systems.
Citation: P. Martin, D. Russel, M. Ben Salem, S. Checkoway, A. Rubin. "Sentinel: Secure Mode Profiling and Enforcement for Embedded Systems." Proc. ACM/IEEE International Conference on Internet-of-Things Design and Implementation (IoTDI 18), 2018.
View paper
Published in Proc. ACM Conference on Bioinformatics, Computational Biology, and Health Informatics (BCB 16), 2016
This paper presents a secure indoor location tracking system using unspoofable Bluetooth Low Energy beacons for healthcare environments. The system enables automatic presentation of relevant patient medical records to physicians as they move through a facility, providing a secondary authentication mechanism that is transparent to the user while strengthening access controls.
Citation: P. Martin, M. Rushanan, T. Tantillo, C. Lehmann, A. Rubin. "Applications of Secure Location Sensing in Healthcare." Proc. ACM Conference on Bioinformatics, Computational Biology, and Health Informatics (BCB 16), 2016.
View paper
Published in Financial Cryptography and Data Security (FC 16), Lecture Notes in Computer Science, vol. 9603, Springer, 2016
KBID is an authentication bracelet that receives a Kerberos ticket upon login to a modified computer terminal through low-energy electrical signals transmitted over the wearer’s skin. The bracelet enables password-free authentication at other terminals throughout a facility and immediately loses the cryptographic secret upon removal from the user.
Citation: J. Carrigan, P. Martin, M. Rushanan. "KBID: Kerberos Bracelet Identification (Short Paper)." Financial Cryptography and Data Security (FC 16), Lecture Notes in Computer Science, vol. 9603, Springer, 2016.
View paper
Published in Proc. ACM Conference on Bioinformatics, Computational Biology, and Biomedical Informatics Health Informatics Symposium (BCB-HIS), 2013
This paper presents a system for enforcing the minimum necessary access principle in healthcare settings by integrating large-scale audit log analysis with access control mechanisms. Using a Hadoop-based application for statistical analysis of electronic medical record audit logs, the system automatically produces human-readable reports and identifies access patterns that may indicate privacy violations. This technology was subsequently patented by Accenture.
Citation: P. Martin, A. Rubin, R. Bhatti. "Enforcing Minimum Necessary Access in Healthcare Through Integrated Audit and Access Control." Proc. ACM Conference on Bioinformatics, Computational Biology, and Biomedical Informatics Health Informatics Symposium (BCB-HIS), September 2013.
View paper
Technical Reports
Published in Technical Report 13-01, Johns Hopkins University, 2013
This technical report presents techniques for fingerprinting and classifying specific versions of network protocol implementations by analyzing observable behavioral differences. The case study focuses on OpenSSL, demonstrating that implementation version information can be inferred from network traffic analysis.
Citation: P. Martin, M. Rushanan, S. Checkoway, M. Green, A. Rubin. "Classifying Network Protocol Implementation Versions: An OpenSSL Case Study." Technical Report 13-01, Johns Hopkins University, December 2013.
View paper
Dissertation
P. Martin, Securing Medical Devices and Protecting Patient Privacy in the Technological Age of Healthcare, Ph.D. Dissertation, Johns Hopkins University, February 2016. [JHU Seminar]
Professional Committees & Standards Bodies
IEEE 7024 Working Group — “Standard for the Procurement, Verification and Validation, and Life Cycle Management of Forensic Technologies” (Current Member)
IEEE Symposium on Security and Privacy — Program Committee Member (Winter 2024, Summer 2024)
