Mohamed R. Elshamy

Mohamed R. Elshamy

Auto-generated profile, based on publicly available information

Mohamed R. Elshamy

Klipsch School of Electrical & Computer Engineering, New Mexico State University

Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
15 July 2022 in Concurrency and Computation: Practice and Experience
Verified
Journal Article
Closed access
MSMohamed R. Shoaib
Mohamed R. Shoaib
MEMohamed R. Elshamy
Mohamed R. Elshamy
TTTaha E. Taha
Taha E. Taha
AEAdel S. El‐Fishawy
Adel S. El‐Fishawy
FEFathi E. Abd El‐Samie
Fathi E. Abd El‐Samie

12 October 2022 in Concurrency and Computation: Practice and Experience
Verified
Journal Article
Closed access
MEMohamed R. Elshamy
Mohamed R. Elshamy
BABelal Abozalam
Belal Abozalam
ASAmged Sayed
Amged Sayed
ENEssam Nabil
Essam Nabil

6 October 2025
Verified
Conference Paper
Open access
MEMehdi Elahi
Mehdi Elahi
MEMohamed R Elshmay
Mohamed R Elshmay
ABAbdel-Hameed Badawy
Abdel-Hameed Badawy
APAhmad Patooghy
Ahmad Patooghy

19 May 2026 in ACM Transactions on Embedded Computing Systems
Verified
Journal Article
Open access
MEMehdi Elahi
Mehdi Elahi
MEMohamed R. Elshamy
Mohamed R. Elshamy
ABAbdel-Hameed Badawy
Abdel-Hameed Badawy
APAhmad Patooghy
Ahmad Patooghy
Dynamic Thermal Management (DTM) systems are critical to the reliable operation of Multiprocessor System-on-Chips (MPSoCs), yet remain vulnerable to sophisticated thermal manipulation attacks. These attacks, executed through hardware trojans or privilege escalation, can compromise the integrity of thermal sensors, causing performance degradation, accelerated aging, and catastrophic hardware failure by disabling thermal throttling mechanisms. Existing countermeasures rely on reactive detection methods and conventional machine learning models that fail to capture the complex physics governing thermal systems, including thermal coupling across cores and power-frequency interdependencies, making them ineffective against multi-stage attacks that exploit DTM decision-making logic. This work presents a novel transformer-based defense framework that leverages self-attention mechanisms to model rich, system-wide feature interactions for detecting adversarial thermal manipulations in real time. The proposed hybrid architecture integrates an adaptive pre-filtering with dynamic thresholding to achieve an 83x throughput improvement (22,798 samples/second versus 274.53 samples/second for transformer-only baseline) and nearly 50% lower GPU utilization, enabling deployment on resource-constrained embedded platforms. Comprehensive on-device validation on the NVIDIA Jetson AGX Orin board demonstrates substantial thermal regulation improvements, reducing average peak temperatures from 103 °C to 98.5 °C while maintaining a model active ratio of only 2.73%. The framework incorporates an adaptive defense system with load-dependent dynamic thresholding that achieves high F1-scores in detecting the thermal attacks discussed in the literature (0.75 to 0.9). This work bridges the critical gap between simulation-based security research and practical embedded system deployment, establishing a new paradigm for lightweight, attention-based anomaly detection in thermally constrained environments.

12 June 2026 in ACM Transactions on Design Automation of Electronic Systems
Verified
Journal Article
Open access
MEMohamed Elshamy
Mohamed Elshamy
MEMehdi Elahi
Mehdi Elahi
APAhmad Patooghy
Ahmad Patooghy
ABAbdel-Hameed Badawy
Abdel-Hameed Badawy
Fine-grained power estimation is fundamental to modern MPSoC design and operation, enabling early power and thermal budgeting before fabrication, predictive thermal control at runtime, post-silicon microarchitectural energy debugging, and hardware security analysis. Existing System-on-Chip (SoC) designs often rely solely on total power consumption measurements because of the cost, size, and complexity of power sensors. However, the lack of precision and granularity in power measurements leads to inefficient power and thermal management and limits the ability to detect failed or compromised sensors. To tackle these challenges, we propose the Improved Clustering Blind Power Identification (ICBPI) approach, an innovative approach for fine-grained power estimation in SoCs. ICBPI leverages total power measurements and per-unit temperature data, even with potentially unreliable sensors, to deliver accurate per-unit power estimates. This dual focus not only enhances power estimation, leading to better power management decisions, but also improves the detection of malicious sensorssuch as those compromised by hardware Trojan attacksthat manipulate sensor data to induce potential system failures. Extensive evaluations across four SoC architectures demonstrate ICBPI’s superior performance in reducing per-core power estimation errors by 77.56% compared to the baseline Blind Power Identification (BPI) and by 68.44% compared to the state-of-the-art approaches. Furthermore, ICBPI achieves 100% detection of malicious thermal sensor attacks and enhances the localization process compared to the Blind Identification Countermeasure (BIC) approach. In the comparative analysis between ICBPI and the most recent model, Alternating Blind Power Identification (ABPI), ICBPI shows error rates as low as 0.5% compared to 2.8% in a 2 × 2 mesh benchmark and outperforms ABPI on heterogeneous architectures. The NVIDIA Jetson Xavier AGX development board serves as a platform for validating ICBPI. Consequently, ICBPI demonstrates an approximate reduction in average error of 79.71% on the CPU and 79.59% on the GPU relative to three existing state-of-the-art approaches.