Haifa Alanazi

Dr. Haifa Alanazi

orcid

Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

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

Representative Publications

View All Publications
1 January 2025 in International Journal of Advanced Computer Science and Applications
Verified
Journal Article
Open access
PDF file
HAHaifa Alanazi
Haifa Alanazi
Instance segmentation is a critical component of medical image analysis, enabling tasks such as tissue and organ delineation, and disease detection. This paper provides a detailed comparative analysis of two fine-tuned one-stage object detection models, YOLOv11-seg and YOLOv9-seg, tailored for instance segmentation in medical imaging. Leveraging transfer learning, both models were initialized with pretrained weights and sub-sequently fine-tuned on the NuInsSeg dataset, which comprises over 30,000 manually segmented nuclei across 665 image patches from various human and mouse organs. This approach facilitated faster convergence and improved generalization, particularly given the limited size and high complexity of the medical dataset. The models were evaluated against key performance metrics. The experimental results reveal that YOLOv11n-seg outperforms YOLOv9c-seg with a precision of 0.87, recall of 0.84, and mAP50 of 0.89, indicating superior segmentation quality and more accurate delineation of nuclei contours. This study highlights the robust performance and efficiency of YOLOv11n-seg, demonstrating its superiority in medical image segmentation tasks, with notable advantages in both accuracy and real-time processing capabilities.

23 March 2026 in Sensors
Verified
Journal Article
Open access
PDF file
AAAbdulaziz Ghabash Alanazi
Abdulaziz Ghabash Alanazi
HAHaifa Alanazi
Haifa Alanazi
Edge computing is now widely used to support real-time and safety-critical IoT services. However, current edge schedulers usually optimize only performance, while security verification and trust assessment are handled as separate modules. This separation creates a practical risk: tasks may be assigned to lightly loaded but compromised edge nodes, or secure nodes may become overloaded, violating latency requirements. We propose EdgeGuard-AI, a unified trust-driven and load-aware scheduling framework inspired by zero-trust security principles for next-generation IoT edge networks. The framework jointly learns dynamic node trust and short-term workload patterns from distributed edge data and integrates both signals into scheduling decisions. Experimental results on a realistic IoT edge security dataset show a task success rate of 97.3 percent, average scheduling latency of 58.1 ms during stress periods, unsafe offloading below 2 percent, and trust discrimination AUC of 0.971.

9 September 2026 in Sensors
Verified
Journal Article
Open access
PDF file
HAHaifa Alanazi
Haifa Alanazi
KAKhaznah Alshammari
Khaznah Alshammari
TinyML models now run inference directly on microcontrollers with only kilobytes of memory, inside sensors, wearables, and industrial devices that often have no cloud link and no way to receive a patch once deployed. This makes them an easy target, since compressed models resist adversarial and extraction attacks less than full-precision models, but most TinyML research still treats security as a step added after compression rather than during it. This paper proposes a three-layer security-by-design architecture built to close that gap. The first layer scores candidate compressed models on a joint fitness function that combines clean accuracy, adversarial robustness, and model size before the final architecture is fixed. The second layer signs and verifies model weights with the lightweight ASCON cipher at boot and during over-the-air updates. The third layer is a runtime monitor that computes a live distribution-drift score on incoming queries and raises local defense strength once that score crosses a fixed threshold, without needing a cloud round trip. The full pipeline was validated on real microcontroller hardware, not simulation, and stays inside a 62 KB RAM budget with inference latency under 10 ms. Against real baseline models on Edge-IIoTset, the proposed method reaches 94.3% accuracy, the highest of the models tested, while cutting attack success rate from over 60–84% down to 38.9% and raising the query cost for model extraction more than five times over the weakest baseline. An ablation study and a five-seed statistical test confirm the gain holds and comes from the combination of all three layers, not from any single part.

3 July 2026 in Network
Verified
Journal Article
Open access
PDF file
AAAbdulaziz Ghabash Alanazi
Abdulaziz Ghabash Alanazi
HAHaifa Alanazi
Haifa Alanazi
NANasser Salem Albalawi
Nasser Salem Albalawi
IoT networks now handle traffic from billions of devices, and edge nodes are under constant pressure to classify that traffic and dispatch tasks within tight latency deadlines. Most existing systems treat classification and scheduling as two separate steps that run one after the other. This sequence adds unnecessary delay and breaks the feedback between the two tasks: the scheduler never sees the traffic type, and the classifier never sees the queue state. We propose ParallelEdge-AI, a system built around a shared flow encoder that feeds two task-specific heads in parallel, one for multi-class traffic classification and one for task-urgency scoring. Both heads are trained end-to-end using a joint loss that combines cross-entropy and pairwise ranking. A load-balance controller then reads the urgency scores alongside live queue lengths to decide, every 200 ms, whether a task stays local or moves to a less-loaded edge node. No global synchronisation is needed. We test the system on three real IoT datasets: RT-IoT2022, N-BaIoT, and CICIoT2023. ParallelEdge-AI reaches 97.63% accuracy and an F1-score of 97.34%, which is 3.16 percentage points above the best baseline. Inference latency is 19.62 ms per batch, the deadline-miss rate is 2.34%, and the load-imbalance index is 0.083, all three are the best results in our comparison. These numbers show that running classification and scheduling together on a shared representation is both faster and more accurate than treating them as separate problems.

3 April 2026 in Computation
Verified
Journal Article
Open access
PDF file
HAHaifa Alanazi
Haifa Alanazi
AAAbdulaziz Ghabash Alanazi
Abdulaziz Ghabash Alanazi
NANasser Salem Albalawi
Nasser Salem Albalawi
As the scale and complexity of Internet of Things (IoT) deployments increase, maintaining resilience in resource-constrained mesh networks becomes a significant challenge. Frequent node failures due to battery depletion, environmental interference, or hardware degradation can disrupt data flows and lead to operational downtime. To address this, we propose EdgeRescue, a novel lightweight AI-driven framework for self-healing in energy-constrained IoT mesh environments. EdgeRescue enables each node to perform local anomaly detection using compact 1D Convolutional Neural Networks (1D-CNNs) and initiates distributed, energy-aware routing reconfiguration when faults are detected. Unlike cloud-dependent methods, EdgeRescue operates entirely at the edge, requiring minimal computation, memory, and communication overhead. Extensive simulations on a 100-node testbed demonstrate that EdgeRescue improves packet delivery by 13.2%, reduces recovery latency by 57%, and lowers average node energy consumption by 18.8% compared to state-of-the-art baselines. These results establish EdgeRescue as a scalable and practical solution for achieving real-time resilience in next-generation IoT mesh networks.

3 July 2026 in Electronics
Verified
Journal Article
Open access
PDF file
AAAbdulaziz Ghabash Alanazi
Abdulaziz Ghabash Alanazi
HAHaifa Alanazi
Haifa Alanazi
NANasser Salem Albalawi
Nasser Salem Albalawi
Running neural networks on battery-powered Internet of Things (IoT) sensor nodes is difficult because flash memory, SRAM, latency, and energy per inference are limited at the same time. Existing TinyML co-design methods usually improve model size or memory use, but runtime voltage–frequency control is often handled as a separate step. This separation limits energy saving because the power policy does not use the layer-wise compute profile of the final compressed model. We propose the Cross-Layer Resource Optimizer (CLRO), a three-stage resource optimization pipeline for TinyML inference on an ARM Cortex-M7 target. The first stage, Mixed-Precision Aware Pruning and Distillation (MPAD), assigns per-layer bit widths and pruning ratios using calibration-set sensitivity scores. The second stage, consisting of the Activation Lifetime-Aware Tensor Scheduler (ALTS), uses the compressed graph to find an execution order that reduces peak live static random-access memory (SRAM). The third stage, Reinforcement Learning-Based Dynamic Voltage and Frequency Scaling (DVFS-RL), trains a tabular Q-learning policy from the multiply–accumulate (MAC) utilization profile of the compressed and scheduled model. The learned voltage–frequency policy is stored as a small flash lookup table, so it adds no runtime decision cost during inference. We evaluate the CLRO on all four MLPerf Tiny tasks using an STM32H743ZI microcontroller with 512 kB SRAM and 2 MB flash. The CLRO reaches 91.7% image classification accuracy, 95.4% keyword-spotting accuracy, 89.6% visual wake words accuracy, and 0.913 anomaly detection AUC. The final deployment uses 198 kB flash and 174 kB peak SRAM, with 387 μJ energy per inference and 38 ms latency. Compared with the MCUNet baseline, the CLRO reduces energy by 58.1% and peak SRAM by 39% while keeping the same accuracy level.