South Asian Journal of Engineering and Technology

Hybrid Encryption and Intelligent Pattern-Matching Architecture for Cloud Security

2026-01-07T09:05:52+00:00

The escalating frequency of data breaches in cloud storage environments has exposed the inadequacies of monolithic security models, where static authentication and opaque server-side encryption fail to protect against credential compromise and insider threats. This work presents Zenith, a secure cloud management platform that unifies multi-layered defense mechanisms across three critical domains: cryptographic access control, hybrid dual-encryption architecture, and automated threat detection. The Security Management Service implements RFC 6238-compliant Time-based One-Time Password (TOTP) authentication using HMAC-SHA1 verification with a flexible 90-second drift tolerance window, achieving verification speeds of 10–20ms while mitigating replay attacks. The encryption framework introduces a hybrid model offering Server-Side Encryption (SSE) using AWS S3 native AES-256 for general performance, and a zero-knowledge Client-Side Encryption (CSE) architecture that derives 256-bit keys using PBKDF2-HMAC-SHA256 with 100,000 iterations and a unique 16-byte salt, ensuring the platform possesses no decryption capability for sensitive data. To proactively prevent data leakage, an Automated Sensitive Data Detection pipeline utilizes regex-based pattern matching to identify Personally Identifiable Information (PII), credit card sequences (13−16 digits), and private IP ranges in real time, automatically triggering mandatory encryption workflows for flagged files. Data durability and disaster recovery are secured through AWS Cross-Region Replication (CRR) spanning approximately 2,800 miles (N. Virginia to Oregon), achieving 99.999999999% (eleven nines) data durability with eventual consistency typically achieved within 15 minutes.Comprehensive session auditing through MongoDB collects the device fingerprint and geolocation information, giving fine-grained visibility to access patterns. Experimental validation looks at that the platform balances rigorous security with operational efficiency, maintaining a client-side encryption overhead of only 200–300ms per megabyte, effectively establishing a compliant, resilient, and transparent foundation for secure cloud data management.

Self-Healing Cybersecurity System for Cloud Environments

2026-01-10T05:47:19+00:00

Cloud computing has become the backbone of today’s digital landscape. It provides high scalability, flexibility, and cost savings. However, its distributed and changing nature makes it susceptible to various cyber threats, such as insider misuse, ransomware, and advanced persistent threats (APTs). This paper introduces a Self-Healing Cybersecurity System (SHCS) for cloud environ- ments that merges intelligent anomaly detection with automatic recovery processes. The framework combines Machine Learning (ML) and Deep Learning (DL) models for real-time threat detection. These models quickly identify harmful activities and system issues in cloud environments. To ensure data integrity, blockchain technology logs information securely and transparently, aiding auditing and regulatory compliance. Experimental results show that this AI-driven method greatly improves system resilience, reduces operational downtime, and builds trust in multi-tenant cloud infrastructures.

Adaptive Sensor Fusion System for Low-Visibility Vehicle Navigation with Hud Point Cloud Rendering

2026-02-18T07:12:03+00:00

This paper describes how I built and tested a patented adaptive sensor-fusion system for autonomous driving in low-visibility conditions. Based on U.S. Patent 12371046B2, the system continuously checks how clear the environment is and automatically switches between camera-based navigation and LiDAR-based navigation when visibility drops. It also projects the LiDAR view (point-cloud information) onto a windshield heads-up display (HUD) so the driver or safety operator can clearly see what’s ahead. To judge visibility, I use a hybrid metric that combines image contrast and edge density, and it identifies conditions accurately (94.5%). For LiDAR perception, the system uses DBSCAN clustering to detect objects from point clouds in real time. In heavy fog, it detects objects reliably up to 50 meters (100% detection rate), while a camera-only approach drops sharply (16.7%). Switching between camera and LiDAR is fast (183 ms), so transitions feel smooth. Across 425 test scenarios, the system greatly increases detection range in heavy fog (316.7% improvement) and keeps end-to-end processing low (73.1 ms), supporting real-time operation at 30 Hz. Overall, it meets and exceeds safety requirements with 99.98% availability, showing the patented approach is practical for real vehicles in bad weather.

ASIL-Aware Circular Logger: A Zero-Interference Diagnostic Logging Framework for Safety- Critical Automotive RTOS

2026-03-14T06:14:59+00:00

High loads frequently lead to throttling of diagnostic logging on mixed-criticality automotive ECUs due to blocking and timing interference created by traditional logger designs. This paper delivers ASIL (Automotive Safety Integrity Level)-Aware Circular Logger (AACL) a FreeRTOS-based framework that incorporates hook-based execution-domain knowledge and ASIL-prioritized buffering to eliminate the possibility of blocking once safety-critical execution paths are defined. A closed-form calibrated deferred-latency upper bound is derived from representative automotive task sets; experimental testing used the FreeRTOS GCC_POSIX port with a Cortex-M4 at 168 MHz reference task model with unique benchmark against a naive logger (directly based on mutexes) will provide evidence hosts but not certified timing (in target systems). Results showed no measured ASIL task blocking for the AACL, subordinate deferred latencies (bounded) with quantifiable tightness bounds, significant throughput/drop-rate robustness game over baseline logger provided, only public FreeRTOS API used in implementation and designed for integration into ISO 26262 timing arguments.