Journal of Network Security Computer Networks

Performance Evaluation of Conflict Resolution Techniques in Edge–Cloud Database Synchronization

2026-07-07T05:18:17+00:00

Edge–cloud database systems face significant challenges in maintaining data consistency while providing low-latency access for edge applications. Concurrent updates at distributed edge nodes often cause conflicts, degrading system performance and reliability. This study presents a systematic empirical evaluation of three widely used conflict resolution strategies: Last-Write-Wins (LWW), version vectors, and Conflict-Free Replicated Data Types (CRDTs), across multiple NoSQL databases (Cassandra, Redis, and CouchDB) under high-concurrency workloads. Experiments assess synchronization latency, throughput, conflict rate, and metadata overhead under varying network conditions and numbers of edge nodes. Results indicate that LWW achieves the lowest latency and highest throughput but suffers from high conflict rates; CRDTs eliminate conflicts at the cost of higher latency and metadata overhead, and version vectors provide a balanced trade-off. These findings provide quantitative guidance for selecting conflict resolution strategies in edge–cloud systems, highlighting the trade-offs between performance, consistency, and scalability. The study also contributes a reproducible benchmark framework for evaluating synchronization strategies under realistic edge–cloud scenarios.

Smart Safe Navigation Using Machine Learning and Network-Based Location

2026-07-01T11:26:33+00:00

The increasing number of road accidents, unsafe travel routes, and delayed emergency responses has highlighted the need for intelligent safety-oriented navigation systems. This paper presents Smart Safe Navigation Using Machine Learning and Network-Based Location, an AI-powered platform designed to enhance public safety through intelligent route guidance, real-time monitoring, and emergency assistance. The proposed system integrates Machine Learning techniques, GPS-based location tracking, accident datasets, traffic analysis, geofencing, and location-based services to identify accident-prone areas and recommend safer travel routes. The system incorporates several advanced features, including live GPS tracking, crime and accident heatmaps, safe route recommendations, geofence-based safety alerts, SOS emergency services, emergency contact management, incident reporting, and an AI chatbot for user assistance. Historical Maharashtra Road accident data from 2022–2024 is utilized to analyze accident trends, identify dangerous zones, and perform risk prediction. An interactive analytics dashboard is also developed to visualize crashes, fatalities, black spots, peak accident timings, and district-wise accident statistics through graphs and heatmaps. By combining machine learning with network-based location technologies, the proposed system provides users with real-time safety information and emergency support. The platform aims to reduce accident risks, improve situational awareness, enhance emergency response efficiency, and promote safer travel experiences. The results demonstrate the effectiveness of intelligent navigation systems in improving road safety and supporting smart city transportation initiatives.

Role of CIA in MD5 to Optimize the Risk

2026-07-02T12:10:10+00:00

The MD5 (Message-Digest 5) hashing algorithm has been widely used for data integrity verification because of its high computational speed and low processing overhead. However, significant cryptographic weaknesses, particularly collision attacks, have rendered MD5 unsuitable for modern security-critical applications. This study evaluates MD5 using the CIA Triad framework—Confidentiality, Integrity, and Availability—and examines its impact on organizational risk optimization within the context of the NIST Risk Management Framework (RMF). The analysis shows that MD5 fails to ensure data integrity because attackers can generate different inputs that produce the same hash value, enabling data forgery and manipulation. From a confidentiality perspective, MD5’s rapid computation makes it vulnerable to brute-force and dictionary attacks when used for password storage. Additionally, continued reliance on MD5 can negatively affect system availability due to compliance restrictions, software incompatibilities, and the rejection of weak cryptographic standards in modern security infrastructures. Despite these limitations, MD5 retains limited utility in low-risk, high-performance environments where cryptographic security is not the primary objective, such as file checksums, duplicate detection, and error verification. To optimize security risk, this study recommends a tiered cryptographic approach in which integrity-critical applications migrate to stronger hashing algorithms such as SHA-256 or SHA-3, while password protection employs adaptive hashing techniques such as bcrypt or Argon2. The findings conclude that MD5 should be deprecated for security-sensitive functions but may remain appropriate for non-adversarial operational tasks. This balanced approach enhances confidentiality, integrity, and availability while maintaining performance and supporting regulatory compliance in contemporary cybersecurity environments.

An Automated IoT Network Threat Detection and Attack Surface Analysis System

2026-05-16T06:21:38+00:00

The issue of network security is rising substantially, in connection with the development of the Internet of Things (IoT) innovations. The most installed one with loose settings is those which have surveillance cameras, routers, and smart home systems, and can be penetrated and were left unrestrained. Although they can represent devices and open ports, traditional network scanning programs like Nmap and Netdiscover will provide raw data, which would need to be interpreted manually and therefore would be time-consuming and prone to error.

To enhance the network security analysis, the proposed research will introduce an automated automation, which will involve smart analysis and scanning networks. There will be the automatic scanning of the WiFi network connected to the system when any change is detected in the network. The scan obtained is utilized in the process of locating the actively used devices and ports that are open. Intelligent Decryption of the results is then performed with the help of a parameter of analysis to find the degree of risk in the network.

The system has a capability to automatically disconnect networks that are considered dangerous and give real-time notification in accordance with risk evaluation. This enhances increased response rate and less human intervention. The proposed strategy can be implemented since the approach converts quite complicated technical data into usable data. The results indicate that the implementation of automation and smart analysis can be used significantly to enhance network security surveillance in IoT environments.