DOI: https://doi.org/10.36719/2789-6919/56/222-227
Shahla Suleymanli
Azerbaijan State Oil and Industry University
Master's student
https://orcid.org/0009-0003-8773-5308
feyzullayeva.shahla.matlab.2024@asoiu.edu.az
A Comparative Study of Implementation Methods for Statistical Memory Analysis Algorithms in JME-Based Systems
Abstract
Efficient memory management is a key factor determining performance and reliability in resource-constrained software environments. Java Micro Edition (JME) is widely used in mobile and embedded systems where memory resources are inherently limited. Therefore, statistical modeling of memory behavior and comparative evaluation of algorithm implementation methods represent an important research problem.
This research systematizes statistical memory analysis algorithms used in JME-based systems, analyses their mathematical foundations, and compares their implementation approaches. Probability distribution models, time-series analysis, adaptive monitoring, and predictive algorithms are evaluated from both theoretical and experimental perspectives. Implementation methods are compared based on computational complexity, memory overhead, real-time responsiveness, prediction accuracy, and system stability.
The comparative analysis demonstrates that hybrid statistical monitoring and adaptive prediction provide the most efficient memory optimization. However, simplified probabilistic models remain more practical and preferable in scenarios with extremely strict resource constraints, where minimal overhead and predictable execution behavior are prioritized over advanced predictive accuracy.
Keywords: JME, memory management, statistical analysis, performance optimization, modeling, embedded systems