International Journal on Soft Computing
Vishnuvardhan Mannava, T. Ramesh · 2026 · 128 citationsRead the paper
Current autonomic computing systems are ad hoc solutions that are designed and implemented from the scratch. When designing software, in most cases two or more patterns are to be composed to solve a bigger problem. A composite design patterns shows a synergy that makes the composition more than just the sum of its parts which leads to ready-made software architectures. As far as we know, there are no studies on composition of design patterns for autonomic computing domain. In this paper we propose pattern-oriented software architecture for self-optimization in autonomic computing system using design patterns composition and multi objective evolutionary algorithms that software designers and/or programmers can exploit to drive their work. Main objective of the system is to reduce the load in the server by distributing the population to clients. We used Case Based Reasoning, Database Access, and Master Slave design patterns. We evaluate the effectiveness of our architecture with and without design patterns compositions. The use of composite design patterns in the architecture and quantitative measurements are presented. A simple UML class diagram is used to describe the architecture.
What the full text shows
Working prototype- Reduction in CPU processing time for the self-optimization module. — 12.456 sec vs. 16.654 sec without patterns (Self-optimization module using aspectual Design Patterns)
- Reduction in total compilation time for the self-optimization module. — 2532 sec vs. 3587 sec without patterns (Self-optimization module using Design Patterns)
- Reduction in total started thread count. — 14 threads vs. 15 threads without patterns (Self-optimization module using Design Patterns)
- Increase in heap memory usage. — 62,345 kbytes vs. 15,691 kbytes without patterns (Self-optimization module using aspectual Design Patterns)
- Setup
- UML class diagram based architecture evaluated by comparing a self-optimization module with and without design pattern compositions.
- How
- The authors proposed a pattern-oriented software architecture for self-optimization in autonomic computing using a composition of Case Based Reasoning, Database Access, and Master Slave design patterns. They integrated Multi-Objective Evolutionary Algorithms (Genetic Algorithms) to distribute the population from a server to multiple clients to reduce server load.
- Limitations
- The paper does not specify the hardware specifications or the specific multi-objective problem used for the GA simulation. · The reported heap memory usage increased significantly when using patterns, contradicting the author's claim that it 'takes less heap memory'.
Summarized by Gemma 4 from the open-access full text · 10 Oct 2026. Read it