研究目的
To extend particle swarm optimization (PSO) from the continuous domain to the binary or discrete domain for solving the nondeterministic polynomial (NP) complete multicast routing problem (MRP) in communication networks.
研究成果
BVDPSO effectively extends PSO to the binary domain for MRP optimization, demonstrating superior solution accuracy and faster convergence speed compared to traditional heuristics and other EC methods. It provides a practical approach for multicast design in communication networks.
研究不足
The study focuses on the Steiner tree problem (STP) as an MRP without QoS constraints. Future work could explore BVDPSO's performance with practical QoS constraints and multiple objectives.
1：Experimental Design and Method Selection:
The study proposes a novel bi-velocity discrete particle swarm optimization (BVDPSO) approach for MRP. It includes the development of a bi-velocity strategy to represent binary characteristics and updates velocity and position based on the original PSO learning mechanism.
2：Sample Selection and Data Sources:
Experiments are conducted on 58 instances from the Operation Research Library (OR-library), categorized into small, medium, and large scales.
3：List of Experimental Equipment and Materials:
The experiments are implemented in VC++ 6.0 and run on a PC with Pentium IV 2.8-GHz CPU and 256-MB memory.
4：0 and run on a PC with Pentium IV 8-GHz CPU and 256-MB memory.
Experimental Procedures and Operational Workflow:
4. Experimental Procedures and Operational Workflow: BVDPSO is tested against several state-of-the-art and recent heuristic algorithms, genetic algorithms, ant colony optimization, and PSO-based methods. Performance is evaluated based on solution accuracy and convergence speed.
5：Data Analysis Methods:
Solution accuracy is measured by relative error (R%), and convergence speed is evaluated by the mean number of fitness evaluations (FEs) to obtain the optimal solution.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
