机构:
Xian Jiaotong Univ, Comp & Informat Technol Inst, Xian 710049, Peoples R ChinaXian Jiaotong Univ, Comp & Informat Technol Inst, Xian 710049, Peoples R China
Cheng, X
[1
]
Hou, YB
论文数: 0引用数: 0
h-index: 0
机构:
Xian Jiaotong Univ, Comp & Informat Technol Inst, Xian 710049, Peoples R ChinaXian Jiaotong Univ, Comp & Informat Technol Inst, Xian 710049, Peoples R China
Hou, YB
[1
]
机构:
[1] Xian Jiaotong Univ, Comp & Informat Technol Inst, Xian 710049, Peoples R China
来源:
2003 INTERNATIONAL CONFERENCE ON MACHINE LEARNING AND CYBERNETICS, VOLS 1-5, PROCEEDINGS
|
2003年
关键词:
source routing;
genetic algorithm;
AntNet;
D O I:
10.1109/ICMLC.2003.1259839
中图分类号:
TP18 [人工智能理论];
学科分类号:
081104 ;
0812 ;
0835 ;
1405 ;
摘要:
In order to improve the routing algorithm AntNet[2], a new algorithm GARA (Genetic Ant Routing Algorithm) is proposed in this paper. A path base is maintained in each node for each destination. Each path bases is evolved by genetic algorithm. Ants are used to explore and evaluate new paths. Based on the path base, GARA is capable of providing 2 new source routing modes suitable for different network traffics. Simulation results show that these new routing modes outperform hop-by-hop routing modes (like AntNet) in terms of throughput and delay, and converge more rapidly.