Global Hybrid Routing for Scale-Free Networks
Xiong Gao,Hongxiang Guo,Yanhu Chen,Yinan Tang,Cen Wang,Shengyao Xu,Jian Wu
DOI: https://doi.org/10.1109/access.2019.2897255
IF: 3.9
2019-01-01
IEEE Access
Abstract:The performance of routing strategies on complex networks can be characterized by two measurements, i.e., the traffic capacity and the average packets travel time. By efficiently synthesizing the degree and the dynamic queue length of nodes, we propose the global hybrid (GH) routing strategy. It can achieve higher traffic capacity, as well as shorter average packets, travel time compared with the state-of-the-art global dynamic (GD) routing strategy and efficient routing (ER) strategy. Moreover, such superiority can be maintained through the queue length information and the corresponding routing paths are updated periodically. The simulation results show that our GH routing strategy can provide the same traffic capacity as the GD routing strategy does, which is more than twice as high as the ER strategy. At the same time, the average packets travel time of the GH routing strategy is more than 20% smaller than that of the GD routing strategy. It is worth noted that longer updating delay makes our GH routing strategy have a greater decline in the average packets travel time. With the updating delay equals 100, the decline can be up to 40%. To illustrate the practicability of our GH routing strategy, we also applied it to a scale-free network-based data center network. The simulation results reveal that it is practical, effective, and can be used in real scenarios to improve network performance.