期刊
IEEE TRANSACTIONS ON NETWORK AND SERVICE MANAGEMENT
卷 18, 期 1, 页码 882-893出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TNSM.2020.3040907
关键词
Optical switches; Neurons; Scheduling; Feature extraction; Training; Switching circuits; Biological neural networks; Intra-datacenter network; spiking neural network; traffic prediction; traffic scheduling
资金
- NSFC project [61871056]
- Young Elite Scientists Sponsorship Program by CAST [2018QNRC001]
- Beijing Natural Science Foundation [4202050]
- Fund of SKL of IPOC (BUPT) [IPOC2018A001, IPOC2019ZT01]
This article proposes a high-accuracy burst traffic prediction method based on SNN framework, as well as a prediction-assisted scheduling algorithm to tackle traffic challenges in hybrid E/O switching DCN. Simulation results demonstrate that this approach can effectively enhance traffic feature extraction and scheduling performance.
Hybrid electrical/optical (E/O) switching data center network (DCN) has recently emerged as a promising paradigm for future DCN architectures. However, there exist two major challenges: 1) the traffic is a mixture of both stable and burst components due to the diverse and heterogeneous user demands; 2) current scheduling algorithms are mostly static and not designed for the complex structure of hybrid E/O switching DCN, provoking frequent burst traffic congestion and performance degradation. This article endeavors to overcome the above challenges as follows. We first construct an error feedback-based spiking neural network (SNN) framework with high accuracy burst traffic prediction. We then design a prediction-assisted scheduling algorithm to handle the worst-case burst traffic. On the one hand, the error feedback-based SNN framework can significantly enhance the extraction of burst traffic features by mimicking the biological neuron system. On the other hand, prediction-assisted scheduling arranges the well-predicted traffic using a global evaluation factor and a traffic scaling factor. The simulation results reveal that our approach can efficiently integrate a spiking neural network into the traffic scheduling scheme and achieve satisfying performance with affordable computational complexity.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据