Chinese Journal on Internet of Things ›› 2018, Vol. 2 ›› Issue (4): 5-13.doi: 10.11959/j.issn.2096-3750.2018.00078

• Theory and Technology • Previous Articles     Next Articles

Crowd-based collaboration caching mechanism in smart identifier network

Haifeng LI1,Wei QUAN1,Nan CHENG2,Hongke ZHANG1,Xuemin SHEN2   

  1. 1 National Engineering Laboratory for Next Generation Internet Technologies,Beijing Jiaotong University,Beijing 100044,China
    2 Department of Electrical and Computer Engineering,University of Waterloo,Waterloo N2L 3G1,Canada
  • Revised:2018-10-30 Online:2018-12-01 Published:2019-01-03
  • Supported by:
    The National Natural Science Foundation of China(61602030);The National Natural Science Foundation of China(91638204);The National Key Research and Development Program of China(2016YFE0122900);The Fundamental Research Funds for the Central Universities(2016RC036)

Abstract:

Smart identifier network (SINET) is an innovative network architecture.Through dynamic collaboration of service resource,function groups and physical components,the network scalability,resource utilization and service quality were effectively improved by SINET,and an effective solution for the development of industrial Internet of things(IoT) could be provided.To promote content delivery in resource-constrained IoT,caching function was introduced in network components by SINET,and the bandwidth waste caused by traffic redundancy in resource-constrained node could be reduced.Therefore,how to efficiently cache content became an important research topic.Based on SINET architecture,a crowd-based collaboration cache (C2Cache) mechanism was proposed in this scheme.According to the actual network topology,the caching function group was dynamically created and optimized by C2Cache,and a function group as crowd minimum unit to execute the maximum benefit cache (MBC) algorithm was defined to maximize the caching space efficiency.With the self-developed emulation system,named EmuStack,the performance of C2Cache was evaluated.The experimental results show that,comparing with LCE,Random,Prob Cache,LCD and Greedy caching mechanisms,the cache hit rate can be improved effectively,then the average access time can be reduced significantly by C2Cache.In the simulated network scenario,the performance increases 15% to 30%.

Key words: smart identifier network, in-network cache, crowd-based collaboration, maximum benefit cache

CLC Number: 

No Suggested Reading articles found!