ndn||mem: an architecture to alleviate the memory bottleneck for named data networking

  • Authors:
  • Mostafa Rezazad;Y. C. Tay

  • Affiliations:
  • National University of Singapore (NUS), Singapore, Singapore;National University of Singapore (NUS), Singapore, Singapore

  • Venue:
  • Proceedings of the 2013 workshop on Student workhop
  • Year:
  • 2013

Quantified Score

Hi-index 0.00

Visualization

Abstract

Named Data Networking (NDN), as one of the candidates for a future Internet, has the potential to overcome many of the current Internet's difficulties (e.g., security, mobility, multicasting, etc.). However, there are also many open problems to be taken care of before it becomes operational. One of the major concerns is that NDN may be memory intensive. Using the router's buffer as cache adds a search time to all requests. If the cache hit rate is not high enough (considering current buffer sizes and traffic rates), cache search time may be a burden on the system instead of solving any problems. Huge fast memories are not affordable yet. To balance the memory speed with incoming rate to a router, we propose a memory architecture, called ndn||mem, along with a content caching strategy, called CCndn. Using these two ideas, only a small fraction of packets need to engage in a time-intensive cache search process. Simulation results demonstrate the efficacy of our architecture