Let data consumers fetch desired data

Higher layers decide whether to fetch, and which piece most urgent

Lower layers decide when/what’s the best way to fetch.

Performance evaluation

measure transport performance: dataset namespace udpate delay

  1. sync interests can be lost.

  2. time between data production and reception by each node in the group.

measure application performance:

  1. Fetching delay: the time period from sending an interest packet till the requested data packet received.
  2. Throughput

The number of interest and data packet sent

  1. By each node
  2. Over each link

# of hops interest (and data) packets traverses

  1. Reflection of interest aggregation and data caching

Pening interest table

  1. Size
  2. PIT entry life time distribution
  3. satisfied/unsatisfied interest

Packet queueing delay: effectiveness of congestion control

  1. “Effective NDN congestion control based on queue size feedback”
  2. The reference list shows many existing works on this topic.

cache hit-ratio

  1. Per node
  2. Per namespace

Cache space occupancy distribution

  1. If the cache management design pre-emptively remove cached content