An invited talk organized by IEEE Sapporo Section YP will be held on March 24, 2021.

Date: To be determined.

Location: R205, Education & Research Building No. 9, Muroran Institute of Technology, 27-1 Mizumoto-cho, Muroran, Hokkaido, 0508585 Japan 


IEEE Sapporo Section Young Professionals (YP)

Emerging Networks and Systems Laboratory (ENeS), Muroran Institute of Technology

IEEE Muroran Institute of Technology Student Branch

As a new network architecture efficiently delivering video content and 
sensor data of IoT, named data networking (NDN), which caches content at 
routers and routes packets using content name, has attracted wide 
attention. In NDN, packets of requesting content items are called 
Interests. Similar to IP routers of the Internet, NDN routers transfer 
Interests by looking up the FIB (forwarding information base). Although 
the FIB in IP routers is matching table between the prefix of 
destination IP address, i.e., the network address, and the output port, 
the FIB in NDN router is matching table between the prefix of content 
name and the output face. Although hosts of publishers providing content 
of the same organization tend to exist in the same area, there is no 
geographical locality in the name of organizations. Therefore, prefix 
aggregation in NDN is more difficult compared with that in IP networks. 
For example, when considering web pages as content, there are about 
10^{11}$ content names, and there are still about $10^{9}$ prefixes of 
content name after aggregating prefixes of the same organization into 
one prefix \cite{Detti12}.

In the Internet, CDN (content delivery network) in which content items 
are delivered from cache servers provided in many networks has been 
widely used to improve the quality of content delivery and reduce the 
amount of traffic in networks. Using NDN, content items are more likely 
delivered from nearby routers to users, so the purpose of CDN, i.e., 
improving user quality and reducing the network traffic, is satisfied by 
NDN. Cache servers of CDN can also advertise prefixes of cached content 
to the networks, and cache servers of CDN can also become the hosts of 
publishers. Therefore, in this presentation, we propose to place copies 
of content at cache servers of CDN so that the aggregation effect of FIB 
entries of NDN routers is improved. Using the measured data of 
web-content location, we evaluate the proposed algorithm of allocating 
web content at routers and show that the proposed method can reduce the 
size of FIBs by about 45% on average.

Noriaki Kamiyama received his M.E. and Ph.D. degrees in communications 
engineering from Osaka University in 1994 and 1996, respectively. From 
1996 to 1997, he was with the University of Southern California as a 
visiting researcher. He joined NTT Multimedia Network Laboratories in 
1997, and he has been at NTT Network Technology Laboratories by 2016. He 
was also with the Osaka University as an invited associate professor 
from 2013 to 2014 and an invited professor in 2015. From 2017, he is a 
professor of Fukuoka University. He has been engaged in research 
concerning content distribution systems, network design, network 
economics, traffic measurement and analysis, and traffic engineering. He 
received the best paper award at the IFIP/IEEE IM 2013. He is a member 
of IEEE, ACM, and IEICE.