Eunsang Cho

A Survey on content-oriented networking for efficient content delivery

As multimedia contents become increasingly dominant and voluminous, the current Internet architecture will reveal its inefficiency in delivering time-sensitive multimedia traffic. To address this issue, there have been studies on contentoriented networking (CON) by decoupling contents from hosts at the networking level. In this article, we present a comprehensive survey on content naming and name-based routing, and discuss further research issues in CON. We also quantitatively compare CON routing proposals, and evaluate the impact of the publish/subscribe paradigm and in-network caching.

Performance comparison of content-oriented networking alternatives: A tree versus a distributed hash table

While the Internet was designed with host-oriented networking applications, recent Internet statistics show that content-oriented traffic has become more and more dominant. Even though content-oriented networking, which tries to resolve this discordance, has received increasing attention, there have been few comprehensive and quantitative studies on how to realize a content-oriented networking architecture. In this paper, we focus on the design alternatives of the content-oriented networking architecture and evaluate their performance: (i) how to locate contents, (ii) how to cache contents, and (iii) how to deliver contents. There are two major infrastructure alternatives in substantiating these mechanisms: a tree and a distributed hash table (DHT).We carry out comprehensive simulation experiments to compare these alternatives in terms of content transfer latency, cache effectiveness, and failure resilience.

Bandwidth Allocation for BitTorrent under Multi-Torrent Environments

BitTorrent has achieved a great success in the field of peer-to-peer (P2P) file sharing. Although BitTorrent allows peers to share files efficiently and scalably, it shows inefficiency when a client participates in multiple torrents where each of them concurrently competing for the limited link bandwidth. In this paper, we propose a new bandwidth allocation algorithm, which greedily increases the bandwidth consumption for downloading, to reduce file transfer time considering the current download/upload status. To compensate overall performance degradation resulting from our greedy allocation, we suggest modifying the choking algorithm of BitTorrent to consider the ratio of seeders and leechers in each torrent. Through comprehensive experiments, we validate the performance gain of the proposed scheme over original BitTorrent in a mix of WiFi and Ethernet testbed and large scale public torrents.

Effects of partial infrastructure on indoor positioning for emergency rescue evacuation support system

Indoor positioning is a key component of an Emergency Rescue Evacuation Support System (ERESS). However, there has been limited works on evaluating indoor positioning schemes in consideration of emergency situations such as fire and earthquakes. Especially, some or all of the indoor infrastructures (e.g. WiFi APs) could be unavailable due to the disasters, which are reference points of the schemes. We design and implement a realistic simulation to evaluate the schemes with ignition of fire, spread of smoke, and malfunction of infrastructure. In the simulation, positioning accuracy of two representative bases of indoor localization schemes is evaluated, which are Pedestrian Dead Reckoning (PDR) and Wi-Fi fingerprinting based on Received Signal Strength (RSS). In both techniques, we use the Particle Filter algorithm to improve the accuracy of localization. For realistic simulations, we model a WiFi infrastructure and its breakdown process using the Anylogic software with smoke propagation data with the Fire Dynamics Simulator (FDS). We also compare the real time pedestrian trace with the estimated locations obtained by the schemes.

TwinPeaks: A new approach for certificateless public key distribution

The current PKI has problems like certificate revocations and fraudulent certificates. To address such issues, we propose TwinPeaks, which is a new infrastructure to distribute public keys of named entities online. TwinPeaks leverages certificateless public key cryptography (CL-PKC), which we extend to make the public key of an entity depend on any combination of its networking parameters; thus TwinPeaks can mitigate spoofing attacks systematically. TwinPeaks needs public key servers, which constitute a DNS-like hierarchical tree. For each parent-child link in the tree, the parent and the child interact in such a way that every named entity has its own public/secret key pair. TwinPeaks removes certificates and hence has no revocation overhead. Instead, each named entity should keep/update its IP address and public key up-to-date in its DNS server and key server, respectively. TwinPeaks also achieves scalable distribution of public keys since public keys can be cached long term without elevating security risks.

LOCON: A lookup-based content-oriented networking framework

While the current Internet has a host-based TCP/IP architecture, the vast majority of Internet usage is attributed to the content retrieval and distribution. This mismatch has proliferated content delivery network (CDN) technologies and P2P file-sharing systems (e.g., BitTorrent); however, the inefficiency of content delivery is not fundamentally solved, and there are business/operation and content copyright/availability issues, respectively. Also, in the research community, there have been efforts to redesign the current architecture from a content-centric perspective by introducing the route-by-name paradigm and in-network caching, which however entails many issues like the routing scalability and processing overhead. We propose a lookup-based communication framework for content-oriented networking (LOCON), whose main components are locator lookup, coordinated caching, and parallel transmissions. LOCON aims to accommodate incremental deployment, support legacy end-hosts, and provide business incentives between network operators and content publishers. To carry out the experiments, we build a networking testbed on top of Amazon data centers around the world. LOCON performs better than other lookup-by-name networking technologies and CCN (i.e., a route-by-name scheme) in terms of content delivery metrics.

LOMO: location resolution of mobile hosts using internet routing structures

The DNS has performed the resolution for quasi-static hosts so far. However, mapping from the names of mobile hosts to their locators is increasingly important. We propose a new resolution architecture, LOMO, that keeps track of the locators of mobile hosts as they are changing the points of attachment. LOMO takes a host-based approach and leverages current IP routing structures to route the resolution traffic in a scalable manner and to balance the workload of resolution.