A Survey on Content-Oriented Networking for Efficient Content Delivery Authors: Jaeyoung Choi, Jinyoung Han, Eunsang Cho, Ted “Taekyoung” Kwon, and Yanghee Choi, Seoul National University Presenters: Prithvi Monangi , 16858394 Amrutha Alaparthy : 66905246

Existing Issue 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 content oriented networking (CON) by decoupling contents from hosts at the networking level.

Contents discussed Content Naming Name based routing Comparison of CON routing proposals Evaluation of the impact of publish/subscribe paradigm and in- network caching

Introduction As the bit rate of multimedia traffic increases , TCP/IP may fail to deliver time sensitive data. Two problems with host based Internet architecture: CPs like youtube, Hulu etc deliver HDTV streaming services multicasting/broadcasting over IP networks (IPTV) TCP/IP architecture fails to efficiently handle the above tasks

Existing Solutions At present, multimedia is delivered to numerous users using P2P systems like BitTorrent BitTorrent : Inefficient in networking perspective: peer can download chunks only from a small subset of peers who may be distantly located Content-oriented applications/services, doesn’t care about hosts, but focusses on contents.

Existing Solutions Host oriented architecture - content delivery This mismatch leads to application/service-specific solutions, which may be costly and/or inefficient. Two representative examples are: Web caches and content delivery networks (CDNs) transparently redirect web clients to a nearby copy of the content file. P2P systems enable users to search and retrieve the content file.

CON vs IP CON node performs routing by content names, not by (host) locators. This means two radical changes: • Identifying hosts is replaced by identifying contents. • The location of a content file is independent of its name. This makes CON location independent.

CON vs IP 2) publish/subscribe paradigm is the main communication model in CON, which helps us to decouple content generation and consumption in time and space….In IP, user should know which source holds the file. 3) Authenticity of contents can easily be verified by leveraging public key cryptography. In IP networking, a host address seen by a user is irrelevant to its content name, which results in phishing attacks

CON Architecture A CON architecture can be characterized by four main building blocks: • How to name the contents • How to locate the contents (routing) • How to deliver/disseminate the contents • How to cache the contents “in” the network

Content Naming Hierarchical Naming Content file is often named by an identifier like web URL. Hierarchical nature can help us mitigate the routing scalability issue since routing entries for contents might be aggregated. components in a hierarchical name (e.g., www.acme.com and logo.jpg) have semantics, which prohibits persistent naming. I.e; if the ownership is changed its name becomes misleading

Content Naming Flat Naming To eliminate shortcomings like persistence naming of hierarchical naming Flat and self-certifying names are employed by defining a content identifier as a cryptographic hash of a public key. Persistence and uniqueness are achieved. Routing scalability decreases as there is no aggregation. An additional resolution is needed b/w application level and human readable names.

Content Naming Attribute Based Naming Identifies content with set of attribute value pairs which facilitates in network searching Drawbacks : An AVP may not be unique or well defined. The semantics of AVPs may be ambiguous. The number of possible AVPs can be huge

Name Based Routing Locating the content based on the name Two types of classification: Unstructured routing : No structure to maintain routing tables Routing advertisement is performed by flooding Network prefixes in IP routing are replaced with content Identifiers As the replication of content file increases, the level of aggregation decreases

Name Based Routing Structured routing : Two structures : Implemented using either DHT or Tree In tree arch, each router contains the routing information of all the contents published in its descendant routers As the level of routing increases, routing burden increases - not scalable In DHT, flatness imposes equal routing burden on routers n(Contents) = C, each router should have log C entries DHT might exhibit longer paths than tree sometimes

Multi Source Dissemination Further issues in CON Multi Source Dissemination Existing internet architecture is designed with point to point connectivity New applications require different connectivities like 1:N (one to many) and M:N (many to many) 1: N - IPTV, online streaming - uses IP multi casting framework CON accommodates 1:N connectivity naturally with its publish- subscribe mechanism in terms of Content naming and group management

Multi Source Dissemination Further issues in CON Multi Source Dissemination M:N Many to Many Multiple sources disseminate different parts of the content to n recipients - Bittorrent and multi-user online gaming Existing Internet requires application specific overlays for M:N Example: P2P networks - Bittorrent P2P systems are application level solutions, cannot exploit network topology information.

Multi Source Dissemination Further issues in CON Multi Source Dissemination

Multi Source Dissemination Further issues in CON Multi Source Dissemination M:N Many to Many CON can efficiently disseminate a content file among subscribers since CON nodes (R2 and R4) will help them download the content file also from the other overlay. CON keeps track of individual sources of the same content allowing us to retrieve different parts of requested content in parallel CON node dynamically decides which part to be received from which source based on traffic dynamics

Multi Source Dissemination Further issues in CON Multi Source Dissemination M:N Many to Many Issue: What routing info needs to be stored at each CON node for multiple sources of same content The more sources of the same content the CON node learns of, the more selectively it may have to propagate the routing information of the sources

Further issues in CON In Network Caching Generally used caching techniques: LRU - Least Recently used and LFU - Least frequently used content file replacement Performance can be further improved by coordinating multiple CON nodes in distributed fashion Study of distributed caching in IP networking is limited to single source dissemination and limited network topologies. we need to reformulate the distributed caching problem in CON environments like multi source dissemination

In Network Caching - continued Further issues in CON In Network Caching - continued Design Issue: How to design a signalling protocol among CON nodes. If a routing protocol is used to facilitate coordinated caching among CON nodes, then here will be no significant signaling traffic overhead The more frequently the content information is updated, the more routing info may have to be advertised - This needs to be controlled

Performance Evaluation We evaluate two things: The effect of routing structures on resolution time to locate a content file How much traffic load can be reduced by in network caching

Two-Tier architecture Hybrid approach: Tree can be formulated with network topology information (e.g., hop count between nodes), tree routing achieves higher throughput than DHT routing. DHT routing is more scalable in terms of routing burden and more resilient to node/link failures due to multiple paths than tree routing. A hybrid approach whose routing structure consists of two tiers: a DHT is the high tier, and a tree is the low tier is introduced.

Two-Tier architecture Hybrid approach: A query for a content file published in the same tree will be serviced within the tree. If a query is for a content file outside the tree structure, the DHT structure is exploited to forward the query

Comparisons Resolution delays: Tree structure outperforms DHT because, the content request goes back and forth among CON nodes of DHT as the topology is constructed without any information on physical topology Performance of two-tier architecture falls between these two Successful resolution ratio as node failure rate increases : Performance gain of DHT over tree because, there are multiple paths among nodes in DHT where as in tree structure, higher level node failure results in node failures

Network traffic load The performance metric is the product of hop count and link bandwidth This diminishes in CON proposals due to cache effect Two-tier exhibits poorer performance than tree due to DHT overlay inefficiency

Conclusion CON vs IP Characteristics of CON - naming, routing etc Research topics in CON environments: M:N dissemination In network caching

THANK YOU