Wide-area content-based routing architecture

• US 20030099237A1
• Filed: 11/15/2002
• Published: 05/29/2003
• Est. Priority Date: 11/16/2001
• Status: Active Grant

• Current Assignee: Electronics and Telecommunications Research Institute
• Original Assignee: Electronics and Telecommunications Research Institute
• Inventors: Maheswaran, Muthucumaru, Ruedal, Jose Alejandro, Mitra, Arindam
• Granted Patent: US 7,339,937 B2
• Time in Patent Office: Days
• US Class Current: 370/393
• CPC Class Codes: H04L 45/00 Routing or path finding of packets in data switching networks routing or path finding in wireless networks H04W40/00
• CPC Class Codes: H04L 45/44 Distributed routing
• CPC Class Codes: H04L 45/50 using label swapping, e.g. multi-protocol label switch [MPLS]
• CPC Class Codes: H04L 45/64 using an overlay routing layer
• CPC Class Codes: H04L 67/104 Peer-to-peer [P2P] networks
• CPC Class Codes: H04L 67/63 Routing a service request depending on the request content or context

Abstract

Content networking is an emerging technology, where the requests for content accesses are steered by “content routers” that examine not only the destinations but also content descriptors such as URLs and cookies. In the current deployments of content networking, “content routing” is mostly confined to selecting the most appropriate back-end server in virtualized web server clusters. This invention presents a novel content-based routing architecture that is suitable for global content networking. In this content-based routing architecture, a virtual overlay network called the “virtual content network” is superimposed over the physical network. The content network contains content routers as the nodes and “pathways” as links. The content-based routers at the edge of the content network may be either a gateway to the client domain or a gateway to the server domain whereas the interior ones correspond to the content switches dedicated for steering content requests and replies. The pathways are virtual paths along the physical network that connect the corresponding content routers. The invention is based on tagging content requests at the ingress points. The tags are designed to incorporate several different attributes of the content in the routing process. The path chosen for routing the request is the optimal path and is chosen from multiple paths leading to the replicas of the content.

Claims

1. A method for directing packets of data based on their content in a telecommunications network,

wherein the physical network comprises a plurality of clients, a plurality of servers for supplying content and a plurality of content aware routers for directing communications over the network;

wherein content represents the payload carried in the packets;

the method comprising;

providing a content intelligent overlay network that is a Virtual Content Network (VCN) created on top of the physical network;

wherein the overlay network and a digest of the content is used to select a preferred path for the packets and the physical network layer is used to transfer the packets from one end to other end of the path;

wherein the Virtual Content Network (VCN), acts as a distributed virtual content router and is transparent to the users;

wherein the VCN is sparser than the underlying physical network and is formed by content edge routers (CERs) placed at the edge and the content switching routers (CSRs) in the core of the VCN;

and wherein the CERs and CSRs are connected using logical content switching paths (CSP).

2. The method according to claim 1 including a method in which all contents hosted by the VCN with similar characteristics are grouped together to create sets referred to as Content-Equivalence Class (CEC) in which the CECs allow handling of large volumes of data by creating content aggregates;

wherein the method comprises the steps;

(a) identification and describing content attributes where the attributes are classified as structural and semantical attributes of the contents and the attributes are used to characterize and classify the contents hosted by the VCN and create the CECs for these contents and where all CECs are unique within the VCN i.e. each CEC will have a unique set of attribute values;

(b) implementing a tagging function to create a fixed sized content-derived tag that is used to identify each CEC where the tag is derived based on the attribute values of each CEC and each tag is unique within the VCN and is used to route packets within the VCN; and

(c) disseminating the tags within the VCN either at periodic intervals or on-demand.

3. The method according to claim 2 wherein each content-derived tag actually corresponds to a finger-print for the content that is requested by the packet and represents a single or a set of servers that hosts the content;

the destination of a packet (or a flow) is chosen from this set of servers during routing.

4. The method according to claim 2 including encapsulation of request packets using the content derived tags by attaching the content derived tag as a content header to the request packets, where the encapsulated packets are then pushed into the VCN and the tag in the content header is used to route the packets.

5. The method according to claim 2 including replacement of the normal IP routing table with the content-derived tag routing table at the content routers within the VCN where each content router within the VCN maintains routing tables to hold the content derived tags for content being hosted by the VCN.

6. The method according to claim 2 including calculating the routing fractions from the load condition of the hosting servers and the condition of the path leading to the servers from any point in the VCN where the tag routing tables will contain the tags and the routing fractions, where the tags are used to select a set of servers and the routing fraction is used to find the best path to the best server and where a routing algorithm uses the tags to identify the requests and route the packets towards the ‘

best’

destination.

7. The method according to claim 2 wherein contents are classified using attributes which allow access to portions of a content hosting site at variable granularities i.e. it is possible to partition/replicate content hosting sites using different attribute values of the content hosted by the sites and no longer requires simplistic information like the domain or locational information of the site for partitioning or replication.

8. The method according to claim 2 wherein request packets are tagged using content derived tags so as to eliminate the round-trip time involved for resolving (i.e., mapping) a request on to a specific server and so that the latency for content discovery is reduced.

9. The method according to claim 2 wherein the routing decision by the CSRs within the VCN core use the content-derived tags so as to eliminate the time required to parse the request packet repeatedly at each CSR along the routing path to make a routing decision.

10. The method according claim 2 wherein the CECs and the content-derived tags are created by the CERs prior to the routing process by the steps;

(a) content characterization wherein identification of the structural and semantical attributes of content;

(b) content classification wherein the content attributes are used to classify content and create the CECs for each autonomous system (AS) in the Internet.

11. The method according to claim 2 including a routing algorithm called Protocol Independent Content Switching (PICS) that uses the content-derived tags to route packets;

wherein the methods of operation of PICS include;

(a) a route push method wherein the content-based routes i.e., the tags are pushed from the server side CERs to all the client-side CERs wherein all CERs and CSRs know about all the tags maintained by the VCN;

(b) a route push and pull method wherein the content-based routes are disseminated and maintained in a distributive manner in the VCN by using a virtual content edge router (VCER).

12. The method according to claim 2 wherein the tags within the VCN are maintained in a distributive manner by using the content edge router VCER, where

(a) the virtual content edge router consists of physical CERs placed in a tree-like fashion;

(b) the lowest level of CERs of the VCER tree maintain tags for the most popular contents, the topmost level maintain tags for the least popular contents while the intermediate VCER levels handle contents of intermediate popularity.

13. The method according to claim 2 wherein the popularity of the contents is changed dynamically by:

(a) the VCER tree is used to define a fast and slow mode of routing.

(b) in the fast mode tags for all most popular contents are stored at the lowest level of the tree and in the slow mode, all less popular contents are stored in the higher levels of the tree.

(c) the tags are pulled down or pushed up the tree as the popularity of the contents change.

14. The method according to claim 2 wherein the selection of the route is facilitated by:

(a) obtaining information about the contents hosted by sites who wish to use the VCN for content delivery (i.e., content providers subscribe to the VCN);

(b) performing analysis of the Internet traffic for these contents;

(c) maintaining the content and corresponding tag information at the edge of the VCN.

15. The method according to claim 4 wherein the content header is used to encapsulate the data packets so as to include the content storage and resource requirements in the content-derived tags in a digest form and this information is used by the CERs and CSRs to make routing decisions.