Layered communications network with adaptively distributed nodal policies

• US 7,346,704 B2
• Filed: 09/14/2001
• Issued: 03/18/2008
• Est. Priority Date: 09/14/2000
• Status: Expired due to Fees

• Current Assignee: British Telecommunications PLC (BT Group PLC)
• Original Assignee: British Telecommunications PLC (BT Group PLC)
• Inventors: Roadknight, Christopher M, Marshall, Ian W
• Primary Examiner(s): Vaughn; William
• Assistant Examiner(s): SERRAO, RANODHI N
• Application Number:
  US10/380,003
• Publication Number:
  US 20030177262A1
• Time in Patent Office: 2,377 Days
• Field of Search: 709/238
• US Class Current: 709/238
• CPC Class Codes: H04L 41/0893 Assignment of logical groups to network elements
• CPC Class Codes: H04L 41/0894 Policy-based network configuration management
• CPC Class Codes: H04L 45/08 Learning-based routing, e.g. using neural networks or artificial intelligence
• CPC Class Codes: H04L 45/566 Routing instructions carried by the data packet, e.g. active networks

Abstract

An emergent network is autonomous at the service level. Network nodes have policies that enable them to process different types of service requests, with the processing earning the nodes ‘rewards’. Successful nodes can pass some or all of their policies to other nodes using the evolutionary biology of bacteria as a model.

Claims

1. A method of operating a multi-service communications network comprising a polygonal topology and including a plurality of nodes located along at least one edge of the network and a plurality of nodes located away from the edge of the network, each node being directly connected to one or more neighboring nodes such that each node is directly or indirectly connected to every other node in the network, the method comprising, at each node:

either (a) selectively performing one or more services upon receipt of a service request in accordance with one or more nodal policies, when implemented by the node, or (b) forwarding a service request to a neighboring node,

each nodal policy including;

(i) a service request type identifier indicating the type of service request to which the policy applies, and (ii) one or more service request criteria which specify the circumstances under which a node implementing the policy will perform a requested service;

wherein, whenever a node receives a service request at option (a), either from a user of the network or from a neighboring node, if the node is operating in accordance with a nodal policy whose service request type identifier corresponds to the type of the received service request and the criteria set out in the nodal policy are satisfied, then the service request is processed by the node or queued by the node for subsequent processing, but otherwise at option (b) the request is forwarded to a neighboring node, unless, in either case, the request has timed out;

maintaining an activity indicator whose value is determined in dependence upon the amount of time the node spends performing services in response to service requests; and

selectively varying a nodal policy or its implementation with a probability which is inversely correlated to the value of the activity indicator,

wherein service requests from users of the network are initially inserted into node locations along at least one edge of the multi-service communications network, and

wherein each node, when forwarding a service request to a neighboring node, selects a neighboring node to which to forward the request to cause a tendency for forwarded service requests to migrate away from said plurality of nodes located along said at least one edge of the network.

2. A method according to claim 1 wherein the service request criteria of one or more nodal policies includes a time to live for a received service request.

3. A method according to claim 1 wherein, if the activity indicator reaches a first upper threshold level, then the node exports one or more of its nodal policies.

4. A method according to claim 3 wherein, if the activity indicator reaches a second upper threshold, then the node replicates all of its nodal policies to generate a clone of the node.

5. A method according to claim 1 wherein, if the activity indicator reaches a first lower threshold, then the node imports a further nodal policy.

6. A method according to claim 5 wherein, if the activity indicator reaches a second lower threshold, then the node deletes an enabled nodal policy and enables a dormant nodal policy.

7. A method according to claim 6 wherein, if the activity indicator reaches a third lower threshold, then a variable within a nodal policy is randomly varied.

8. A data storage medium containing computer code which, when executed by a computer at each of plural nodes in a multi-services communications network causes the network to carry out the method of claim 1.

9. A multi-service communications network comprising a polygonal topology and including a plurality of nodes located along at least one edge of the network and a plurality of nodes located away from the edge of the network, each node being directly connected to one or more neighboring nodes such that each node is directly or indirectly connected to every other node in the network, wherein each node comprises a processor, said processor having:

means for either (a) selectively performing one or more services upon receipt of a service request in accordance with one or more nodal policies, when implemented by the node, or (b) forwarding a service request to a neighboring node,

each nodal policy including;

(i) a service request type identifier indicating the type of service request to which the policy applies, and (ii) one or more service request criteria which specify the circumstances under which a node implementing the policy will perform a requested service;

wherein whenever a node receives a service request at option (a), either from a user of the network or from a neighboring node, if the node is operating in accordance with a nodal policy whose service request type identifier corresponds to the type of the received service request and the criteria set out in the nodal policy are satisfied, then the service request is processed by the node or queued by the node for subsequent processing, but otherwise at option (b) the request is forwarded to a neighboring node, unless, in either case, the request has timed out;

means for maintaining an activity indicator whose value is determined in dependence upon the amount of time the node spends performing services in response to service requests; and

means for selectively varying a nodal policy or its implementation with a probability which is inversely correlated to the value of the activity indicator,

wherein the network is arranged such that service requests from users of the network are initially inserted into node locations along at least one edge of the multi-service communications network, and

wherein each node is operable, when forwarding a service request to a neighboring node, to select a neighboring node to which to forward the request to cause a tendency for forwarded service requests to migrate away from said plurality of nodes located along said at least one edge of the network.