NEURAL NETWORK-BASED MOBILITY MANAGEMENT FOR MOBILE AD HOC RADIO NETWORKS

• US 20070299946A1
• Filed: 06/26/2006
• Published: 12/27/2007
• Est. Priority Date: 06/26/2006
• Status: Active Grant

• Current Assignee: The Boeing Co.
• Original Assignee: The Boeing Co.
• Inventors: El-Damhougy, Hesham
• Granted Patent: US 7,742,425 B2
• Time in Patent Office: Days
• US Class Current: 709/223
• CPC Class Codes: H04L 43/0811 by checking connectivity
• CPC Class Codes: H04L 67/12 specially adapted for proprietary or special-purpose networking environments, e.g. medical networks, sensor networks, networks in vehicles or remote metering networks

Abstract

A self managed ad hoc communications network and method of managing the network. The network includes wireless devices or nodes that include a neural network element and the ad hoc network operates as a neural network. One of the nodes is designated as a Network Management System (NMS) that provides overall network management. Clusters of nodes are organized around cluster leaders. Each cluster leader manages a cluster of nodes and communications between node clusters. Each cluster may also have other nodes identified as lower order cluster leaders.

Claims

1. An ad hoc communications network comprising:

a plurality of wireless devices connectively coupled, each of said wireless devices including a neural network element;

a Network Management System (NMS) providing overall network management, said NMS being one of said plurality of wireless devices; and

at least one cluster leader, each cluster leader managing at least one said wireless devices.

2. An ad hoc communications network as in claim 1, wherein at least one said group includes a plurality of leaders, one of said plurality of leaders being identified as the cluster leader and the remaining ones of said plurality of leaders being lower order cluster leaders.

3. An ad hoc communications network as in claim 1, wherein each of said cluster leaders monitors connectivity of wireless devices in a respective said group and passes identified changes to said NMS.

4. An ad hoc communications network as in claim 1, wherein said NMS leaders monitor connectivity of said plurality of wireless devices and passes identified changes to each said cluster leader.

5. An ad hoc communications network as in claim 1, wherein each of said wireless devices is a node in a cluster of said plurality of wireless devices.

6. An ad hoc communications network as in claim 5, wherein each said node in each said cluster contains a copy of a connectivity matrix indicating connectivity within a respective said cluster.

7. An ad hoc communications network as in claim 6, wherein each said connectivity matrix indicates connectivity of all nodes within said network.

8. An ad hoc communications network as in claim 5, wherein each of said plurality of cluster leaders has a minimum cluster leader objective function value for its respective cluster.

9. A method of managing an ad hoc communications network comprising the steps of:

a) determining a cluster leader objective function value for each node in a cluster, said cluster having a plurality of nodes;

b) providing each cluster leader objective function value to a current cluster leader identified for said cluster;

c) selecting a minimum value from provided cluster leader objective function values, the selected said value identifying a cluster leader for said cluster; and

d) broadcasting the identity of said cluster leader to said plurality of nodes.

10. A method as in claim 9, wherein whenever, the identified cluster leader is a different node than said current cluster leader, said method further comprises the steps of:

e) initiating a handover from said current cluster leader to said identified cluster leader, said identified cluster leader node becoming said current cluster leader; and

f) sending a hello message from said current cluster leader.

11. A method as in claim 10, wherein the step (c) further comprises selecting lower order cluster leaders from said plurality of nodes.

12. A method as in claim 11, wherein said lower order cluster leaders include a secondary cluster leader and a tertiary cluster leader.

13. A method as in claim 11, wherein if said hello message is not received by said lower order cluster leaders after a maximum waiting period, one of said lower order cluster leader becomes current cluster leader and sends a second hello message.

14. A method as in claim 13, wherein if said second hello message is not received by remaining said lower order cluster leaders after another maximum waiting period, one of said remaining lower order cluster leader becomes current cluster leader and sends a hello message.

15. A method as in claim 9, wherein said cluster is one of a plurality of clusters and the step (d) of broadcasting further broadcasts the identity of said cluster leader to other ones of said plurality of clusters.

16. A method as in claim 9, wherein each time entries in a connectivity matrix for said plurality of nodes in said cluster change, each node returns to step (a) and determines said cluster leader objective function value.

17. A method as in claim 9, wherein for each (N_k(t)) within Voronoi region k at time t, said cluster leader objective function value has the form

18. A method as in claim 9, wherein for each (N_k(t)) within Voronoi region k at time t, said cluster leader objective function value has the form

19. A method of organizing and managing an ad hoc communications network comprising the steps of:

a) selecting a Network Management System (NMS) from a plurality of nodes forming an ad hoc communications network;

b) identifying nodes with a minimum cluster leader objective function value in each cluster of nodes in an operating coverage area, identified said nodes being selected as cluster leaders;

c) providing training results to said each cluster of nodes; and

d) monitoring connectivity within said each cluster of nodes.

20. A method as in claim 19, wherein whenever the monitoring step (d) identifies network changes in a cluster, said method further comprises the steps of:

e) determining said cluster leader objective function value for each node in said cluster;

f) providing each cluster leader objective function value to a current cluster leader identified for said cluster;

g) selecting a minimum value from provided cluster leader objective function values, the selected said value identifying a cluster leader for said cluster;

h) selecting lower order cluster leaders from remaining nodes in said cluster; and

i) broadcasting the identity of said cluster leader to said nodes in said cluster and to other said clusters.

21. A method as in claim 20, wherein whenever the identified cluster leader is a different node than said current cluster leader, said method further comprises the steps of:

j) initiating a handover from said current cluster leader to said identified cluster leader, said identified cluster leader node becoming said current cluster leader; and

k) sending a hello message from said current cluster leader.

22. A method as in claim 21, wherein said lower order cluster leaders include a secondary cluster leader and a tertiary cluster leader.

23. A method as in claim 22, wherein if said hello message is not received by said secondary cluster leaders after a maximum waiting period, said secondary cluster leader assumes current cluster leader and sends a second hello message.

24. A method as in claim 23, wherein if said second hello message is not received by said tertiary cluster leaders after another maximum waiting period, said tertiary cluster leader assumes current cluster leader and sends a hello message.

25. A method as in claim 24, wherein for each (N_k(t)) within Voronoi region k at time t, said cluster leader objective function value has the form

26. A method as in claim 24, wherein for each (N_k(t)) within Voronoi region k at time t, said cluster leader objective function value has the form

27. A method of organizing an ad hoc communications network as in claim 20, before the step (a) of selecting the NMS, said method further comprising the steps of:

a1) providing operating environment parameters; and

a2) tessellating an operating coverage area, the tessellated said operating coverage area indicating communication between said plurality of nodes;

a3) identifying clustered nodes within the tessellated said operating coverage area.

28. A mobile communications station comprising:

a wireless communications unit for connecting to other mobile communications stations;

a storage unit storing network tables and matrices indicating wireless connectivity and connection to said other mobile communications stations; and

an Artificial Neural Network (ANN) element trained to determine a cluster leader objective function value for each node in a cluster of a plurality of nodes, each said node being another mobile communications station, select a minimum value from determined said cluster leader objective function values, said mobile communications node being the cluster leader for said cluster whenever said minimum value corresponds to said mobile communications node, and wirelessly broadcast a cluster leader identity to said plurality of nodes in said cluster.

29. A mobile communications station as in claim 28, wherein said mobile communications node is trainable to hand off cluster leader whenever said minimum value corresponds to another node.

30. A mobile communications station as in claim 28, wherein said mobile communications node is trainable to selectively assume cluster leader whenever a previously identified said cluster leader fails to become cluster leader.

31. An mobile communications station as in claim 28, wherein said mobile communications station is coupled to an unmanned ground vehicle.

32. An mobile communications station as in claim 28, wherein said mobile communications station is coupled to an unmanned air vehicle.