Method and system for detecting failures of network nodes

1. A central controller that detects device failures in a network, said network including a plurality of nodes coupled to the central controller, wherein a first of said nodes communicates with the central controller via a second of said nodes, said central controller having a processor that executes program instructions stored in a memory device that cause the central controller to:

record a failure alert received from the second node in a log stored on the memory device, said second node providing said failure alert based on a determination that the first node has not transmitted at least one message within a predefined time period;

determine, by the processor, that the first node has failed based on a set of failure alerts recorded in the log in association with the first node; and

generate, when the processor determines that the first node has failed, a failure notification causing a notification device to indicate a failure.

2. The central controller of claim 1 , wherein the network is a grid network and messages transmitted from the first node to the central controller are relayed to the central controller through a subset of the plurality of second nodes that lie in one or more predetermined routes through the grid network to the central controller.

3. The central controller of claim 2 ,

wherein only a node that is a direct neighbor of the first node and resides along one of the plurality of routes to the central controller provides failure alerts regarding the first node.

4. The central controller of claim 1 , wherein the at least one message is a message broadcast by the first node at a scheduled frequency and time.

5. The central controller of claim 4 , wherein the broadcast message is a beaconing message.

6. The central controller of claim 1 , wherein the processor determines that the first node has failed when the log includes a predetermined number of alerts received from a predetermined number of nodes within a predetermined period of time.

7. The central controller of claim 1 , wherein the processor determines whether the first node has failed based on one or more policies.

8. The central controller of claim 7 , wherein the processor determines whether the first node has failed based on a policy selected from a plurality of policies, said policy being dynamically selected by the processor based on the current condition of the network.

9. The central controller of claim 1 , wherein the failed node is removed from the log.

10. The system of claim 1 , wherein the processor of the central controller executes the program instructions to cause the central controller to:

receive a failure alert from at least one third node;

record the failure alert received from the at least one third node in the log in association with the failure alert received from the second node; and

determine that the first node has failed by determining a number of failure alerts for the first node received from the second node in relation to a number of failure alerts received from the at least one third node.

11. A method for a central controller to detect device failures in a network comprised of a plurality of communicatively-coupled network nodes, wherein a first node communicates with the central controller via a second node that is in communication with the central controller, said method comprising:

recording a failure alert received from the second node in a log stored on the memory device, said second node providing said failure alert based on a determination that the first node has not transmitted at least one message within a predefined time period;

determining, by the processor, that the first node has failed based on a set of failure alerts recorded in the log in association with the first node; and

generating, when the processor determines that the first node has failed, a failure notification causing a notification device to indicate a failure.

12. The method of claim 11 , wherein the network is a grid network and messages transmitted from the first node to the central controller are relayed to the central controller through a subset of the plurality of second nodes that lie in one or more predetermined routes through the grid network to the central controller.

13. The method of claim 12 , wherein only a node that is a direct neighbor of the first node and resides along one of said routes to the central controller provides failure alerts regarding the first node.

14. The method of claim 11 , wherein the at least one message is a message broadcast by the first node at a scheduled frequency and time.

15. The method of claim 14 , wherein the broadcast message is a beaconing message.

16. The method of claim 11 , wherein the first node is determined to have failed when the log includes a predetermined number of alerts received from a predetermined number of nodes within a predetermined period of time.

17. The method of claim 11 , wherein the processor determines whether the first node has failed based on one or more policies.

18. The method of claim 17 , wherein the processor determines whether the first node has failed based on a policy selected from a plurality of policies, said policy being dynamically selected by the processor based on the current condition of the network.

19. The system of claim 11 , wherein the failed node is removed from the log.

20. The method of claim 11 , comprising:

receiving a failure alert from at least one third node;

recording the failure alert received from the at least one third node in the log in association with the failure alert received from the second node; and

determining that the first node has failed by determining a number of failure alerts for the first node received from the second node in relation to a number of failure alerts received from the at least one third node.