Interference management in a multi-hop wireless network

1. A method comprising, by an interference-management system of a multi-hop wireless network:

accessing an interference map indicating interference among network nodes of the multi-hop wireless network;

identifying a plurality of links between network nodes, wherein one or more of the plurality of links are identified as golden links;

generating a factor-graph representation comprising a set of first vertices corresponding to a respective set of second vertices, wherein each vertex pair is associated with an identified link, wherein, for each vertex pair, the first vertex is a variable node representing beamforming weight variables associated with the identified link and each second vertex is a function node representing a capacity equation associated with the identified link, and wherein each vertex pair is assigned a capacity weight from a set of non-equal capacity weights based at least in part on the identified golden links; and

determining one or more adjustments to one or more beamforming weights to reduce interference among the network nodes.

2. The method of claim 1 , wherein each variable node further represents a signal phase shift variable associated with the identified link.

3. The method of claim 1 , wherein each variable node further represents a time division multiplexing variable associated with the identified link.

4. The method of claim 1 , wherein the one or more golden links are identified based on a proximity to a vital network node.

5. The method of claim 4 , wherein the vital network node connects the multi-hop wireless network to a backbone network.

6. The method of claim 1 , wherein, for one or more of the identified links, the one or more adjustments are determined at least in part based on one of more messages passed from the first vertex to the second vertex.

7. The method of claim 6 , wherein, for each of the one or more identified links, the one or more messages are derived at least in part based on one or more messages previously passed from the first vertex to the second vertex.

8. The method of claim 7 , wherein a set of possible values for each variable node is identified based on the one or more previously passed messages.

9. The method of claim 8 , wherein the set of possible values for each variable node corresponds to a set of micro-routes and nano-routes on the multi-hop wireless network.

10. The method of claim 9 , wherein the set of possible values for each variable node comprises two, three, or four possible values.

11. The method of claim 1 , wherein the one or more adjustments to the one or more beamforming weights are determined by applying a belief-propagation algorithm to a portion of the factor-graph representation.

12. The method of claim 11 , wherein the determined adjustments optimize a sum-capacity of the multi-hop wireless network.

13. The method of claim 11 , wherein the determined adjustments optimize a sum-capacity of a portion of the multi-hop wireless network associated with the portion of the factor-graph representation.

14. The method of claim 11 , wherein the determined adjustments comprise adjusting a signal phase shift of the identified link.

15. The method of claim 11 , wherein the determined adjustments comprise a time division multiplexing scheme.

16. The method of claim 15 , wherein the time division multiplexing scheme is configured to permit one or more sectors of one or more of the network nodes to transmit or receive during designated time slots in a time-slot-exclusion schedule.

17. The method of claim 11 , wherein the determined adjustments exclude one or more micro-routes on the multi-hop wireless network.

18. The method of claim 11 , further comprising:

sending at least one of the determined adjustments to at least one of the network nodes of the multi-hop wireless network.

19. A system comprising: one or more processors; and a memory coupled to the processors comprising instructions executable by the processors, the processors operable when executing the instructions to:

access an interference map indicating interference among network nodes of the multi-hop wireless network;

identify a plurality of links between network nodes, wherein one or more of the plurality of links are identified as golden links;

generate a factor-graph representation comprising a set of first vertices corresponding to a respective set of second vertices, wherein each vertex pair is associated with an identified link, wherein, for each vertex pair, the first vertex is a variable node representing beamforming weight variables associated with the identified link and each second vertex is a function node representing a capacity equation associated with the identified link, and wherein each vertex pair is assigned a capacity weight from a set of non-equal capacity weights based at least in part on the identified golden links; and

determine one or more adjustments to one or more beamforming weights to reduce interference among the network nodes.

20. One or more computer-readable non-transitory storage media embodying software that is operable when executed to:

access an interference map indicating interference among network nodes of the multi-hop wireless network;

identify a plurality of links between network nodes, wherein one or more of the plurality of links are identified as golden links;

generate a factor-graph representation comprising a set of first vertices corresponding to a respective set of second vertices, wherein each vertex pair is associated with an identified link, wherein, for each vertex pair, the first vertex is a variable node representing beamforming weight variables associated with the identified link and each second vertex is a function node representing a capacity equation associated with the identified link, and wherein each vertex pair is assigned a capacity weight from a set of non-equal capacity weights based at least in part on the identified golden links; and

determine one or more adjustments to one or more beamforming weights to reduce interference among the network nodes.