Multi-hop wireless network planning

1. A method comprising, by one or more computing devices:

identifying, from a point cloud comprising a plurality of point-cloud points, a first set of point clusters, wherein the point cloud is generated by measuring the distance from an initial measuring point to each point-cloud point, wherein each point-cloud point corresponds to a location on a surface of an object located in a region in a three-dimensional space,

and wherein each point cluster comprises one or more point-cloud points that are located within a grid segment on a two-dimensional grid derived from the three-dimensional space;

determining, for at least a plurality of point clusters from the first set of point clusters, that each point cluster represents a vertical-linear structure or a portion of a vertical-linear structure; and

identifying one or more point-cluster pairs from the plurality of point clusters, wherein each point-cluster pair is identified based on the two point clusters having (1) a separation distance within a threshold distance in the three-dimensional space and

(2) a line-of-sight connection in the three-dimensional space.

2. The method of claim 1 , wherein determining that each point cluster of the plurality of point clusters represents a vertical-linear structure or a portion of a vertical-structure is done using Principle Component Analysis (PCA).

3. The method of claim 2 , wherein the PCA results in three eigenvalues λ 1 , λ 2 , λ 3 ,wherein:

a linear structure is characterized by λ 1 >>λ 2 ≅λ 3 ;

a planar structure is characterized by λ 1 ≅λ 2 >>λ 3 ; and

a volumetric structure is characterized by λ 1 ≅λ 2 ≅λ 3 .

4. The method of claim 1 , wherein a line-of-sight connection exists between two point clusters in a point-cluster pair if a number of point-cloud points between the two point clusters is below a threshold number.

5. The method of claim 1 , wherein the point cloud is obtained using light detection and ranging (LiDAR).

6. The method of claim 1 , further comprising:

for each point cluster determined to represent a vertical-linear structure or portion of the vertical-linear structure, applying one or more filter constraints to determine that the vertical-linear structure or portion of the vertical-linear structure is a utility pole or portion of a utility pole, respectively.

7. The method of claim 6 , further comprising:

retrieving one or more point-cloud points that are within a threshold distance of the vertical-linear structure or portion of the vertical-linear structure, wherein the one or more retrieved point-cloud points correspond to a location on a non-vertical component of the utility pole.

8. The method of claim 1 , further comprising:

selecting a plurality of point clusters for antenna sites for a multi-hop wireless network, the selection based at least in part on:

whether a line-of-sight connection exists between a respective point cluster and at least one other point cluster; and

one or more demand measures associated with the region in the three-dimensional space.

9. The method of claim 8 , wherein the one or more demand measures comprise a set of point clusters that have been determined to require a network antenna.

10. The method of claim 8 , wherein the one or more demand measures comprise an amount of traffic at one or more locations within the region in the three-dimensional space.

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

identify, from a point cloud comprising a plurality of point-cloud points, a first set of point clusters, wherein the point cloud is generated by measuring the distance from an initial measuring point to each point-cloud point, wherein each point-cloud point corresponds to a location on a surface of an object located in a region in a three-dimensional space,

and wherein each point cluster comprises one or more point-cloud points that are located within a grid segment on a two-dimensional grid derived from the three-dimensional space;

determine, for at least a plurality of point clusters from the first set of point clusters, that each point cluster represents a vertical-linear structure or a portion of a vertical-linear structure; and

identify one or more point-cluster pairs from the plurality of point clusters, wherein each point-cluster pair is identified based on the two point clusters having (1) a separation distance within a threshold distance in the three-dimensional space and (2) a line-of-sight connection in the three-dimensional space.

12. The media of claim 11 , wherein determining that each point cluster of the plurality of point clusters represents a vertical-linear structure or a portion of a vertical-structure is done using Principle Component Analysis (PCA).

13. The method of claim 2 , wherein the PCA results in three eigenvalues λ 1 , λ 2 , λ 3 ,wherein:

a linear structure is characterized by λ 1 >>λ 2 ≅λ 3 ;

a planar structure is characterized by λ 1 ≅λ 2 >>λ 3 ; and

a volumetric structure is characterized by λ 1 ≅λ 2 ≅λ 3 .

14. The media of claim 11 , wherein a line-of-sight connection exists between two point clusters in a point-cluster pair if a number of point-cloud points between the two point clusters is below a threshold number.

15. The media of claim 11 , wherein the point cloud is obtained using light detection and ranging (LiDAR).

16. The media of claim 11 , wherein the software is further operable when executed to:

for each point cluster determined to represent a vertical-linear structure or portion of the vertical-linear structure, apply one or more filter constraints to determine that the vertical-linear structure or portion of the vertical-linear structure is a utility pole or portion of a utility pole, respectively.

17. The media of claim 16 , wherein the software is further operable when executed to:

retrieve one or more point-cloud points that are within a threshold distance of the vertical-linear structure or portion of the vertical-linear structure, wherein the one or more retrieved point-cloud points correspond to a location on a non-vertical component of the utility pole.

18. The media of claim 11 , wherein the software is further operable when executing the instructions to:

select a plurality of point clusters for antenna sites for a multi-hop wireless network, the selection based at least in part on:

whether a line-of-sight connection exists between a respective point cluster and at least one other point cluster; and

one or more demand measures associated with the region in the three-dimensional space.

19. The media of claim 18 , wherein the one or more demand measures comprise a set of point clusters that have been determined to require a network antenna.

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

identify, from a point cloud comprising a plurality of point-cloud points, a first set of point clusters, wherein the point cloud is generated by measuring the distance from an initial measuring point to each point-cloud point, wherein each point-cloud point corresponds to a location on a surface of an object located in a region in a three-dimensional space,

and wherein each point cluster comprises one or more point-cloud points that are located within a grid segment on a two-dimensional grid derived from the three-dimensional space;

determine, for at least a plurality of point clusters from the first set of point clusters, that each point cluster represents a vertical-linear structure or a portion of a vertical-linear structure; and

identify one or more point-cluster pairs from the plurality of point clusters, wherein each point-cluster pair is identified based on the two point clusters having (1) a separation distance within a threshold distance in the three-dimensional space and

(2) a line-of-sight connection in the three-dimensional space.