Assigning a subset of access points in a wireless network to a high priority

1. A method comprising:

assigning, by a network device, a plurality of client devices in a wireless network into a number of clusters;

calculating, by the network device, a cluster centroid for each cluster of client devices based on locations of the client devices;

calculating, by the network device, an average distance between each client device in a particular cluster and the original cluster centroid location for the particular cluster;

determining, by the network device, that the number of clusters is associated with a low average distance from each client device in a respective cluster to a respective cluster centroid;

including, by the network device, an access point located closest to the respective cluster centroid in a subset of access points in the wireless network; and

assigning, by the network device, the subset of access points in the wireless network to a high priority.

2. The method of claim 1 , further comprising:

assigning, by the network device, each client device in the plurality of client devices to a cluster whose cluster centroid location is closest to the location of the each client device.

3. The method of claim 1 , further comprising:

determining, by a network device, locations of a plurality of client devices within the wireless network;

assigning, by the network device, the plurality of client devices into the clusters; and

iteratively adjusting, by the network device, the number of clusters and assignment of the plurality of client devices to determine a low number of cluster and a low average distance from each client device in a respective cluster to a respective cluster centroid location; and

calculating, by the network device, an original cluster centroid location for each cluster of client devices.

4. The method of claim 3 , wherein iteratively adjusting the number of clusters and assignment of the plurality of client devices further comprises:

iteratively adjusting, by the network device, the number of clusters and the assignment of the plurality of client devices using K-means clustering.

5. The method of claim 1 , wherein assigning the plurality of client devices in the wireless network into the clusters further comprising:

randomly selecting a subset of client devices from the plurality of client devices, wherein a total number of selected client devices equals to the number of clusters;

associating each of remaining client devices to one client device in the subset that is closest in distance to each respective remaining client device; and

for each client device in the selected subset, identifying associated remaining client devices as belonging to the same cluster of client devices.

6. The method of claim 5 , further comprising:

giving each client device in the plurality of client devices a weight value based on a client parameter; and

in response to a respective client device having a high weight value, selecting the respective client device frequently into the subset of client devices.

7. The method of claim 6 , further comprising:

determining a closest cluster centroid location for each of the plurality of client devices;

associating each respective client device to a new cluster whose centroid location has closest distance to the each respective client device; and

calculating a new cluster centroid location based on client devices associated to the new cluster.

8. The method of claim 7 , further comprising:

determining whether a distance between a respective new cluster centroid location and original cluster centroid location is less than a predetermined threshold; and

in response to a distance between a respective new cluster centroid location and original cluster centroid location being less than a predetermined threshold, assigning the high priority to the subset of access points having closest distance to each new cluster centroid location.

9. The method of claim 1 , wherein the subset of access points are (1) associated with a high priority traffic profile, (2) associated with a high priority Quality-of-Service (QoS) profile, and (3) given priority for selecting their operating channels.

10. A network device, comprising at least:

a memory;

a processor executing instructions from the memory to:

assign a plurality of client devices in a wireless network into a number of clusters;

calculate a cluster centroid for each cluster of client devices based on locations of the client devices;

calculate an average distance between each client device in a particular cluster and the original cluster centroid location for the particular cluster;

determine that the number of clusters is associated with a low average distance from each client device in a respective cluster to a respective cluster centroid;

include an access point located closest to the respective cluster centroid in a subset of access points in the wireless network; and

assign the subset of access points in the wireless network to a high priority.

11. The network device of claim 10 , wherein the processor further executes the instructions from the memory to:

assign each client device in the plurality of client devices to a cluster whose cluster centroid location is closest to the location of the each client device.

12. The network device of claim 10 , wherein the processor further executes the instructions from the memory to:

determine locations of a plurality of client devices within the wireless network;

assign the plurality of client devices into the clusters; and

iteratively adjust the number of clusters and assignment of the plurality of client devices to determine a low number of cluster and a low average distance from each client device in a respective cluster to a respective cluster centroid location; and

calculate an original cluster centroid location for each cluster of client devices.

13. The network device of claim 12 , wherein the processor further executes the instructions from the memory to:

iteratively adjust the number of clusters and the assignment of the plurality of client devices using K-means clustering.

14. The method of claim 10 , wherein the processor further executes the instructions from the memory to:

randomly select a subset of client devices from the plurality of client devices, wherein a total number of selected client devices equals to the number of clusters;

associate each of remaining client devices to one client device in the subset that is closest in distance to each respective remaining client device; and

for each client device in the selected subset, identify associated remaining client devices as belonging to the same cluster of client devices.

15. The network device of claim 14 , wherein the processor further executes the instructions from the memory to:

give each client device in the plurality of client devices a weight value based on a client parameter; and

in response to a respective client device having a high weight value, select the respective client device frequently into the subset of client devices.

16. The network device of claim 15 , wherein the processor further executes the instructions from the memory to:

determine a closest cluster centroid location for each of the plurality of client devices;

associate each respective client device to a new cluster whose centroid location has closest distance to the each respective client device; and

calculate a new cluster centroid location based on client devices associated to the new cluster.

17. The network device of claim 16 , wherein the processor further executes the instructions from the memory to:

determine whether a distance between a respective new cluster centroid location and original cluster centroid location is less than a predetermined threshold; and

in response to a distance between a respective new cluster centroid location and original cluster centroid location being less than a predetermined threshold, assign the high priority to the subset of access points having closest distance to each new cluster centroid location.

18. The network device of claim 10 , wherein the subset of access points are (1) associated with a high priority traffic profile, (2) associated with a high priority Quality-of-Service (QoS) profile, and (3) given priority for selecting their operating channels.

19. A non-transitory machine-readable storage medium encoded with instructions executable by at least one processor of a network device, the machine-readable storage medium comprising instructions to:

assign a plurality of client devices in a wireless network into a number of clusters;

calculate a cluster centroid for each cluster of client devices based on locations of the client devices;

calculate an average distance between each client device in a particular cluster and the original cluster centroid location for the particular cluster;

determine that the number of clusters is associated with a low average distance from each client device in a respective cluster to a respective cluster centroid;

include an access point located closest to the respective cluster centroid in a subset of access points in the wireless network; and

assign the subset of access points in the wireless network to a high priority.

20. The non-transitory machine-readable storage medium of claim 19 , wherein the machine-readable storage medium further comprises instructions to:

determine locations of a plurality of client devices within the wireless network;

assign the plurality of client devices into the clusters; and

iteratively adjust the number of clusters and assignment of the plurality of client devices to determine a low number of cluster and a low average distance from each client device in a respective cluster to a respective cluster centroid location; and

calculate an original cluster centroid location for each cluster of client devices;

determine a closest cluster centroid location for each of the plurality of client devices;

assign each respective client device to a new cluster whose centroid location has closest distance to the each respective client device;

calculate a new cluster centroid location based on client devices associated to the new cluster;

determine whether a distance between a respective new cluster centroid location and original cluster centroid location is less than a predetermined threshold; and

in response to a distance between a respective new cluster centroid location and original cluster centroid location being less than a predetermined threshold, assign the high priority to the subset of access points having closest distance to each new cluster centroid location.