Calculating mean wireless signal strengths using a gaussian process approach incorporating predictive standard deviations
Disclosed are apparatus and methods for providing outputs; e.g., location estimates, based on measurement bins (MBs) determined by a computing device. An MB can be associated with a wireless signal emitter (WSE), and can include a mean signal strength value (SSV) and a standard deviation of SSVs for each WSE associated with the MB. The computing device can designate a WSE. The computing device can determine a collection of the MBs associated with the designated WSE. The computing device can train a mean Gaussian process for the designated WSE based on the mean SSV and the standard deviation of SSVs of the collection of MBs. The mean Gaussian process can be associated with a covariance matrix having a diagonal entry based on a standard deviation of SSVs of an MB in the collection of MBs. The computing device can provide an estimated location based on the trained mean Gaussian process.
1. A method, comprising:
determining by a computing device a plurality of measurement bins, wherein a particular measurement bin of the plurality of measurement bins is associated with one or more wireless signal emitters, and wherein the particular measurement bin comprises a mean signal strength value and a standard deviation of signal strength values for each wireless signal emitter of the one or more wireless signal emitters associated with the particular measurement bin;
determining by the computing device a designated wireless signal emitter;
determining by the computing device a collection of measurement bins of the plurality of measurement bins, wherein a particular measurement bin in the collection of measurement bins is associated with the designated wireless signal emitter;
training by the computing device a mean Gaussian process to model signals emitted by the designated wireless signal emitter based on the mean signal strength values of the collection of measurement bins and the standard deviation of signal strength values of the collection of measurement bins, wherein the mean Gaussian process is associated with a covariance matrix, and wherein a particular diagonal entry of the covariance matrix is based upon a standard deviation of signal strength values of a corresponding measurement bin in the collection of measurement bins; and
providing by the computing device an estimated location based on the trained mean Gaussian process.
2. The method of claim 1 , wherein providing the estimated location comprises:
receiving a request related to locating a mobile device;
determining by the computing device an estimate of the location of the mobile device based on the trained mean Gaussian process;
generating by the computing device the estimated location comprising the estimate of the location; and
providing the estimated location.
3. The method of claim 1 , further comprising:
training, by the computing device, a standard deviation Gaussian process for the designated wireless signal emitter based on the mean signal strength values of the collection of measurement bins and the standard deviation of signal strength values of the collection of measurement bins.
4. The method of claim 1 , wherein each non-diagonal entry of the covariance matrix is equal to zero.
5. The method of claim 1 , further comprising:
providing by the computing device a representation of the trained mean Gaussian process.
6. The method of claim 1 , further comprising:
determining, by the computing device, whether at least one bin in the collection of measurement bins is modified; and
after determining that at least one bin in the collection of measurement bins is modified, training a modified trained mean Gaussian process for the designated wireless signal emitter based on the mean signal strength values of the collection of measurement bins including the modified measurement bin and the standard deviation of signal strength values of the collection of measurement bins.
7. The method of claim 6 , further comprising:
determining, by the computing device, whether a difference between the trained mean Gaussian process and the modified trained mean Gaussian process exceeds a threshold; and
after determining that the difference between the trained mean Gaussian process and the modified trained mean Gaussian process exceeds the threshold, providing a second output based on the modified trained mean Gaussian process.
8. The method of claim 6 , wherein the trained mean Gaussian process is associated with a confidence value, and wherein the method further comprises:
determining whether a difference between the trained mean Gaussian process and the modified trained mean Gaussian process exceeds a threshold; and
after determining that the difference between the trained mean Gaussian process and the modified trained mean Gaussian process does not exceed the threshold, increasing the confidence value associated with the trained mean Gaussian process.
9. A computing device, comprising:
one or more processors; and
data storage, configured to store at least computer-readable program instructions, wherein the instructions are configured to, upon execution by the one or more processors, cause the computing device to perform functions comprising:
determining a plurality of measurement bins, wherein a particular measurement bin of the plurality of measurement bins is associated with one or more wireless signal emitters, and wherein the particular measurement bin comprises a mean signal strength value and a standard deviation of signal strength values for each wireless signal emitter of the one or more wireless signal emitters associated with the particular measurement bin;
determining a designated wireless signal emitter;
determining a collection of measurement bins of the plurality of measurement bins, wherein a particular measurement bin in the collection of measurement bins is associated with the designated wireless signal emitter;
training a mean Gaussian process to model signals emitted by the designated wireless signal emitter based on the mean signal strength values of the collection of measurement bins and the standard deviation of signal strength values of the collection of measurement bins, wherein the mean Gaussian process is associated with a covariance matrix, and wherein a particular diagonal entry of the covariance matrix is based upon a standard deviation of signal strength values of a corresponding measurement bin in the collection of measurement bins; and
providing an estimated location based on the trained mean Gaussian process.
10. The computing device of claim 9 , wherein providing the estimated location comprises:
receiving a request related to locating a mobile device;
determining an estimate of the location of the mobile device based on the trained mean Gaussian process;
generating the estimated location comprising the estimate of the location; and
providing the estimated location.
11. The computing device of claim 9 , wherein the functions further comprise:
training a standard deviation Gaussian process for the designated wireless signal emitter based on the mean signal strength values of the collection of measurement bins and the standard deviation of signal strength values of the collection of measurement bins.
12. The computing device of claim 9 , wherein each non-diagonal entry of the covariance matrix is equal to zero.
13. The computing device of claim 9 , wherein the functions further comprise:
providing a representation of the trained mean Gaussian process.
14. The computing device of claim 9 , wherein the functions further comprise:
determining whether at least one bin in the collection of measurement bins is modified; and
after determining that at least one bin in the collection of measurement bins is modified, training a modified trained mean Gaussian process for the designated wireless signal emitter based on the mean signal strength values of the collection of measurement bins including the modified measurement bin and the standard deviation of signal strength values of the collection of measurement bins.
15. The computing device of claim 14 , wherein the functions further comprise:
determining whether a difference between the trained mean Gaussian process and the modified trained mean Gaussian process exceeds a threshold; and
after determining that the difference between the trained mean Gaussian process and the modified trained mean Gaussian process exceeds the threshold, providing a second output based on the modified trained mean Gaussian process.
16. The computing device of claim 14 , wherein the trained mean Gaussian process is associated with a confidence value, and wherein the functions further comprise:
determining whether a difference between the trained mean Gaussian process and the modified trained mean Gaussian process exceeds a threshold; and
after determining that the difference between the trained mean Gaussian process and the modified trained mean Gaussian process does not exceed the threshold, increasing the confidence value associated with the trained mean Gaussian process.
17. An article of manufacture including a computer-readable storage medium having instructions stored thereon that, in response to execution by one or more processors, cause the one or more processors to perform functions comprising:
determining a plurality of measurement bins, wherein a particular measurement bin of the plurality of measurement bins is associated with one or more wireless signal emitters, and wherein the particular measurement bin comprises a mean signal strength value and a standard deviation of signal strength values for each wireless signal emitter of the one or more wireless signal emitters associated with the particular measurement bin;
determining a designated wireless signal emitter;
determining a collection of measurement bins of the plurality of measurement bins, wherein a particular measurement bin in the collection of measurement bins is associated with the designated wireless signal emitter;
training a mean Gaussian process to model signals emitted by the designated wireless signal emitter based on the mean signal strength values of the collection of measurement bins and the standard deviation of signal strength values of the collection of measurement bins, wherein the mean Gaussian process is associated with a covariance matrix, and wherein a particular diagonal entry of the covariance matrix is based upon a standard deviation of signal strength values of a corresponding measurement bin in the collection of measurement bins; and
providing an estimated location based on the trained mean Gaussian process.
18. The article of manufacture of claim 17 , wherein providing the estimated location comprises:
receiving a request related to locating a mobile device;
determining an estimate of the location of the mobile device based on the trained mean Gaussian process;
generating the estimated location comprising the estimate of the location; and
providing the estimated location.
19. The article of manufacture of claim 17 , wherein the functions further comprise:
training a standard deviation Gaussian process for the designated wireless signal emitter based on the mean signal strength values of the collection of measurement bins and the standard deviation of signal strength values of the collection of measurement bins.
20. The article of manufacture of claim 17 , wherein each non-diagonal entry of the covariance matrix is equal to zero.