Computing statistics for mixture distributions
A computing device includes a processor and memory storing instructions that are executable to determine a median of a first mixture distribution. The instructions are also executable to determine a parent mean, a parent standard deviation, and boundaries for each of multiple segments in the first mixture distribution. The instructions are also executable to determine a segment mean and a segment second moment for each segment based on the parent mean, the parent standard deviation, and the boundaries for the respective segment. The instructions are also executable to determine a scaled probability for each segment. The instructions are also executable to determine a mixture mean and a mixture standard deviation for the first mixture distribution based on the segment mean, the segment second moment, and the scaled probability for each segment in the first mixture distribution.
1. A computing device for computing statistics for a first mixture distribution, comprising:
a processor;
memory in electronic communication with the processor; and
instructions stored in a memory, the instructions being executable to:
determine a median of the first mixture distribution;
determine a parent mean, a parent standard deviation, and boundaries for each of multiple segments in the first mixture distribution;
determine a segment mean and a segment second moment for each segment based on the parent mean, the parent standard deviation, and the boundaries for the respective segment;
determine a scaled probability for each segment; and
determine a mixture mean and a mixture standard deviation for the first mixture distribution based on the segment mean, the segment second moment, and the scaled probability for each segment in the first mixture distribution.
2. The computing device of claim 1 , wherein the median of the first mixture distribution and the parent mean, the parent standard deviation, and the boundaries for each of the segments are determined using an irregular terrain model (ITM).
3. The computing device of claim 1 , wherein the scaled probability for a particular segment is determined based on a ratio of an area under a curve of the particular segment relative to a parent distribution lower bound and a parent distribution upper bound.
4. The computing device of claim 1 , wherein the instructions being executable to determine the mixture mean and the mixture standard deviation comprise instructions being executable to:
determine the mixture mean by summing a plurality of probability-scaled segment means, wherein each probability-scaled segment mean is a segment mean scaled by its corresponding scaled probability.
5. The computing device of claim 4 , wherein the instructions being executable to determine the mixture mean and the mixture standard deviation further comprise instructions being executable to:
determine a mixture variance by computing a difference of a summed plurality of probability-scaled segment second moments and the summed probability-scaled segment means.
6. The computing device of claim 1 , further comprising instructions being executable to:
determine a mixture mean and a mixture standard deviation for each of a plurality of other mixture distributions;
determine an aggregate linear mean and an aggregate linear variance for an aggregate distribution based on the mixture means and the mixture standard variations for the first and other mixture distributions; and
determine a log-scale 95th percentile value for the aggregate distribution based on the aggregate linear mean and the aggregate linear variance.
7. The computing device of claim 6 , wherein the instructions being executable to determine the aggregate linear mean and the aggregate linear variance comprise instructions being executable to:
determine the aggregate linear mean as a linear sum of the mixture means; and
determine the aggregate linear variance as the linear sum of the mixture variances.
8. The computing device of claim 6 , wherein the instructions being executable to determine the aggregate linear mean and the aggregate linear variance comprise instructions being executable to:
convert the aggregate linear mean to a log-scale mean;
convert the aggregate linear variance to a log-scale variance; and
determine the log-scale 95th percentile value by determining a 95th percentile value of a normal distribution with a mean of the log-scale mean and a variance of the log-scale variance.
9. The computing device of claim 6 , wherein the computing device is a spectrum access system (SAS) in a Citizens Broadband Radio Service (CBRS) system.
10. The computing device of claim 9 , wherein the SAS uses the log-scale 95th percentile value for the aggregate distribution to determine a subset of Citizens Broadband Radio Service devices (CBSDs) to remove from a communication channel.
11. A method for computing statistics for a first mixture distribution, comprising:
determining a median of the first mixture distribution;
determining a parent mean, a parent standard deviation, and boundaries for each of multiple segments in the first mixture distribution;
determining a segment mean and a segment second moment for each segment based on the parent mean, the parent standard deviation, and the boundaries for the respective segment;
determining a scaled probability for each segment; and
determining a mixture mean and a mixture standard deviation for the first mixture distribution based on the segment mean, the segment second moment, and the scaled probability for each segment in the first mixture distribution.
12. The method of claim 11 , wherein the median of the first mixture distribution and the parent mean, the parent standard deviation, and the boundaries for each of the segments are determined using an irregular terrain model (ITM).
13. The method of claim 11 , wherein the scaled probability for a particular segment is determined based on a ratio of an area under a curve of the particular segment relative to a parent distribution lower bound and a parent distribution upper bound.
14. The method of claim 11 , wherein the determining the mixture mean and the mixture standard deviation comprises:
determining the mixture mean by summing a plurality of probability-scaled segment means, wherein each probability-scaled segment mean is a segment mean scaled by its corresponding scaled probability.
15. The method of claim 14 , wherein determining the mixture mean and the mixture standard deviation further comprises:
determining a mixture variance by computing a difference of a summed plurality of probability-scaled segment second moments and the summed probability-scaled segment means.
16. The method of claim 11 , further comprising:
determining a mixture mean and a mixture standard deviation for each of a plurality of other mixture distributions;
determining an aggregate linear mean and an aggregate linear variance for an aggregate distribution based on the mixture means and the mixture standard variations for the first and other mixture distributions; and
determining a log-scale 95th percentile value for the aggregate distribution based on the aggregate linear mean and the aggregate linear variance.
17. The method of claim 16 , wherein determining the aggregate linear mean and the aggregate linear variance comprises:
determining the aggregate linear mean as a linear sum of the mixture means; and
determining the aggregate linear variance as the linear sum of the mixture variances.
18. The method of claim 16 , wherein determining the aggregate linear mean and the aggregate linear variance comprises:
converting the aggregate linear mean to a log-scale mean;
converting the aggregate linear variance to a log-scale variance; and
determining the log-scale 95th percentile value by determining a 95th percentile value of a normal distribution with a mean of the log-scale mean and a variance of the log-scale variance.
19. The method of claim 16 , wherein the method is performed by a spectrum access system (SAS) in a Citizens Broadband Radio Service (CBRS) system.
20. The method of claim 19 , wherein the SAS uses the log-scale 95th percentile value for the aggregate distribution to determine a subset of Citizens Broadband Radio Service devices (CBSDs) to remove from a communication channel.