Method for correlating congestion to performance metrics in internet traffic
View Patent ↗For each of a number of network performance metrics, an associated value r pm that represents a difference between a first correlation coefficient r 1 computed for a baseline data set and a second correlation coefficient r 2 computed for a second data set that includes the baseline data set and other data points classified as duration outliers is computed. The first and second correlation coefficients for each network performance metric represent correlation between that network performance metric and durations of network connections. The network performance metric that has a largest associated r pm value of all r pm values computed is selected as representing the probable root cause of the duration outliers.
1. A computer based method, comprising:
for each of a number of network performance metrics collected by a network monitor, computing an associated value r pm that represents a difference between a first correlation coefficient r 1 computed by the network monitor for a baseline data set and a second correlation coefficient r 2 computed by the network monitor for a second data set that includes the baseline data set and other data points classified as duration outliers, the first and second correlation coefficients for each network performance metric representing correlation between that network performance metric and durations of network connections; and
selecting that network performance metric that has a largest associated r pm value of all r pm values computed as representing a root cause of the duration outliers.
2. The method of claim 1 wherein the network performance metrics include one or more of round trip time, server response time, packet loss and file size.
3. The method of claim 1 wherein the duration outliers are classified using one or more of Grubb's Test, Rosner's Test and Walsh's Test.
4. The method of claim 1 wherein the duration outliers are classified using information from the network performance metrics.
5. The method of claim 4 wherein the network performance metrics include one or more of round trip time, server response time, packet loss and file size.
6. A computer based method, comprising:
for a first set of data points regarding Internet traffic collected by a network monitor, each data point having a number of associated parameters, eliminating outliers as determined according to statistical properties of at least one of the associated parameters from the set of data points collected by the network monitor to produce a baseline;
computing, for each of one or more of the associated parameters, a first correlation coefficient r 1 between connection duration and that associated parameter for the baseline;
computing, for each of the one or more of the associated parameters, a second correlation coefficient r 2 between connection duration and that associated parameter for the first set of data points;
computing, for each of the one or more of the associated parameters, a difference r pm between r 2 and r 1 for that associated parameter; and
selecting a corresponding one of the associated parameters that has a largest r pm value of all r pm values computed as representing a root cause of the outliers.
7. The method of claim 6 wherein the associated parameters include file size, server response time, packet loss and latency.
8. The method of claim 7 wherein the at least one of the associated parameters comprises connection duration.
9. The method of claim 6 wherein the outliers are determined using one or more of Grubb's Test, Rosner's Test and Walsh's Test.
10. The method of claim 6 wherein the outliers are determined using information from the associated parameters.