System for Estimating Download Speed From Passive Measurements

1. A system for passive estimation of throughput, the system comprising: a memory that stores instructions; a processor that executes the instructions to perform operations, the operations comprising: annotating a flow record for each data flow of a plurality of data flows to include an application field and a content provider field, wherein the application field indicates an application protocol and the content provider field indicates a content provider with which each data flow is in communication; determining a flow type of each data flow based on the application field and the content provider field of the flow record; selecting each data flow for which the flow type is determined to have a non-rate-limited flow type; and estimating an average maximum throughput for each data flow selected.

2. The system of claim 1 , wherein the operations further comprise generating a throughput index, wherein the throughput index comprises the flow type determined for each data flow if the flow type is determined to have the non-rate-limited flow type.

3. The system of claim 2 , wherein the operations further comprise selecting, from the throughput index, each data flow for which the flow type is determined to have the non-rate-limited flow type.

4. The system of claim 2 , wherein the operations further comprise rejecting, from the throughput index, each data flow determined to not have the non-rate-limited flow type.

5. The system of claim 1 , wherein the operations further comprise determining if each data flow of the plurality of data flows has a flow size for achieving a steady-state throughput.

6. The system of claim 5 , wherein the operations further comprise excluding each data flow determined not to have the flow size for achieving the steady-state throughput when estimating the average maximum throughput.

7. The system of claim 1 , wherein the flow type is selected from the group comprising a rate-capped flow type, a partially rate-limited flow type, and the non-rate-limited flow type.

8. The system of claim 1 , wherein the operations further comprise validating the average maximum throughput estimated for each data flow selected by comparing the average maximum throughput to a set of active measurements measured in a network.

9. The system of claim 1 , wherein the operations further comprise recording the set of active measurements measured in the network by utilizing probes positioned along various locations in the network.

10. The system of claim 1 , wherein the operations further comprise accessing the flow record for each data flow of the plurality of data flows during a predetermined time interval.

11. The system of claim 10 , wherein the operations further comprise adjusting the predetermined time interval, and wherein the operations further comprise estimating the average maximum throughput for each data flow selected based on the adjusted predetermined time interval.

12. A method for passive estimation of throughput, the method comprising: annotating a flow record for each data flow of a plurality of data flows to include an application field and a content provider field, wherein the application field indicates an application protocol and the content provider field indicates a content provider with which each data flow is in communication; determining a flow type of each data flow based on the application field and the content provider field of the flow record; selecting each data flow for which the flow type is determined to have a non-rate-limited flow type; and estimating, by utilizing instructions from memory that are executed by a processor, an average maximum throughput for each data flow selected.

13. The method of claim 12 , further comprising validating the average maximum throughput estimated for each data flow selected by comparing the average maximum throughput to a set of active measurements measured in a network.

14. The method of claim 13 , further comprising recording the set of active measurements measured in the network by utilizing probes positioned along various locations in the network.

15. The method of claim 12 , further comprising determining if each data flow of the plurality of data flows has a flow size for achieving a steady-state throughput by determining if enough bytes were transmitted in each data flow to exit a slow-start phase during transmission of each data flow.

16. The method of claim 15 , further comprising excluding each data flow determined not to have the flow size for achieving the steady-state throughput when estimating the average maximum throughput.

17. The method of claim 12 , further comprising generating a throughput index, wherein the throughput index comprises the flow type determined for each data flow if the flow type is determined to have the non-rate-limited flow type.

18. The method of claim 17 , further comprising rejecting, from the throughput index, each data flow determined to not have the non-rate-limited flow type.

19. The method of claim 12 , further comprising storing the flow record for each data flow of the plurality of data flows during a predetermined time interval.

20. A computer-readable device comprising instructions, which, when loaded and executed by a processor, cause the process to perform operations comprising: annotating a flow record for each data flow of a plurality of data flows to include an application field and a content provider field, wherein the application field indicates an application protocol and the content provider field indicates a content provider with which each data flow is in communication; determining a flow type of each data flow based on the application field and the content provider field of the flow record; selecting each data flow for which the flow type is determined to have a non-rate-limited flow type; and estimating an average maximum throughput for each data flow selected.