Techniques for establishing a communication handoff threshold using user feedback

1. At least one machine-readable non-transitory medium comprising a set of instructions that in response to being executed on a computing device cause the computing device to:

receive user feedback comprising perceived VoIP audio quality correlated with a plurality of VoIP communication link parameters for an ongoing packet data communication link between a hybrid mobile device and a call server; and

fit a statistical model to the user feedback training data set, the statistical model operative to instantaneously estimate an outcome associated with VoIP audio quality when fit to a subsequent data set of VoIP communication link parameters wherein the outcome of fitting the statistical model to subsequent data sets of VoIP communication link parameters yields either acceptable or poor VoIP audio quality,

wherein the VoIP communication link parameters comprising at least two of: a received signal strength (RSSI) between the hybrid mobile device and an IP access point, a link speed between the hybrid mobile device and an IP access point, jitter, latency, the number of dropped packets in a packet data stream, the number of holes in a packet data stream (num holes), the average hole size, the maximum hole size, and transformed parameters.

2. The at least one non-transitory machine-readable medium of claim 1 , comprising instructions that in response to being executed on the computing device cause the computing device to:

create transformed VoIP communication link parameters by applying mathematical functions to one or more VoIP communication link parameters, the mathematical functions including translation, negation, summation, multiplication, power functions, logarithms, exponentiation, and step functions such that the resulting transformed parameters are not identical to any individual VoIP communication link parameter; and

re-determine the statistical model to include the transformed VoIP communication link parameters.

3. The at least one machine-readable non-transitory medium of claim 1 , comprising instructions that in response to being executed on the computing device cause the computing device to:

shift the received user feedback backwards along the packet data stream by a pre-determined time interval to compensate for a lag between a user's observation of the VoIP communication link parameters and the user's sending of the observed VoIP communication link parameters.

4. An apparatus, comprising:

a processor component;

communication link handoff (CLH) logic to be executed by the processor component to:

receive user feedback comprising perceived VoIP audio quality correlated with a plurality of VoIP communication link parameters for an ongoing packet data communication link between a hybrid mobile device and a call server; and

fit a statistical model to the user feedback training data set, the statistical model operative to instantaneously estimate an outcome associated with VoIP audio quality when fit to a subsequent data set of VoIP communication link parameters wherein the outcome of fitting the statistical model to subsequent data sets of VoIP communication link parameters yields either acceptable or poor VoIP audio quality;

wherein the VoIP communication link parameters comprising at least two of: a received signal strength (RSSI) between the hybrid mobile device and an IP access point, a link speed between the hybrid mobile device and an IP access point, jitter, latency, the number of dropped packets in a packet data stream, the number of holes in a packet data stream (num holes), the average hole size, the maximum hole size, and transformed parameters.

5. The apparatus of claim 4 , the CLH logic executed by the processor component further operative to:

create transformed VoIP communication link parameters by applying mathematical functions to one or more VoIP communication link parameters, the mathematical functions including translation, negation, summation, multiplication, power functions, logarithms, exponentiation, and step functions such that the resulting transformed parameters are not identical to any individual VoIP communication link parameter; and

re-determine the statistical model to include the transformed VoIP communication link parameters.

6. The apparatus of claim 5 , the statistical model comprising a logistic regression model operative to:

determine a coefficient for each of the individual and/or transformed VoIP communication link parameters; and

calculate an audio quality threshold raw score comprised of the sum of each individual and/or transformed VoIP communication link parameter multiplied by its respective coefficient.

7. The apparatus of claim 4 , the communication link handoff (CLH) logic to be executed by the processor component to:

shift the received user feedback backwards along the packet data stream by a pre-determined time interval to compensate for a lag between a user's observation of the VoIP communication link parameters and the user's sending of the observed VoIP communication link parameters.

8. A computer implemented method operative on a processor component, the method comprising:

receiving user feedback comprising perceived VoIP audio quality correlated with a plurality of VoIP communication link parameters for an ongoing packet data communication link between a hybrid mobile device and a call server; and

fitting a statistical model to the user feedback training data set, the statistical model operative to instantaneously estimate an outcome associated with VoIP audio quality when fit to a subsequent data set of VoIP communication link parameters wherein the outcome of fitting the statistical model to subsequent data sets of VoIP communication link parameters yields either acceptable or poor VoIP audio quality;

wherein the VoIP communication link parameters comprising at

least two of: a received signal strength (RSSI) between the hybrid mobile device and an IP access point, a link speed between the hybrid mobile device and an IP access point, jitter, latency, the number of dropped packets in a packet data stream, the number of holes in a packet data stream (num holes), the average hole size, the maximum hole size, and transformed parameters.

9. The method of claim 8 , wherein the statistical model comprises a logistic regression model and further comprising:

creating transformed VoIP communication link parameters by applying mathematical functions to one or more VoIP communication link parameters, the mathematical functions including translation, negation, summation, multiplication, power functions, logarithms, exponentiation, and step functions such that the resulting transformed parameters are not identical to any individual VoIP communication link parameter; and

determining the logistic regression model to include the transformed VoIP communication link parameters.

10. The method of claim 9 , further comprising:

determining a coefficient for each of the individual and/or transformed VoIP communication link parameters; and

calculating an audio quality threshold raw score comprised of the sum of each individual and/or transformed VoIP communication link parameter multiplied by its respective coefficient.

11. The method of claim 8 , further comprising:

shifting the received user feedback backwards along the packet data stream by a pre-determined time interval to compensate for a lag between a user's observation of the VoIP communication link parameters and the user's sending of the observed VoIP communication link parameters.