Wireless system performance modeling

1. A method comprising:

modeling a performance of a physical layer of a network for a plurality of data modulation and coding schemes for a particular application program running on a computer and communicating data over the physical layer of the network, the computer including a processor, wherein the computer is in communication with the network, wherein said modeling comprises:

receiving performance data of said physical layer for said plurality of data modulation and coding schemes, wherein said performance data comprises a plurality of performance curves, wherein the plurality of performance curves comprises a plurality of sigmoidal curves, each of said plurality of sigmoidal curves representing a sigmoidal relationship between the throughput of said physical layer and a condition of a data transmission channel used by said physical layer for each of said plurality of data modulation and coding schemes, wherein said condition comprises a signal to interference plus noise ratio (“SINR”) of said data transmission channel, and

modeling the plurality of performance curves using a single sigmoid function to approximate an envelope formed by upper knees of said plurality of sigmoidal curves, and adapting said single sigmoid function to changing conditions of said data transmission channel; and

predicting the performance of said network for the particular application program based at least in part on the single sigmoid function based performance model of said physical layer by the processor.

2. The method of claim 1 , wherein predicting the performance of said network for an application comprises:

receiving the SINR of said data transmission channel used by said physical layer; and

estimating the performance of said physical layer under said SINR based on said single sigmoid function based performance model of said physical layer.

3. The method of claim 2 , wherein predicting the performance of said network for an application further comprises:

estimating the performance of an application layer of said network based at least in part on said estimated performance of said physical layer under said SINR; and

producing a data stream for transmitting said data required by said application at a rate according to the estimated application layer performance.

4. The method of claim 3 , wherein predicting the performance of said network for an application further comprises estimating the quality of said data required by said application, when received, based on said data stream.

5. The method of claim 3 , wherein estimating the performance of an application layer of said network comprises taking into account overheads added to said data required by said application by network layers higher than said physical layer.

6. A method comprising:

receiving performance data of a physical layer in a network for a plurality of data modulation and coding schemes for a particular application program running on a computer, wherein said performance data includes a plurality of sigmoidal curves for each of said plurality of data modulation and coding schemes, each of said plurality of sigmoidal curves representing a sigmoidal relationship between the throughput of said physical layer and a condition of a data transmission channel used by said physical layer for the corresponding one of said plurality of data modulation and coding schemes, wherein said performance data of said physical layer for a plurality of data modulation and coding schemes comprises the condition of the data transmission channel used by said physical layer for each of said plurality of data modulation and coding schemes, wherein said condition comprises a signal to interference plus noise ratio (“SINR”) of said data transmission channel; and

abstracting said performance data comprising modeling said plurality of sigmoidal curves using a single sigmoid function to approximate an envelope formed by upper knees of said plurality of sigmoidal curves, and adapting said single sigmoid function to changing conditions of said data transmission channel to model the performance of the particular application program communicating data over the physical layer of the network by a processor in communication with the network.

7. The method of claim 6 , wherein abstracting said performance data further comprises obtaining parameters of said sigmoid function based at least in part on said envelope.

8. An apparatus comprising:

a physical layer (PHY) performance modeling mechanism configured to:

receive performance data of said physical layer for a plurality of data modulation and coding schemes, wherein said performance data comprises a plurality of performance curves, wherein said plurality of performance curves comprises a plurality of sigmoidal curves, each of said plurality of sigmoidal curves representing a sigmoidal relationship between the throughput of said physical layer and a condition of a data transmission channel used by said physical layer for each of said plurality of data modulation and coding schemes, wherein said condition comprises a signal to interference plus noise ratio (“SINR”) of said data transmission channel, and

model the performance for a particular application program running on a computer and communicating data over said physical layer based on modeling the plurality of performance curves by using a single sigmoid function to approximate an envelope formed by upper knees of said plurality of sigmoidal curves, and adapting said single sigmoid function to changing conditions of said data transmission channel; and

a quality estimator configured to predict the performance of said network for the particular application program based at least in part on the obtained single sigmoid function based performance model of said physical layer.

9. The apparatus of claim 8 , further comprising a PHY performance estimator to receive the SINR of said data transmission channel used by said physical layer, and to estimate the performance of said physical layer under said SINR based on said sigmoid function based performance model of said physical layer.

10. The apparatus of claim 9 , further comprising an application performance estimator to estimate the performance of an application layer of said network based at least in part on said estimated performance of said physical layer under said SINR, and to produce a data stream for transmitting said data required by said application at a rate according to the estimated application layer performance.

11. The apparatus of claim 10 , wherein said application performance estimator estimates the performance of an application layer of said network by taking into account overheads added to said data required by said application by network layers higher than said physical layer.

12. The apparatus of claim 10 , wherein said quality estimator predicts the performance of said network for an application by estimating the quality of said data required by said application, when received, based on said data stream.

13. An article comprising:

a non-transitory machine readable medium that contains instructions, which when executed by a processing platform, cause said processing platform to perform operations comprising:

modeling a performance of a physical layer of a network for a plurality of data modulation and coding schemes for a particular application program running on a computer and communicating data over the physical layer of the network, wherein said modeling comprises:

receiving performance data of said physical layer for said plurality of data modulation and coding schemes, wherein said performance data comprises a plurality of performance curves, wherein the plurality of performance curves comprises a plurality of sigmoidal curves, each of said plurality of sigmoidal curves representing a sigmoidal relationship between the throughput of said physical layer and a condition of a data transmission channel used by said physical layer for each of said plurality of data modulation and coding schemes, wherein said condition comprises a signal to interference plus noise ratio (“SINR”) of said data transmission channel, and

modeling the plurality of performance curves using a single sigmoid function to approximate an envelope formed by upper knees of said plurality of sigmoidal curves, and adapting said single sigmoid function to changing conditions of said data transmission channel; and

predicting the performance of said network for the particular application program based at least in part on the obtained single sigmoid function based performance model of said physical layer.

14. The article of claim 13 , wherein said predicting the performance of said network for an application comprises:

receiving the SINR of said data transmission channel used by said physical layer;

estimating the performance of said physical layer under said SINR based on said sigmoid function based performance model of said physical layer;

estimating the performance of an application layer of said network based at least in part on said estimated performance of said physical layer under said SINR;

producing a data stream for transmitting said data required by said application at a rate according to the estimated application layer performance; and

estimating the quality of said data required by said application, when received, based on said data stream.

15. A method comprising:

modeling a performance of a physical layer of a network for a plurality of data modulation and coding schemes for a particular application program running on a computer and communicating data over the physical layer of the network, the computer including a processor, wherein the computer is in communication with the network, wherein said modeling includes modeling a combination of a plurality of curves using a sigmoid function, wherein each of the plurality of curves is associated with one of the plurality of data modulation and coding schemes; and

predicting the performance of said network for the particular application program based at least in part on the sigmoid function based performance model of said physical layer by the processor, wherein said predicting comprises:

receiving a signal to interference plus noise ratio (“SINR”) of a data transmission channel used by said physical layer;

estimating the performance of said physical layer under said SINR based on said sigmoid function based performance model of said physical layer;

estimating the performance of an application layer of said network based at least in part on said estimated performance of said physical layer under said SINR, and taking into account overheads added to said data required by said application by network layers higher than said physical layer; and

producing a data stream for transmitting said data required by said application at a rate according to the estimated application layer performance.

16. An apparatus comprising:

a physical layer (PHY) performance modeling mechanism to model the performance for a particular application program running on a computer and communicating data over a physical layer of a network for a plurality of data modulation and coding schemes based on modeling a combination of a plurality of curves using a single sigmoid function, wherein each of the plurality of curves is associated with one of the plurality of data modulation and coding schemes;

a quality estimator to predict the performance of said network for the particular application program based at least in part on the obtained sigmoid function based performance model of said physical layer;

a PHY performance estimator to receive a signal to interference plus noise ratio (“SINR”) of a data transmission channel used by said physical layer, and to estimate the performance of said physical layer under said SINR based on said sigmoid function based performance model of said physical layer; and

an application performance estimator to estimate the performance of an application layer of said network based at least in part on said estimated performance of said physical layer under said SINR and by taking into account overheads added to said data required by said application by network layers higher than said physical layer, and to produce a data stream for transmitting said data required by said application at a rate according to the estimated application layer performance.

17. A method comprising:

modeling a performance of a physical layer of a network for a plurality of data modulation and coding schemes for a particular application program running on a computer and communicating data over the physical layer of the network, the computer including a processor, wherein the computer is in communication with the network, wherein said modeling the performance of the physical player comprises:

representing a plurality of physical layer performance relationships comprising a plurality of sigmoidal curves collectively as a single sigmoid function such that said single sigmoid function approximates an envelope formed by upper knees of said plurality of sigmoidal curves, wherein each of the plurality of physical layer performance relationships indicates a relationship between a throughput of said physical layer and a condition of a data transmission channel used by said physical layer for one of said plurality of data modulation and coding schemes, wherein said condition comprises a signal to interference plus noise ratio (“SINR”) of said data transmission channel.