Selective convolution encoding of data transmitted over degraded links

1. A system, comprising a processor and logic integrated with and/or executable by the processor, the logic being configured to cause the processor to:

determine that a link configured to transmit data between two end points in a networking system has become degraded;

encode data prior to transmitting the data across the degraded link, the logic configured to encode the data being configured to:

segment the data to be encoded into k-bit segments;

mark each k-bit segment of the data to be encoded with a marker indicating convolution encoding; and

convolution encode the k-bit segments using m stages of delay via a partial response maximum likelihood (PRML) encoding scheme to produce encoded data, wherein k is greater than or equal to 1 and m is greater than or equal to 1,

wherein a 2/3 PRML is used to convolution encode the data in response to a determination that greater than 50% of data packets from the data are resent over the link, and

wherein a lower error code rate PRML is used to convolution encode the data in response to a determination that 50% or less of the data packets from the data are resent over the link; and

transmit the encoded data across the degraded link provided that data transmitted across a normal link is not encoded.

2. The system as recited in claim 1 , wherein the logic is further configured to cause the processor to determine that the link is normal by:

counting a number of retries over the link during a period of time;

comparing the number of retries against a threshold; and

reporting the link as normal when the number of retries does not exceed the threshold.

3. The system as recited in claim 1 , wherein the processor determines that the link has become degraded by:

counting a number of retries over the link during a period of time;

comparing the number of retries against a threshold; and

reporting the link as degraded when the number of retires exceeds the threshold.

4. The system as recited in claim 1 , wherein the logic is further configured to cause the processor to:

receive the encoded data;

apply an XOR function to the encoded data with pathing allowed by the convolution encoding to retrieve error-corrected data; and

identify the error-corrected data as an overall path with zero Hamming distance.

5. The system as recited in claim 1 , wherein the logic configured to convolution encode the k-bit segments using the m stages of delay is configured to generate 2 km states with 2 k branches entering each state and 2 k branches leaving each state.

6. The system as recited in claim 1 , wherein the logic is further configured to cause the processor to:

monitor the degraded link over one or more predetermined periods of time;

determine that the degraded link is normal; and

send subsequently received data across the normal link without encoding the subsequently received data.

7. A computer program product for selectively overcoming a degraded link in a networking system, the computer program product comprising a non-transitory computer readable storage medium having program code embodied therewith, the embodied program code being readable/executable by a processor to cause the processor to:

determine, by the processor, that a link configured to transmit data between two end points has become degraded;

encode, by the processor, data prior to transmitting the data across the degraded link, the embodied program code further causing the processor to:

segment the data to be encoded into k-bit segments;

mark each k-bit segment of the data to be encoded with a marker indicating convolution encoding; and

convolution encode the k-bit segments using m stages of delay via a partial response maximum likelihood (PRML) encoding scheme to produce encoded data, wherein k is greater than or equal to 1 and m is greater than or equal to 1,

wherein a 2/3 PRML is used to convolution encode the data in response to a determination that greater than 50% of data packets from the data are resent over the link, and

wherein a lower error code rate PRML is used to convolution encode the data in response to a determination that 50% or less of the data packets from the data are resent over the link; and

transmit, by the processor, the encoded data across the degraded link provided that data transmitted across a normal link is not encoded.

8. The computer program product as recited in claim 7 , wherein the program code is further readable/executable by the processor to:

determine that the link is normal by:

counting a number of retries over the link during a period of time,

comparing the number of retries against a threshold, and

reporting the link as normal when the number of retries does not exceed the threshold; and

transmit the data across the normal link without encoding the data,

wherein the processor determines that the link has become degraded when the number of retires exceeds the threshold.

9. The computer program product as recited in claim 7 , wherein the program code is further readable/executable by the processor to:

monitor the degraded link over one or more predetermined periods of time;

determine that the degraded link is normal and no longer degraded; and

send subsequently received data across the normal link without encoding the subsequently received data.

10. A method for selectively overcoming a degraded link in a networking system, the method comprising:

determining that a link configured to transmit data between two end points has become degraded;

encoding data prior to transmitting the data across the degraded link, the encoding the data comprising:

segmenting the data to be encoded into k-bit segments;

marking each k-bit segment of the data to be encoded with a marker indicating convolution encoding; and

convolution encoding the k-bit segments using m stages of delay via a partial response maximum likelihood (PRML) encoding scheme to produce encoded data, wherein k is greater than or equal to 1 and m is greater than or equal to 1,

wherein a 2/3 PRML is used to convolution encode the data in response to a determination that greater than 50% of data packets from the data are resent over the link, and

wherein a lower error code rate PRML is used to convolution encode the data in response to a determination that 50% or less of the data packets from the data are resent over the link; and

transmitting the encoded data across the degraded link.

11. The method as recited in claim 10 , further comprising:

determining that the link is normal by:

counting a number of retries over the link during a period of time;

comparing the number of retries against a threshold; and

reporting the link as normal when the number of retries does not exceed the threshold; and

transmitting the data across the normal link without encoding the data,

wherein the link is determined to have become degraded when the number of retires exceeds the threshold.

12. The method as recited in claim 10 , further comprising:

receiving the encoded data; and

decoding the encoded data to retrieve the data.

13. The method as recited in claim 10 , further comprising:

receiving the encoded data; and

applying an XOR function to the encoded data with pathing allowed by the convolution encoding to retrieve error-corrected data; and

identifying the error-corrected data as an overall path with zero Hamming distance.

14. The method as recited in claim 10 , wherein the convolution encoding the k-bit segments using m stages of delay comprises generating 2 km states with 2 k branches entering each state and 2 k branches leaving each state.

15. The method as recited in claim 10 , further comprising:

monitoring the degraded link over one or more predetermined periods of time;

determining that the degraded link is normal; and

sending subsequently received data across the normal link without encoding the subsequently received data.