| 摘要 |
Certain embodiments of the present disclosure control whether or not certain protocol stack operation are performed in a hardware (HW) protocol stack accelerator based on ARQ/HARQ re-transmission rate. Latency penalties associated with using the HW accelerator are typically higher than the data movement overhead when this data does not need to be processed by the HW accelerator, such as when a re-transmission error occurs. According to certain embodiments, the HW accelerator is activated if the ARQ/HARQ re-transmission rate is below a threshold value. Otherwise, if the ARQ/HARQ re-transmission rate is above a threshold, at least a portion of the HW accelerator may be de-activated, which may reduce overhead associated with moving data between the protocol stack and the HW accelerator. |
