Adaptive Causal Network Coding with Feedback for Delay and Throughput Guarantees
We propose a novel causal coding scheme with forward error correction (FEC) for a point-to-point communication link with delayed feedback. The proposed model can learn the erasure pattern in the channel, and adaptively adjust its transmission and FEC rate based on the burstiness of the channel and the feedback. We investigate the throughput, and the in-order delivery delay of the adaptive causal coding algorithm, and contrast its performance with the one of the selective repeat (SR) ARQ. We demonstrate via an experimental study of the protocol that our model can double the throughput gains, and triple the gain in terms of mean in-order delivery delay when the channel is bursty, while keeping the difference between the maximum and mean in-order delivery delay is much smaller than SR ARQ. Closing the delay gap along with boosting the throughput is very promising for enabling ultra-reliable low-latency communications applications. We validate the performance of data delivery under the traces of Intel.
READ FULL TEXT