Feedback-based online network coding

Abstract: Current approaches to the practical implementation of network coding are batch-based, and often do not use feedback, except possibly to signal completion of a file download. In this paper, the various benefits of using feedback in a network coded system are studied. It is shown that network coding can be performed in a completely online manner, without the need for batches or generations, and that such online operation does not affect the throughput. Although these ideas are presented in a single-hop packet erasure broadcast setting, they naturally extend to more general lossy networks which employ network coding in the presence of feedback. The impact of feedback on queue size at the sender and decoding delay at the receivers is studied. Strategies for adaptive coding based on feedback are presented, with the goal of minimizing the queue size and delay. The asymptotic behavior of these metrics is characterized, in the limit of the traffic load approaching capacity. Different notions of decoding delay are considered, including an order-sensitive notion which assumes that packets are useful only when delivered in order. Our work may be viewed as a natural extension of Automatic Repeat reQuest (ARQ) schemes to coded networks.