Impact of Channel Memory on the End-to-end Communication Latency
Qixing Guan, Xiaoli Xu · 2021 IEEE Globecom Workshops (GC Wkshps) · 2021
In this paper, we investigate the impact of channel memory on the end-to-end (E2E) communication latency in the network with Bernoulli packet arrival and Gilbert-Elliott (GE) erasure channel. We first consider the network without feedback channel, by analyzing the E2E latency achieved with a random coding strategy. The explicit expression of the E2E latency is derived based on the infinite Markov chain model. We further investigate the impact of channel memory on the E2E latency for networks with feedback, by analyzing the automatic repeat query (ARQ) scheme in the presence of instantaneous feedback and considering the existing coding strategy based on delayed feedback. It is revealed that the E2E latency grows with the channel memory η(0 ≤ η < 1) by $\frac{\eta }{{1 - \eta }}$, for both scenarios without feedback and with instantaneous feedback.