Packet loss models can miss risks of prolonged update failures
Survival-Time Risk Beyond Low-Order Channel Models: Identifiability and Scheduling
Networking and Internet ArchitectureInformation Theory
Summary
People use simple models to predict how often data packets get lost in a network, especially in bursts. This paper shows that such simple models can miss bigger risks where many updates in a row fail. The authors look at more complex channel behaviors that still match the simple models but have very different chances of long failure periods. They find that how and when retries happen can change which channel is riskier, so network scheduling needs deeper measurements to predict risks accurately.
What this means in practice
- •For network schedulers: Improve scheduling strategies by using channel statistics sensitive to retry timing, reducing unexpected long update failures during bursty losses.
- •For iot device managers: Design update transmission plans accounting for complex loss behaviors beyond simple loss rate and burst length to avoid prolonged update outages.
Authors
Ruifeng Zheng, Zhihan Xu, Pit Hofmann, Frank H. P. Fitzek
Abstract
This paper investigates whether low-order packet-loss statistics suffice to assess survival-time risk under bursty packet losses, i.e., the probability that no update succeeds for a prescribed number of consecutive slots. Conventional models, such as the Gilbert model, characterize packet losses using the loss rate and mean burst length, but channels sharing these statistics may exhibit substantially different failure risks. We characterize the Markov channels one order above a given low-order fit that share its statistics and analyze their risks under acknowledgment-based scheduling with transmission budgets. We show that, with dense transmission attempts, risk is primarily governed by loss persistence beyond the modeled memory, and the worst-case channel need not be the most persistent one. Under limited budgets, retry spacing makes loss relapse probabilities important and can reverse the risk ordering of compatible channels. Numerical and measurement results show that low-order models can substantially underestimate the risk, highlighting the need for scheduling-dependent channel statistics.