m3taversal
be8ff41bfe
Wrote sourced_from: into 414 claim files pointing back to their origin source. Backfilled claims_extracted: into 252 source files that were processed but missing this field. Matching uses author+title overlap against claim source: field, validated against 296 known-good pairs from existing claims_extracted. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
40 lines
No EOL
2.2 KiB
Markdown
40 lines
No EOL
2.2 KiB
Markdown
---
|
|
type: claim
|
|
domain: internet-finance
|
|
description: "Higher variance-to-mean ratio requires more capacity to maintain same congestion level"
|
|
confidence: proven
|
|
source: "Liu et al. (NC State), 'Modeling and Simulation of Nonstationary Non-Poisson Arrival Processes' (2019)"
|
|
created: 2026-03-11
|
|
sourced_from:
|
|
- inbox/archive/internet-finance/2019-00-00-liu-modeling-nonstationary-non-poisson-arrival-processes.md
|
|
---
|
|
|
|
# Arrival process burstiness increases required capacity for fixed service level
|
|
|
|
Congestion measures (queue length, wait time, utilization) are increasing functions of arrival process variability. For a fixed average arrival rate and service rate, a bursty arrival process requires more capacity than a smooth (Poisson) arrival process to maintain the same service level.
|
|
|
|
This means that modeling arrivals as Poisson when they are actually bursty (higher variance-to-mean ratio) will systematically underestimate required capacity, leading to service degradation.
|
|
|
|
## Evidence
|
|
|
|
Liu et al. establish that "congestion measures are increasing functions of arrival process variability — more bursty = more capacity needed." This is a fundamental result in queueing theory: variance in the arrival process translates directly to variance in system state, which manifests as congestion.
|
|
|
|
The CIATA method explicitly models the "asymptotic variance-to-mean (dispersion) ratio" as a separate parameter from the rate function, recognizing that burstiness is a first-order determinant of system performance, not a second-order correction.
|
|
|
|
## Application to Research Pipeline Capacity
|
|
|
|
For pipelines processing research sources that arrive in bursts:
|
|
|
|
1. A Poisson model with the same average rate will underestimate queue lengths and wait times
|
|
2. Capacity sized for Poisson arrivals will experience congestion during burst periods
|
|
3. The dispersion ratio (variance/mean) must be measured and incorporated into capacity planning
|
|
|
|
The MMPP framework provides a tractable way to model this: the state-switching structure naturally generates higher variance than Poisson while remaining analytically tractable for capacity calculations.
|
|
|
|
---
|
|
|
|
Relevant Notes:
|
|
- domains/internet-finance/_map
|
|
|
|
Topics:
|
|
- core/mechanisms/_map |