Five Blind Men and the Internet: Towards an Understanding of Internet Traffic

Ayush Mishra; Ege Cem Kirci; Laurent Vanbever

arxiv: 2509.06515 · v2 · submitted 2025-09-08 · 💻 cs.NI

Five Blind Men and the Internet: Towards an Understanding of Internet Traffic

Ege Cem Kirci , Ayush Mishra , Laurent Vanbever This is my paper

Pith reviewed 2026-05-18 18:30 UTC · model grok-4.3

classification 💻 cs.NI

keywords internet trafficexchange pointstraffic measurementnetwork growthself-similaritydiurnal patternstraffic proxyIXP statistics

0 comments

The pith

Traffic statistics from Internet exchange points provide a reliable proxy for overall Internet growth and behavior.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper seeks to show that data from hundreds of public Internet Exchange Points can give a clear picture of how Internet traffic is growing and how it is used around the world. This matters because the Internet has no central authority collecting detailed traffic information, making it hard to understand its scale and changes. The authors examined two years of statistics covering a large share of exchange point capacity and found steady growth along with consistent daily patterns in different regions. They also checked for biases in the data and found that the patterns closely match what one would expect from the entire Internet. This approach opens the door to ongoing, open monitoring of the network's health and expansion.

Core claim

By gathering publicly available traffic statistics from 472 Internet Exchange Points that together account for 87 percent of global IXP port capacity, the study measures a 49.2 percent increase in aggregate traffic over two years, identifies distinct regional diurnal cycles and sudden anomalies tied to events, and observes stable link utilization rates. Through bias analysis and verification of strong self-similarity, the work concludes that IXP traffic serves as a robust and representative indicator of broader Internet traffic volumes and usage trends.

What carries the argument

The combination of bias analysis in the IXP dataset and the observed high degree of self-similarity in traffic patterns, which together justify using IXP measurements as a stand-in for global Internet activity.

If this is right

Global Internet traffic increased by 49.2 percent between 2023 and 2024, averaging 24.5 percent growth per year.
Traffic patterns vary by region, with different peak times and responses to events.
Utilization of IXP infrastructure remains stable, indicating that capacity scales in line with demand.
The released dataset enables replicable studies of Internet dynamics over time.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

Researchers could apply similar methods to track how emerging applications like AI services affect network load.
Network planners might use these public metrics to anticipate where new exchange points are needed.
Comparing IXP data with other sources could highlight the role of private peering in overall traffic distribution.

Load-bearing premise

The selected Internet Exchange Points capture a representative and unbiased sample of worldwide traffic volumes and daily usage patterns.

What would settle it

A direct comparison showing that total global Internet traffic growth differs substantially from the growth rate measured at these IXPs over the same two-year period.

Figures

Figures reproduced from arXiv: 2509.06515 by Ayush Mishra, Ege Cem Kirci, Laurent Vanbever.

**Figure 2.** Figure 2: Example of OCR-based data extraction from [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 3.** Figure 3: Map of all IXPs listed in PeeringDB as of January 2023, classified by collection status. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

**Figure 4.** Figure 4: Stacked plot of daily mean traffic volume [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗

**Figure 6.** Figure 6: Global IXP traffic increased significantly over [PITH_FULL_IMAGE:figures/full_fig_p006_6.png] view at source ↗

**Figure 7.** Figure 7: Traffic surged substantially across all regions, [PITH_FULL_IMAGE:figures/full_fig_p007_7.png] view at source ↗

**Figure 9.** Figure 9: Same-day traffic from three IXPs, showing [PITH_FULL_IMAGE:figures/full_fig_p008_9.png] view at source ↗

**Figure 10.** Figure 10: As traffic volume increases, daily patterns [PITH_FULL_IMAGE:figures/full_fig_p009_10.png] view at source ↗

**Figure 12.** Figure 12: Distinct weekly traffic patterns across global [PITH_FULL_IMAGE:figures/full_fig_p010_12.png] view at source ↗

**Figure 13.** Figure 13: Sensitivity of regional IXP traffic patterns [PITH_FULL_IMAGE:figures/full_fig_p011_13.png] view at source ↗

**Figure 15.** Figure 15: Port capacity of IXPs increases significantly [PITH_FULL_IMAGE:figures/full_fig_p012_15.png] view at source ↗

**Figure 16.** Figure 16: Regional network utilization varies but re [PITH_FULL_IMAGE:figures/full_fig_p013_16.png] view at source ↗

read the original abstract

The Internet, the world's largest and most pervasive network, lacks a transparent, granular view of its traffic patterns, volumes, and growth trends, hindering the networking community's understanding of its dynamics. This paper leverages publicly available Internet Exchange Point traffic statistics to address this gap, presenting a comprehensive two-year study (2023-2024) from 472 IXPs worldwide, capturing approximately 300 Tbps of peak daily aggregate traffic by late 2024. Our analysis reveals a 49.2% global traffic increase (24.5% annualized), uncovers regionally distinct diurnal patterns and event-driven anomalies, and demonstrates stable utilization rates, reflecting predictable infrastructure scaling. By analyzing biases and confirming high self-similarity, we establish IXP traffic as a robust proxy for overall Internet growth and usage behavior. With transparent, replicable data--covering 87% of the worldwide IXP port capacity--and plans to release our dataset, this study offers a verifiable foundation for long-term Internet traffic monitoring. In particular, our findings shed light on the interplay between network design and function, providing an accessible framework for researchers and operators to explore the Internet's evolving ecosystem.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper delivers a large public IXP dataset with a clear 49% growth number over two years, but the claim that this serves as a robust proxy for overall Internet traffic rests on an untested representativeness assumption.

read the letter

The main thing here is the scale: two years of daily stats from 472 IXPs that together handle 87% of reported IXP port capacity, with a measured 49.2% aggregate traffic rise and plans to release the data. That gives the community a concrete, replicable baseline where none existed at this breadth. The regional diurnal patterns and event spikes are also straightforward to read and potentially useful for operators doing capacity work or anomaly spotting. Stable utilization rates across the period add a small practical observation about how infrastructure has kept up so far. Self-similarity is noted but is not a new finding on its own; it mainly serves to support the proxy argument. The dataset release is the part that actually lowers the barrier for follow-on work. The soft spot is the proxy conclusion. Port capacity share does not equal traffic volume share, and the vantage point misses private peering, content-provider backbones, and intra-AS flows that never traverse public IXPs. The paper states it analyzed biases, yet the abstract gives no external cross-check against non-IXP measurements, so the claim that the observed growth and patterns stand in for the unobserved remainder stays assumptive. Self-similarity within the sample does not close that gap. This is worth a serious referee for anyone who needs empirical traffic numbers rather than theory. Researchers and operators working on measurement, capacity planning, or traffic engineering will get direct value from the numbers and the released traces even if they treat the proxy language with caution. I would send it to review; the data collection is substantive and the representativeness question is one the community should debate in print rather than dismiss at the desk.

Referee Report

2 major / 2 minor

Summary. The manuscript presents a two-year (2023-2024) analysis of publicly available traffic statistics from 472 Internet Exchange Points (IXPs) worldwide, reporting an aggregate peak of approximately 300 Tbps by late 2024 and a 49.2% global traffic increase (24.5% annualized). It identifies regionally distinct diurnal patterns, event-driven anomalies, and stable utilization rates, and after performing bias analysis and confirming high self-similarity, concludes that IXP traffic constitutes a robust proxy for overall Internet growth and usage behavior. The study covers 87% of worldwide IXP port capacity, emphasizes transparent and replicable methods, and states plans to release the underlying dataset.

Significance. If the proxy claim is substantiated, the work supplies a verifiable, publicly grounded empirical baseline for tracking Internet traffic volumes, growth trends, and usage patterns at scale. The emphasis on replicable data collection and dataset release strengthens its potential utility for the networking research community.

major comments (2)

[§4] §4 (Bias Analysis subsection): The claim that biases have been analyzed to support the proxy conclusion requires explicit comparison to traffic that bypasses public IXPs (e.g., private peering volumes or intra-AS flows). Port-capacity coverage of 87% does not automatically imply representativeness of observed traffic volumes or patterns; if the bias quantification is limited to variation within the sampled IXPs, the central proxy argument rests on an untested assumption about completeness of the vantage point.
[§5.1] §5.1 (Self-similarity results): The observation of high self-similarity is presented as supporting evidence for the proxy. However, self-similarity is a well-documented property of aggregate Internet traffic; demonstrating it within the IXP sample does not, by itself, establish that the unobserved remainder (private peering, content-provider backbones) exhibits the same growth rates or diurnal structure.

minor comments (2)

[Abstract] The abstract states that biases were analyzed but does not summarize the specific statistical checks or external validation steps; adding one sentence on the method would improve clarity for readers.
[Tables/Figures] Table or figure captions describing the 472 IXPs should explicitly note whether the 87% figure refers to port capacity or to measured traffic volume.

Simulated Author's Rebuttal

2 responses · 1 unresolved

We thank the referee for their constructive and detailed feedback on our manuscript. We address each major comment point by point below, indicating where we agree and how we will revise the paper.

read point-by-point responses

Referee: [§4] §4 (Bias Analysis subsection): The claim that biases have been analyzed to support the proxy conclusion requires explicit comparison to traffic that bypasses public IXPs (e.g., private peering volumes or intra-AS flows). Port-capacity coverage of 87% does not automatically imply representativeness of observed traffic volumes or patterns; if the bias quantification is limited to variation within the sampled IXPs, the central proxy argument rests on an untested assumption about completeness of the vantage point.

Authors: We appreciate the referee highlighting this important limitation. Our bias analysis in §4 compared traffic statistics across IXP size, region, and peering policy to quantify variation within the sampled set. We agree that 87% port-capacity coverage does not automatically establish representativeness for traffic bypassing public IXPs. Direct empirical comparison to private peering or intra-AS flows is not feasible with publicly available data. We will revise the Bias Analysis subsection to explicitly state this scope limitation and moderate the proxy claim to apply to public IXP traffic. revision: partial
Referee: [§5.1] §5.1 (Self-similarity results): The observation of high self-similarity is presented as supporting evidence for the proxy. However, self-similarity is a well-documented property of aggregate Internet traffic; demonstrating it within the IXP sample does not, by itself, establish that the unobserved remainder (private peering, content-provider backbones) exhibits the same growth rates or diurnal structure.

Authors: We acknowledge that self-similarity is a well-known property of Internet traffic. In §5.1 we report high self-similarity specifically within the IXP dataset and link it to observed stability in utilization and diurnal patterns. We agree this does not by itself demonstrate identical properties in unobserved traffic segments. We will revise §5.1 and the conclusions to clarify that self-similarity supports consistency inside the sampled data rather than extending the proxy argument to private peering or backbone traffic. revision: yes

standing simulated objections not resolved

Direct quantitative comparison to private peering volumes and intra-AS flows, as no public datasets exist for these components.

Circularity Check

0 steps flagged

No circularity: proxy claim rests on external data and empirical observations

full rationale

The paper's central derivation uses publicly available traffic statistics from 472 IXPs (covering 87% of worldwide port capacity) to analyze biases and confirm high self-similarity as an empirical finding. This supports the claim that IXP traffic serves as a proxy for overall Internet growth. No step reduces a reported result to a fitted parameter defined inside the paper, a self-citation chain, or a definitional equivalence; the analysis draws on external vantage points and standard traffic properties without tautological construction. The derivation chain is self-contained against the provided data sources.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The claims rest on the accuracy and representativeness of public IXP statistics plus standard statistical assumptions about self-similarity; no new free parameters, invented physical entities, or ad-hoc axioms are introduced beyond domain-standard assumptions about measurement validity.

axioms (1)

domain assumption Publicly reported IXP traffic statistics accurately reflect actual exchanged volumes without systematic under- or over-reporting.
The entire measurement study depends on this premise; it is invoked when treating the collected numbers as ground truth for growth calculations.

pith-pipeline@v0.9.0 · 5734 in / 1313 out tokens · 44023 ms · 2026-05-18T18:30:58.187620+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

IndisputableMonolith/Foundation/RealityFromDistinction.lean reality_from_one_distinction unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

By analyzing biases and confirming high self-similarity, we establish IXP traffic as a robust proxy for overall Internet growth and usage behavior.

What do these tags mean?

matches: The paper's claim is directly supported by a theorem in the formal canon.
supports: The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends: The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses: The paper appears to rely on the theorem as machinery.
contradicts: The paper's claim conflicts with a theorem or certificate in the canon.
unclear: Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

30 extracted references · 30 canonical work pages

[1]

Bernhard Ager, Nikolaos Chatzis, Anja Feldmann, Nadi Sarrar, Steve Uhlig, and Walter Willinger. 2012. Anatomy of a Large European IXP. InProceedings of the ACM SIGCOMM 2012 Conference on Applications, Technologies, Architectures, and Protocols for Computer Communication (Sigcomm ’12). Association for Computing Machinery, New York, NY, USA, 163–174. https:...

work page doi:10.1145/2342356.2342393 2012
[2]

David Belson. 2023. Cloudflare 2023 Year in Review. https://blog. cloudflare.com/radar-2023-year-in-review/

work page 2023
[3]

David Belson. 2024. Cloudflare 2024 Year in Review. https://blog. cloudflare.com/radar-2024-year-in-review/

work page 2024
[4]

Timm Böttger, Felix Cuadrado, Gareth Tyson, Ignacio Castro, and Steve Uhlig. 2018. Open Connect Everywhere: A Glimpse at the Internet Ecosystem through the Lens of the Netflix CDN.SIGCOMM Comput. Commun. Rev.48, 1 (April 2018), 28–34. https://doi.org/10. 1145/3211852.3211857

work page arXiv 2018
[5]

Timm Böttger, Ghida Ibrahim, and Ben Vallis. 2020. How the Internet Reacted to Covid-19: A Perspective from Facebook’s Edge Network. InProceedings of the ACM Internet Measurement Conference (IMC ’20). Association for Computing Machinery, New York, NY, USA, 34–41. https://doi.org/10.1145/3419394.3423621

work page doi:10.1145/3419394.3423621 2020
[6]

Lloyd Brown, Albert Gran Alcoz, Frank Cangialosi, Akshay Narayan, Mohammad Alizadeh, Hari Balakrishnan, Eric Friedman, Ethan Katz- Bassett, Arvind Krishnamurthy, Michael Schapira, and Scott Shenker

work page
[7]

InProceedings of the ACM SIGCOMM 2024 Conference (ACM SIGCOMM ’24)

Principles for Internet Congestion Management. InProceedings of the ACM SIGCOMM 2024 Conference (ACM SIGCOMM ’24). Association for Computing Machinery, New York, NY, USA, 166–180. https: //doi.org/10.1145/3651890.3672247

work page doi:10.1145/3651890.3672247 2024
[8]

CAIDA UCSD. 2017. CAIDA UCSD PeeringDB Dataset, 2023-01-01 – 2024-12-31. https://www.caida.org/catalog/datasets/peeringdb/

work page 2017
[9]

Juan Camilo Cardona Restrepo and Rade Stanojevic. 2012. IXP Traffic: A Macroscopic View. InProceedings of the 7th Latin American Net- working Conference (Lanc ’12). Association for Computing Machinery, New York, NY, USA, 1–8. https://doi.org/10.1145/2382016.2382018

work page doi:10.1145/2382016.2382018 2012
[10]

Nikolaos Chatzis, Georgios Smaragdakis, Anja Feldmann, and Walter Willinger. 2013. There Is More to IXPs than Meets the Eye.SIGCOMM Comput. Commun. Rev.43, 5 (Nov. 2013), 19–28. https://doi.org/10. 1145/2541468.2541473

work page arXiv 2013
[11]

2020.Cisco Annual Internet Report (2018–

Cisco Systems. 2020.Cisco Annual Internet Report (2018–

work page 2020
[12]

White Paper C11-741490

White Paper. White Paper C11-741490. Cisco Systems, Inc. https://www.cisco.com/c/en/us/solutions/collateral/executive- perspectives/annual-internet-report/white-paper-c11-741490.pdf

work page
[13]

Marek Denis, Yuanjun Yao, Ashley Hatch, Qin Zhang, Chiun Lin Lim, Shuqiang Zhang, Kyle Sugrue, Henry Kwok, Mikel Jimenez Fer- nandez, Petr Lapukhov, Sandeep Hebbani, Gaya Nagarajan, Omar Baldonado, Lixin Gao, and Ying Zhang. 2023. EBB: Reliable and Evolvable Express Backbone Network in Meta. InProceedings of the ACM SIGCOMM 2023 Conference (Acm Sigcomm ’2...

work page doi:10.1145/3603269.3604860 2023
[14]

Anja Feldmann, Oliver Gasser, Franziska Lichtblau, Enric Pujol, Ing- mar Poese, Christoph Dietzel, Daniel Wagner, Matthias Wichtlhuber, Juan Tapiador, Narseo Vallina-Rodriguez, Oliver Hohlfeld, and Geor- gios Smaragdakis. 2020. The Lockdown Effect: Implications of the COVID-19 Pandemic on Internet Traffic. InProceedings of the ACM Internet Measurement Con...

work page arXiv 2020
[15]

Hadas Gold. 2020. Netflix and YouTube Are Slowing down in Europe to Keep the Internet from Breaking | CNN Business. https://www. cnn.com/2020/03/19/tech/netflix-internet-overload-eu

work page 2020
[16]

Grafana Labs Contributors. 2025. Grafana/Grafana: The Open-Source Platform for Monitoring and Observability. https://github.com/ grafana/grafana

work page 2025
[17]

INEX – Internet Neutral Exchange Association. 2025. IXP Manager — World’s Most Trusted IXP Platform. https://www.ixpmanager.org/

work page 2025
[18]

Craig Labovitz, Scott Iekel-Johnson, Danny McPherson, Jon Ober- heide, and Farnam Jahanian. 2010. Internet Inter-Domain Traffic. In Proceedings of the ACM SIGCOMM 2010 Conference (Sigcomm ’10). Association for Computing Machinery, New York, NY, USA, 75–86. https://doi.org/10.1145/1851182.1851194

work page doi:10.1145/1851182.1851194 2010
[19]

Samir Marwaha. 2024. Sandvine’s 2024 Global Internet Phenom- ena Report: Global Internet Usage Continues to Grow. https: //www.applogicnetworks.com/blog/sandvines-2024-global-internet- phenomena-report-global-internet-usage-continues-to-grow

work page 2024
[20]

2016.IXP Technical Requirements — OIX-1 (Version 2.0)

Open-IX Association. 2016.IXP Technical Requirements — OIX-1 (Version 2.0). Standard OIX-1. Open-IX Association. https://storage.googleapis.com/website-v4_media/documents/ IXPTechnicalRequirements2016.10.19.pdf

work page 2016
[21]

Open-IX Association. 2025. About Us – Open-IX Association. https: //testing.oix.org/about/

work page 2025
[22]

PeeringDB Community. 2025. PeeringDB — the Interconnection Data- base. https://www.peeringdb.com/

work page 2025
[23]

Maxime Piraux, Louis Navarre, Nicolas Rybowski, Olivier Bonaven- ture, and Benoit Donnet. 2022. Revealing the Evolution of a Cloud Provider through Its Network Weather Map. InProceedings of the 22nd ACM Internet Measurement Conference (Imc ’22). Association for Computing Machinery, New York, NY, USA, 298–304. https: //doi.org/10.1145/3517745.3561462

work page doi:10.1145/3517745.3561462 2022
[24]

Philipp Richter, Georgios Smaragdakis, Anja Feldmann, Nikolaos Chatzis, Jan Boettger, and Walter Willinger. 2014. Peering at Peer- ings: On the Role of IXP Route Servers. InProceedings of the 2014 Conference on Internet Measurement Conference (IMC ’14). Associa- tion for Computing Machinery, New York, NY, USA, 31–44. https: //doi.org/10.1145/2663716.2663757

work page doi:10.1145/2663716.2663757 2014
[25]

Matthew Roughan, Albert Greenberg, Charles Kalmanek, Michael Rumsewicz, Jennifer Yates, and Yin Zhang. 2002. Experience in Mea- suring Backbone Traffic Variability: Models, Metrics, Measurements and Meaning. InProceedings of the 2nd ACM SIGCOMM Workshop on Internet Measurment (Imw ’02). Association for Computing Machinery, New York, NY, USA, 91–92. https:...

work page doi:10.1145/637201.637213 2002
[26]

Sandvine. 2023. Sandvine’s 2023 Global Internet Phenomena Report Shows 24% Jump in Video Traffic, with Netflix Volume Overtaking YouTube. https://www.applogicnetworks.com/press- releases/sandvines-2023-global-internet-phenomena-report-shows- 24-jump-in-video-traffic-with-netflix-volume-overtaking-youtube

work page 2023
[27]

Tanzania Internet eXchange (TIX). 2025. TIX Traffic Statistics Portal. https://stats.tix.or.tz/

work page 2025
[28]

Tesseract OCR Contributors. 2024. Tesseract-Ocr/Tesseract: Tesseract Open Source OCR Engine. https://github.com/tesseract-ocr/tesseract

work page 2024
[29]

TimescaleDB Contributors. 2025. Timescale/Timescaledb: A Time- Series Database for High-Performance Real-Time Analytics. https: //github.com/timescale/timescaledb

work page 2025
[30]

Tong Yang, Jie Jiang, Peng Liu, Qun Huang, Junzhi Gong, Yang Zhou, Rui Miao, Xiaoming Li, and Steve Uhlig. 2018. Elastic Sketch: Adaptive and Fast Network-Wide Measurements. InProceedings of the 2018 Conference of the ACM Special Interest Group on Data Communication (Sigcomm ’18). Association for Computing Machinery, New York, NY, USA, 561–575. https://do...

work page doi:10.1145/3230543.3230544 2018

[1] [1]

Bernhard Ager, Nikolaos Chatzis, Anja Feldmann, Nadi Sarrar, Steve Uhlig, and Walter Willinger. 2012. Anatomy of a Large European IXP. InProceedings of the ACM SIGCOMM 2012 Conference on Applications, Technologies, Architectures, and Protocols for Computer Communication (Sigcomm ’12). Association for Computing Machinery, New York, NY, USA, 163–174. https:...

work page doi:10.1145/2342356.2342393 2012

[2] [2]

David Belson. 2023. Cloudflare 2023 Year in Review. https://blog. cloudflare.com/radar-2023-year-in-review/

work page 2023

[3] [3]

David Belson. 2024. Cloudflare 2024 Year in Review. https://blog. cloudflare.com/radar-2024-year-in-review/

work page 2024

[4] [4]

Timm Böttger, Felix Cuadrado, Gareth Tyson, Ignacio Castro, and Steve Uhlig. 2018. Open Connect Everywhere: A Glimpse at the Internet Ecosystem through the Lens of the Netflix CDN.SIGCOMM Comput. Commun. Rev.48, 1 (April 2018), 28–34. https://doi.org/10. 1145/3211852.3211857

work page arXiv 2018

[5] [5]

Timm Böttger, Ghida Ibrahim, and Ben Vallis. 2020. How the Internet Reacted to Covid-19: A Perspective from Facebook’s Edge Network. InProceedings of the ACM Internet Measurement Conference (IMC ’20). Association for Computing Machinery, New York, NY, USA, 34–41. https://doi.org/10.1145/3419394.3423621

work page doi:10.1145/3419394.3423621 2020

[6] [6]

Lloyd Brown, Albert Gran Alcoz, Frank Cangialosi, Akshay Narayan, Mohammad Alizadeh, Hari Balakrishnan, Eric Friedman, Ethan Katz- Bassett, Arvind Krishnamurthy, Michael Schapira, and Scott Shenker

work page

[7] [7]

InProceedings of the ACM SIGCOMM 2024 Conference (ACM SIGCOMM ’24)

Principles for Internet Congestion Management. InProceedings of the ACM SIGCOMM 2024 Conference (ACM SIGCOMM ’24). Association for Computing Machinery, New York, NY, USA, 166–180. https: //doi.org/10.1145/3651890.3672247

work page doi:10.1145/3651890.3672247 2024

[8] [8]

CAIDA UCSD. 2017. CAIDA UCSD PeeringDB Dataset, 2023-01-01 – 2024-12-31. https://www.caida.org/catalog/datasets/peeringdb/

work page 2017

[9] [9]

Juan Camilo Cardona Restrepo and Rade Stanojevic. 2012. IXP Traffic: A Macroscopic View. InProceedings of the 7th Latin American Net- working Conference (Lanc ’12). Association for Computing Machinery, New York, NY, USA, 1–8. https://doi.org/10.1145/2382016.2382018

work page doi:10.1145/2382016.2382018 2012

[10] [10]

Nikolaos Chatzis, Georgios Smaragdakis, Anja Feldmann, and Walter Willinger. 2013. There Is More to IXPs than Meets the Eye.SIGCOMM Comput. Commun. Rev.43, 5 (Nov. 2013), 19–28. https://doi.org/10. 1145/2541468.2541473

work page arXiv 2013

[11] [11]

2020.Cisco Annual Internet Report (2018–

Cisco Systems. 2020.Cisco Annual Internet Report (2018–

work page 2020

[12] [12]

White Paper C11-741490

White Paper. White Paper C11-741490. Cisco Systems, Inc. https://www.cisco.com/c/en/us/solutions/collateral/executive- perspectives/annual-internet-report/white-paper-c11-741490.pdf

work page

[13] [13]

Marek Denis, Yuanjun Yao, Ashley Hatch, Qin Zhang, Chiun Lin Lim, Shuqiang Zhang, Kyle Sugrue, Henry Kwok, Mikel Jimenez Fer- nandez, Petr Lapukhov, Sandeep Hebbani, Gaya Nagarajan, Omar Baldonado, Lixin Gao, and Ying Zhang. 2023. EBB: Reliable and Evolvable Express Backbone Network in Meta. InProceedings of the ACM SIGCOMM 2023 Conference (Acm Sigcomm ’2...

work page doi:10.1145/3603269.3604860 2023

[14] [14]

Anja Feldmann, Oliver Gasser, Franziska Lichtblau, Enric Pujol, Ing- mar Poese, Christoph Dietzel, Daniel Wagner, Matthias Wichtlhuber, Juan Tapiador, Narseo Vallina-Rodriguez, Oliver Hohlfeld, and Geor- gios Smaragdakis. 2020. The Lockdown Effect: Implications of the COVID-19 Pandemic on Internet Traffic. InProceedings of the ACM Internet Measurement Con...

work page arXiv 2020

[15] [15]

Hadas Gold. 2020. Netflix and YouTube Are Slowing down in Europe to Keep the Internet from Breaking | CNN Business. https://www. cnn.com/2020/03/19/tech/netflix-internet-overload-eu

work page 2020

[16] [16]

Grafana Labs Contributors. 2025. Grafana/Grafana: The Open-Source Platform for Monitoring and Observability. https://github.com/ grafana/grafana

work page 2025

[17] [17]

INEX – Internet Neutral Exchange Association. 2025. IXP Manager — World’s Most Trusted IXP Platform. https://www.ixpmanager.org/

work page 2025

[18] [18]

Craig Labovitz, Scott Iekel-Johnson, Danny McPherson, Jon Ober- heide, and Farnam Jahanian. 2010. Internet Inter-Domain Traffic. In Proceedings of the ACM SIGCOMM 2010 Conference (Sigcomm ’10). Association for Computing Machinery, New York, NY, USA, 75–86. https://doi.org/10.1145/1851182.1851194

work page doi:10.1145/1851182.1851194 2010

[19] [19]

Samir Marwaha. 2024. Sandvine’s 2024 Global Internet Phenom- ena Report: Global Internet Usage Continues to Grow. https: //www.applogicnetworks.com/blog/sandvines-2024-global-internet- phenomena-report-global-internet-usage-continues-to-grow

work page 2024

[20] [20]

2016.IXP Technical Requirements — OIX-1 (Version 2.0)

Open-IX Association. 2016.IXP Technical Requirements — OIX-1 (Version 2.0). Standard OIX-1. Open-IX Association. https://storage.googleapis.com/website-v4_media/documents/ IXPTechnicalRequirements2016.10.19.pdf

work page 2016

[21] [21]

Open-IX Association. 2025. About Us – Open-IX Association. https: //testing.oix.org/about/

work page 2025

[22] [22]

PeeringDB Community. 2025. PeeringDB — the Interconnection Data- base. https://www.peeringdb.com/

work page 2025

[23] [23]

Maxime Piraux, Louis Navarre, Nicolas Rybowski, Olivier Bonaven- ture, and Benoit Donnet. 2022. Revealing the Evolution of a Cloud Provider through Its Network Weather Map. InProceedings of the 22nd ACM Internet Measurement Conference (Imc ’22). Association for Computing Machinery, New York, NY, USA, 298–304. https: //doi.org/10.1145/3517745.3561462

work page doi:10.1145/3517745.3561462 2022

[24] [24]

Philipp Richter, Georgios Smaragdakis, Anja Feldmann, Nikolaos Chatzis, Jan Boettger, and Walter Willinger. 2014. Peering at Peer- ings: On the Role of IXP Route Servers. InProceedings of the 2014 Conference on Internet Measurement Conference (IMC ’14). Associa- tion for Computing Machinery, New York, NY, USA, 31–44. https: //doi.org/10.1145/2663716.2663757

work page doi:10.1145/2663716.2663757 2014

[25] [25]

Matthew Roughan, Albert Greenberg, Charles Kalmanek, Michael Rumsewicz, Jennifer Yates, and Yin Zhang. 2002. Experience in Mea- suring Backbone Traffic Variability: Models, Metrics, Measurements and Meaning. InProceedings of the 2nd ACM SIGCOMM Workshop on Internet Measurment (Imw ’02). Association for Computing Machinery, New York, NY, USA, 91–92. https:...

work page doi:10.1145/637201.637213 2002

[26] [26]

Sandvine. 2023. Sandvine’s 2023 Global Internet Phenomena Report Shows 24% Jump in Video Traffic, with Netflix Volume Overtaking YouTube. https://www.applogicnetworks.com/press- releases/sandvines-2023-global-internet-phenomena-report-shows- 24-jump-in-video-traffic-with-netflix-volume-overtaking-youtube

work page 2023

[27] [27]

Tanzania Internet eXchange (TIX). 2025. TIX Traffic Statistics Portal. https://stats.tix.or.tz/

work page 2025

[28] [28]

Tesseract OCR Contributors. 2024. Tesseract-Ocr/Tesseract: Tesseract Open Source OCR Engine. https://github.com/tesseract-ocr/tesseract

work page 2024

[29] [29]

TimescaleDB Contributors. 2025. Timescale/Timescaledb: A Time- Series Database for High-Performance Real-Time Analytics. https: //github.com/timescale/timescaledb

work page 2025

[30] [30]

Tong Yang, Jie Jiang, Peng Liu, Qun Huang, Junzhi Gong, Yang Zhou, Rui Miao, Xiaoming Li, and Steve Uhlig. 2018. Elastic Sketch: Adaptive and Fast Network-Wide Measurements. InProceedings of the 2018 Conference of the ACM Special Interest Group on Data Communication (Sigcomm ’18). Association for Computing Machinery, New York, NY, USA, 561–575. https://do...

work page doi:10.1145/3230543.3230544 2018