Michael Welzl

• Hovind, Ingebrigt; Welzl, Michael; Islam, Safiqul & Fairhurst, Gorry (2026). An Open Source Implementation of IETF Transport Services over QUIC, TCP and UDP. I Krishnan, Suresh (Red.), ANRW '26: Proceedings of the 2026 Applied Networking Research Workshop. Association for Computing Machinery (ACM). ISSN 9798400728761. doi: 10.1145/3822163.3827933. Fulltekst i vitenarkiv Vis sammendrag

  IETF Transport Services provide applications with an easy to use, generic API that makes them independent of the underlying network interfaces and transport protocols at compile time. This simplicity is achieved by relocating functions to choose, configure and race transport protocols below the API. To date, there is no open-source implementation of TAPS that uses QUIC; this paper describes the first one, in C, using Picoquic. The implementation process has highlighted which features of QUIC can be preserved with IETF Transport Services, and which limitations occur.

• Hamrah, Hamid; Biswas, Kamanashis; Islam, Safiqul; Welzl, Michael & Feng, Boning (2026). A Pareto-Based Carbon-Aware Routing Protocol for Inter-Domain Routing. I Solmaz, Gürkan & Vahabi, Maryam (Red.), 2026 IEEE 51st Conference on Local Computer Networks (LCN). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9798319531483. doi: 10.1109/lcn67947.2026.11660809. Fulltekst i vitenarkiv Vis sammendrag

  Existing inter-domain routing algorithms do not consider carbon intensity in their route selection process. Recent research aims to address this limitation by proposing carbon-aware inter-domain routing protocols that select greener routes based on available carbon information along the path. However, these approaches often choose longer paths and therefore lead to increased end-to-end latency. In this paper, we argue that effective route selection must consider both path cost and carbon intensity to manage the trade-off between latency efficiency and carbon emissions. To this end, we propose a Pareto-Based Carbon-Aware Routing (PBCR) protocol, which applies path length restriction and regret-based ranking to balance between these two routing metrics. The experimental outcomes indicate that PBCR achieves significant carbon reduction while also minimising end-to-end latency compared to existing carbon-aware routing protocols.

• Bian, Jie & Welzl, Michael (2026). Linking Rationale to Decision on Internet Standards: A Retrieval-Based Approach Using Synthetic Data. I Piperidis,, Stelios; Bel, Núria; Heuvel, Henk van den; Ide, Nancy; Krek, Simon & Toral, Antonio (Red.), Proceedings of the Fifteenth Language Resources and Evaluation Conference (LREC 2026). European Language Resources Association. ISSN 9782493814494. s. 7149–7162. doi: 10.63317/3szh4omfcsxb. Fulltekst i vitenarkiv Vis sammendrag

  The Internet Engineering Task Force (IETF) develops Internet-Drafts (I-Ds) and Requests for Comments (RFCs) as formal specifications for Internet Protocols. While these documents capture finalized technical standards, the rich design rationales and deliberations that shape them are often buried in informal discussions across mailing lists. These discussions are rarely linked explicitly to the specifications they inform, making it difficult to trace the origins of specific design decisions. We address this gap by generating synthetic data that explicitly links discussion threads to their corresponding RFC/I‑D sections, producing roughly 350 000 such aligned instances. This data enables training a semantic embedding-based information retrieval (IR) system that, given an email discussion, retrieves the most relevant specification content. Our experiments show that this synthetic supervision helps models learn associations between informal discourse and formal documentation, though the task remains challenging due to the implicit and context-dependent nature of the links.

• Clemm, Alexander; Kutscher, Dirk; Welzl, Michael; Westphal, Cedric; Zilberman, Noa & Ferlin-Reiter, Simone (2025). Greening Networking: Toward a Net Zero Internet (Dagstuhl Seminar 24402). Dagstuhl Reports. 14(9), s. 167–192. doi: 10.4230/DAGREP.14.9.167. Fulltekst i vitenarkiv Vis sammendrag

  This is the report of the Dagstuhl Seminar 24402 on Greening Networking: Toward a Net Zero Internet that took place from September 29th to October 2nd 2024. The seminar discussed the most impactful networking improvements for reducing carbon emissions in three different areas: 1) applications, systems, and stakeholders; 2) network technologies; and 3) lifecycle and control loops. As a major result of the seminar, the following problems and topics for future research were identified: 1) characterizing the Internet footprint on carbon emissions accurately; 2) understanding attributional and consequential accounting of carbon emissions in networked systems; and 3) identifying potential solutions to give network systems more flexibility in better supporting energy grids and connecting to renewable energy sources. One of the concrete results of this seminar is a list of technologies and research opportunities for which we estimated the potential impact and time horizons.

• Bian, Jie; Arefyev, Nikolay; Mühlhäuser, Max & Welzl, Michael (2025). Automated Insights Into GitHub Collaboration Dynamics. IEEE Access. 13, s. 85526–85542. doi: 10.1109/access.2025.3566309. Fulltekst i vitenarkiv Vis sammendrag

  Today, GitHub (Trademark) is the most widely used platform for open-source software development. Large projects may comprise hundreds of distributed collaborators and thousands of GitHub “issues” (structured discussions). It includes basic support for dealing with issues, via “Pull Requests” (PRs)—document changes that can manually be defined to “close” them (i.e., they address and thereby conclude the issue discussion). Unresolved issues can pile up. For example, at the time of writing, the Kubernetes repository has almost 2000 open issues; finding which ones a PR might close is a hard task by itself. We address this by automatically identifying issue-PR relationships using language models (LMs) and leveraging Information Retrieval (IR) techniques. To foster further research, we contribute a carefully curated novel dataset called CodeConvo, reflecting the most influential open-source repositories for code development as well as some technical document repositories. We use this dataset to benchmark several state-of-the-art (SOTA) non-proprietary models that show exceptional performance on the MTEB benchmark, as well as to train and evaluate the performance of Smart Insights into GitHub Issue-PR Relations (SIGIR), our tailored model. The best SIGIR model/data combination yields an average Mean Reciprocal Rank (MRR) above 0.7, around 20% higher than the best baseline performance. Notably, ablation studies revealed that knowledge transfer occurs not only between different programming languages but also between code and technical documents, albeit to a lesser extent. We believe these results are encouraging and can stimulate the practical application of LMs for taming the complexity of very large projects, in GitHub and beyond.

• Bian, Jie & Welzl, Michael (2025). Empowering IETF Collaboration with NLP Search Innovations and LLM-Enhanced RFC Writing. I Brunstrom, Anna & Beckett, Ryan (Red.), ANRW '25: Proceedings of the 2025 Applied Networking Research Workshop. Association for Computing Machinery (ACM). ISSN 9798400720093. s. 24–31. doi: 10.1145/3744200.3744761. Fulltekst i vitenarkiv Vis sammendrag

  The Internet Engineering Task Force (IETF) produces extensive textual data, including discussions in email archives and GitHub repositories and Internet-Drafts (I-Ds), which are preliminary versions of Requests for Comments (RFCs). The sheer volume and complexity of this material present significant workflow challenges, contributing to the duration of the standardization process (it is common to take several years from an initial draft to the final RFC). This paper explores the potential of Natural Language Processing (NLP) using Large Language Models (LLMs) to streamline IETF workflows. We use Information Retrieval (IR) to i) build a search system that helps users locate comments related to details in an I-D, and ii) partially automate RFC writing.

• Bian, Jie & Welzl, Michael (2025). Instruction Tuning TextFlow Semi-automatic RFCs Generation. I Ichise, Ryutaro (Red.), Natural Language Processing and Information Systems NLDB 2025. Springer Nature. ISSN 9783031971402. s. 350–364. doi: 10.1007/978-3-031-97141-9_24. Fulltekst i vitenarkiv Vis sammendrag

  Instruction tuning enhances language models’ (LMs) performance by refining them with specific guidelines. This paper explores the application of instruction tuning in generating and editing Internet Drafts (I-Ds), the preliminary versions of Request for Comments (RFCs). The process involves training models with detailed instructions based on earlier drafts and feedback from Working Groups (WGs) to improve the text. This approach enables the generation of drafts that adhere to established Internet Engineering Task Force (IETF) standards, significantly reducing the need for extensive manual revisions that can span years. This work marks a promising start for the future development of network protocols. By combining instruction tuning models with human expertise, we are moving towards more efficient and accurate technical documentation.

• Khan, Al Nafeu; Hasan, Mahmudul; Islam, Safiqul; Salehin, Khondaker; Welzl, Michael & Feng, Boning [Vis alle 7 forfattere av denne artikkelen] (2025). Flow-Label Trends in IPv6 Traffic: A 9-Year Analysis of a Dataset Collected in Japan. I Shiomoto, Kohei & Maier, Guido Alberto (Red.), 2025 IEEE 26th International Conference on High Performance Switching and Routing (HPSR). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9798331529918. doi: 10.1109/hpsr64165.2025.11038848. Fulltekst i vitenarkiv Vis sammendrag

  RFC 6437 introduces the 20-bit flow label field in the IPv6 header and recommends it to be used alongside source and destination IP addresses for efficient flow classification. A recent paper from Google authors explains how they have begun to use this field in their IPv6 network to select an alternative path, thereby preventing congestion or outages. However, there is no prior research investigating whether connections make use of this field on the Internet. In this paper, we present a comprehensive analysis of the 9-year MAWI dataset (2016-2024), the Internet traces gathered from a large backbone link in Japan, and examine whether the IPv6 flow label is indeed employed on the Internet. Our findings reveal an increasing upward trend in the adoption of the flow label field. In particular, this trend correlates with the use of QUIC, whose specification—different from TCP—explicitly repeats the recommendation to utilize the flow label.

• Fornes, Håkon Medhus; Birketvedt, Elias Hoel; Griwodz, Carsten; Skjegstad, Magnus; Welzl, Michael & Alay, Ozgu (2025). Acceptable Latency in Predictable First-Person VR Cloud Games. I Orduna, Marta (Red.), MMVE ’25 Proceedings of The 17th International Workshop on IMmersive Mixed and Virtual Environment Systems. Association for Computing Machinery (ACM). ISSN 9798400714689. s. 58–64. doi: 10.1145/3712677.3720467. Fulltekst i vitenarkiv Vis sammendrag

  Over the past three decades, Virtual Reality (VR) technology has seen remarkable advancements in both hardware and content creation. As these technologies continue to evolve, it becomes essential to evaluate the contemporary VR user experience. Latencies exceeding 100 ms are usual between countries, yet many modern users who engage in VR cloud gaming with such latencies report satisfactory experiences. Since this includes fast-paced games like table tennis, we have wonder whether the difference between anecdotal evidence from players and the 100 ms limit for motion-to-photon latency (MTP) frequently reported in the literature are due to choice of games for studies, which require immediate reactions to the game's or opponents' actions. We take a close look at two first-person games that give players a moderate amount of planning time, Beat Saber and Superhot VR. We find indications that the VR setup used in our experiments (which could resemble a typical home user's PC VR gaming scenario) can already have a baseline MTP of just below 100 ms, and we have evaluated user experience with additional latencies of 0 ms, 50 ms, and 100 ms. We observe that a combined latency of 150 ms is still deemed tolerable, but that user experience is more negatively impacted in Beat Saber compared to Superhot VR.

• Welzl, Michael (2025). TCP Pacing in the Linux Kernel, 2025 International Conference on Computing, Networking and Communications (ICNC). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9798331520960. s. 133–139. doi: 10.1109/icnc64010.2025.10994011. Fulltekst i vitenarkiv Vis sammendrag

  Pacing introduces short pauses in between data packets upon transmission; this is often beneficial. For example, pacing can reduce the chance of packet loss, and improve operation with very short network queues. For TCP, it has been available in the Linux kernel for many years, and it applies in the same way to all TCP congestion control variants that do not influence pacing by themselves (at the time of writing, only BBR is such an exception). Several papers offer a high level description of its operation; here, we give the first in-depth explanation of how Linux TCP pacing really works. We illustrate its behavior with a trace from a testbed experiment, and provide a companion guide that makes the relevant kernel code easier to understand.

• Welzl, Michael; Islam, Safiqul; Ciko, Kristjon; Barhaugen, Petter & Haukebøe, Oskar (2025). NETHINT: Monitoring Local Network Interference With Passive Measurements, 2025 International Conference on Computing, Networking and Communications (ICNC). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9798331520960. s. 668–674. doi: 10.1109/icnc64010.2025.10993931. Fulltekst i vitenarkiv Vis sammendrag

  The "quality" of a home Internet connection is hard to quantify, as it is a subjective measure encompassing many factors. One of them, the detection of local (as opposed to remote) traffic interference, is not provided by existing tools. We address this with NETHINT, a tool to detect whether users and their applications disturb each other in a home network. NETHINT identifies traffic correlations via passive online latency monitoring. We explain the inner workings of NETHINT, and show how it can be successfully applied for live online root cause analysis of delay in a local network.

• Guibert, Loïc; Reynaud, Sébastien; Weppe, Olivier; Ciko, Kristjon; Welzl, Michael & Rumley, Sébastien (2025). TRAFFIC: Testbed foR Assessing energy eFFiciency In throughput Computing. I Palumbo, Francesca; Tumeo, Antonino; Varbanescu, Ana Lucia & Simmhan, Yogesh (Red.), CF '25 Companion: Proceedings of the 22nd ACM International Conference on Computing Frontiers: Workshops and Special Sessions. Association for Computing Machinery (ACM). ISSN 9798400713934. s. 99–107. doi: 10.1145/3706594.3727961. Fulltekst i vitenarkiv Vis sammendrag

  Information and Communication Technologies (ICT) represent a significant share of global resource usage. Notably, the energy consumption of data centres has emerged as a critical concern, escalating rapidly, especially with the rise of generative Artificial Intelligence (AI) models. This surge in energy demand calls for efforts focused on measuring and reducing the energy and carbon footprints of components used in ICT service delivery. However, the variety and multitude of devices involved in these services make it challenging to accurately measure and evaluate these footprints. In this paper, we propose an alternative approach to assess energy impacts by considering data centres for what they are: throughput computing systems offering ICT services. We define a simple experimental testbed and evaluate several machines with different hardware capacities. These machines serve two ICT services: a loop incrementing a counter until a defined limit, and an AI inference predicting the next token based on an input. Our experiment highlights three main take-aways: (a) the CPU usage is a poor predictor of the power consumption, (b) the energy or CO2 quantity associated with a service visit highly depends on the total load the service is facing, and (c) modern machines do not yield better energy figures compared to older ones in all circumstances.

• Welzl, Michael; Islam, Safiqul & Stephanides, Maximilian Von (2024). Real-Time TCP Packet Loss Prediction Using Machine Learning. IEEE Access. 12, s. 159622–159634. doi: 10.1109/ACCESS.2024.3488511. Fulltekst i vitenarkiv Vis sammendrag

  Congestion and resulting packet loss in TCP connections can lead to performance degradation and reduce the Quality of Experience (QoE) for end users. Many common TCP congestion control algorithms therefore adjust their sending rate proactively, using heuristics based on measured delay and/or throughput. We have investigated whether it would be possible to replace such heuristics with Machine Learning (ML), such that Internet hosts could meaningfully react when a packet loss is predicted to happen. For example, such a reaction could involve the use of Forward Erasure Correction (FEC) to protect against the loss, or to dynamically tune parameters of TCP. We present two ML models that were trained for Reno and Cubic flows, respectively, and show that they can be beneficially applied in real time. We do this by informing a TCP sender to reduce its rate via a local, artificially generated Explicit Congestion Notification (ECN) signal at the sender before packet loss would have happened.

• Yuan, Gina; Sotoudeh, Matthew; Zhang, David; Welzl, Michael; Mazières, David & Winstein, Keith (2024). Sidekick: In-Network Assistance for Secure End-to-End Transport Protocols. I Vanbever, Laurent & Zhang, Irene (Red.), 21st USENIX Symposium on Networked Systems Design and Implementation (NSDI 24). USENIX - The Advanced Computing Systems Association. ISSN 9781939133397. Fulltekst i vitenarkiv Vis sammendrag

  In response to concerns about protocol ossification and privacy, post-TCP transport protocols such as QUIC and Web- RTC include end-to-end encryption and authentication at the transport layer. This makes their packets opaque to middleboxes, freeing the transport protocol to evolve but preventing some in-network innovations and performance improvements. This paper describes sidekick protocols: an approach to innetwork assistance for opaque transport protocols where innetwork intermediaries help endpoints by sending information adjacent to the underlying connection, which remains opaque and unmodified on the wire. A key technical challenge is how the sidekick connection can efficiently refer to ranges of packets of the underlying connection without the ability to observe cleartext sequence numbers.We present a mathematical tool called a quACK that concisely represents a selective acknowledgment of opaque packets, without access to cleartext sequence numbers. In real-world and emulation-based evaluations, the sidekick improved performance in several scenarios: early retransmission over lossy Wi-Fi paths, proxy acknowledgments to save energy, and a path-aware congestion-control mechanism we call PACUBIC that emulates a “split” connection.

• Meuser, Tobias; Loven, Lauri; Bhuyan, Monowar; Patil, Shishir G.; Dustdar, Schahram & Aral, Atakan [Vis alle 18 forfattere av denne artikkelen] (2024). Revisiting Edge AI: Opportunities and Challenges. IEEE Internet Computing. ISSN 1089-7801. 28(4), s. 49–59. doi: 10.1109/MIC.2024.3383758. Fulltekst i vitenarkiv Vis sammendrag

  Edge AI is an innovative computing paradigm that aims to shift the training and inference of machine learning models to the edge of the network. This paradigm offers the opportunity to significantly impact our everyday lives with new services such as autonomous driving and ubiquitous personalized healthcare. Nevertheless, bringing intelligence to the edge involves several major challenges, which include the need to constrain model architecture designs, the secure distribution and execution of the trained models, and the substantial network load required to distribute the models and data collected for training. In this article, we highlight key aspects in the development of edge AI in the past and connect them to current challenges. This article aims to identify research opportunities for edge AI, relevant to bring together the research in the fields of artificial intelligence and edge computing.

• Bian, Jie; Welzl, Michael; Kutuzov, Andrei & Arefyev, Nikolay (2024). Tell Me Why: Language Models Help Explain the Rationale Behind Internet Protocol Design. I LI, Geoffrey Ye; Liang, Le; Gündüz, Deniz & Antón-Haro, Carles (Red.), 2024 IEEE International Conference on Machine Learning for Communication and Networking (ICMLCN). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9798350343199. s. 447–453. doi: 10.1109/ICMLCN59089.2024.10624781. Fulltekst i vitenarkiv Vis sammendrag

  Request for Comments (RFCs) serve as guidebooks for the implementation of Internet protocols or network mechanisms. They reveal how these protocols and mechanisms work, but the underlying reasons for their operation are not always available in RFCs. We present an attempt to automate the discovery of these reasons in mailing list archives via natural language processing methods. Our approach leverages the known relationship between text changes and the discussions leading to these changes in the Working Group (WG) GitHub repositories of the Internet Engineering Task Force (IETF) to obtain labeled training data. We find that our model is able to generalize, and it can indeed discover emails that have led to formulations in RFCs. This is a first step towards facilitating a deeper understanding of these often complex documents, which can be helpful for developers, protocol designers, and educators alike.

• Islam, Safiqul; Welzl, Michael; Hapnes, Erlend & Feng, Boning (2024). Using the IPv6 Flow Label for Path Consistency: A Large-Scale Measurement Study, ICC 2024 - IEEE International Conference on Communications. IEEE (Institute of Electrical and Electronics Engineers). ISSN 9781728190549. doi: 10.1109/ICC51166.2024.10622542. Fulltekst i vitenarkiv Vis sammendrag

  RFC 6437 specifies the usage of an IPv6 3-tuple (flow label, source and destination address fields) for flow classification in routers. Is this the only flow classification approach in use with IPv6 when the flow label is not zero? If so, end hosts could benefit from knowing that multiple flows between the same host pair traverse the same network bottleneck (for example, single-path congestion control coupling mechanisms could then reduce loss and delay). This paper attempts to answer this question with an investigation of whether path consistency is achievable via the IPv6 3-tuple in the Internet. We employ a novel method that extends Paris traceroute to control the path behavior solely using IPv6 network-layer header data, and conduct a large-scale measurement study from globally stationed vantage points to examine i) the extent to which the flow label is delivered to its destination, and ii) how it influences the behavior of load-balancing algorithms. Our results suggest that the flow label always passes through the network unaltered, but the 3-tuple is not a reliable flow identifier in the current Internet.

• Islam, Safiqul; Morina, Visar; Welzl, Michael & Feng, Boning (2024). TeachNet: A Lightweight and User-Friendly Network Visualization and Experimentation Tool. I Hendrantoro, Gamantyo & Cho, Sungrae (Red.), 29th Asia Pacific Conference on Communications (APCC). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9798331517816. s. 523–526. doi: 10.1109/APCC62576.2024.10767938. Fulltekst i vitenarkiv Vis sammendrag

  Network researchers, academics, and students often need to deal with complex scripting in order to design a network topology, configure network nodes, automate experimentation, and visualize network performance metrics. Current simulation or emulation tools lack the ability to provide a comprehensive understanding of the network. This paper introduces TeachNet, a lightweight and user-friendly network tool that provides a graphical user interface for designing, configuring, and automating network experiments. We explain TeachNet’s design and present results to demonstrate its effectiveness and efficiency using our TCP experimentation-focused prototype, and Mininet in the backend. TeachNet creates a repository of log files and offers a framework for future research that is easily extensible to accommodate other tools such as OMNeT++ and ns3.

• Teymoori, Peyman; Welzl, Michael & Hayes, David Andrew (2023). LGCC: A Novel High-Throughput and Low Delay Paradigm Shift in Multi-Hop Congestion Control. IEEE/ACM Transactions on Networking. ISSN 1063-6692. 32(1), s. 761–776. doi: 10.1109/TNET.2023.3301291. Fulltekst i vitenarkiv Vis sammendrag

  Technological advancements have provided wireless links with very high data rate capacity for 5G/6G mobile networks and WiFi 6, which will be widely deployed by 2025. However, the capacity can have substantial fluctuations, violating the assumption at the transport layer that the capacity is (almost) steady. In this paper, we present a general and efficient, yet deployable solution to this problem through a novel design empowered with a rich theory, allowing a significantly improved experience in using new technologies, especially mobile cellular services. We employ the well-known theory of food-chain models in biology, where a bottleneck link can be modeled as prey, while flows are predators. We extend this model to a chain of predators and preys to form a multi-hop congestion controller, called LGCC. Through simulation evaluation with real-life 5G traces we show the effectiveness of LGCC, compared with the state-of-the-art ABC (Accel-Brake Control). Our results show an order of magnitude bottleneck queuing delay decrease, with only a small decrease in throughput because LGCC tries to never exceed link capacities. LGCC’s design can additionally open a new paradigm in stable multi-hop congestion control and flow aggregation.

• Welzl, Michael; Ott, Jörg; Perkins, Colin; Islam, Safiqul & Kutscher, Dirk (2023). How not to IETF: Lessons Learned From Failed Standardization Attempts, 2023 IEEE International Conference on Pervasive Computing and Communications Workshops and other Affiliated Events (PerCom Workshops). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9781665453813. doi: 10.1109/PerComWorkshops56833.2023.10150250. Fulltekst i vitenarkiv Vis sammendrag

  Protocol standards work is an interesting mixture of technical, political, financial, and human factors. Standardization processes require stamina as they may be lengthy, and they demand frustration resistance as they may hold surprises at all stages. While this certainly bears some similarity to academic endeavors, the need to build broader consensus and the potential of far reaching industry impact, among other factors, lead to different incentives and value systems. Peer review perspectives may also differ notably. In this paper, we discuss issues we came across in the past when trying to develop and advance technologies in the IETF or push presumed solid technology solutions towards standardization. We summarize our personal perspectives on the lessons learned.

• Ciko, Kristjon; Welzl, Michael & Teymoori, Peyman (2023). Going Dark: A Software "Light Switch" for Internet Servers. IEEE Workshop on Local and Metropolitan Area Networks. ISSN 1944-0367. doi: 10.1109/LANMAN58293.2023.10189419. Fulltekst i vitenarkiv Vis sammendrag

  To fight global warming, carbon emissions must urgently be reduced. In this paper, we look at opportunities to do so by diminishing the operational energy usage of an important always-on element of the Internet: server systems. Our measurements from a server host indicate that there is probably not much to be gained by making a significant software change - but, depending on the system, a simple, easily overlooked configuration update may make a difference at no performance cost. This difference is small per server (in the order of an LED light bulb). Given the multitude of permanently operational servers in the world, it may however be significant at scale.

• Welzl, Michael (2022). Not a Trade-Off: On the Wi-Fi Energy Efficiency of Effective Internet Congestion Control, 2022 17th Wireless On-Demand Network Systems and Services Conference (WONS). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9783903176461. doi: 10.23919/WONS54113.2022.9764413. Fulltekst i vitenarkiv
• Yuan, Gina; Zhang, David; Sotoudeh, Matthew; Welzl, Michael & Winstein, Keith (2022). Sidecar: In-Network Performance Enhancements in the Age of Paranoid Transport Protocols, HotNets '22: Proceedings of the 21st ACM Workshop on Hot Topics in Networks. Association for Computing Machinery (ACM). ISSN 9781450398992. s. 221–227. doi: 10.1145/3563766.3564113. Fulltekst i vitenarkiv
• Islam, Safiqul; Hiorth, Kristian Andreas; Griwodz, Carsten & Welzl, Michael (2022). Is it really necessary to go beyond a fairness metric for next-generation congestion control? ANRW '22: Proceedings of the Workshop on Applied Networking Research. Association for Computing Machinery (ACM). ISSN 9781450394444. s. 1–7. doi: 10.1145/3547115.3547192. Fulltekst i vitenarkiv Vis sammendrag

  A recent article suggests that the potential for deployment of congestion control mechanisms in the future Internet should be evaluated using a new concept called “harm” instead of measuring “fairness”. While there are good arguments in favor of this new approach, its practical benefits have not yet been experimentally evaluated, and calculating harm requires producing more experimental data. We apply the harm concept to data produced in real-life experiments with competing pairs of various TCP variants: Cubic vs. Reno, BBR vs. Cubic, and Reno vs. Vegas. These experiments cover various levels of “aggression” as well as different feedback types that the controls are based upon. We present a new linear representation of relative harm between scenarios, which can help us to assess the differences in harm between a variety of situations. Among other results, we can see that BBR is on average 1.6 times more harmful to Cubic in high- BDP situations (when Cubic is most aggressive) than Cubic is to Reno.

• Ciko, Kristjon; Teymoori, Peyman & Welzl, Michael (2022). LGC-ShQ: Datacenter Congestion Control with Queueless Load-Based ECN Marking. Computer communication review. ISSN 0146-4833. 52(4), s. 2–11. doi: 10.1145/3577929.3577931. Fulltekst i vitenarkiv Vis sammendrag

  We present LGC-ShQ, a new ECN-based congestion control mechanism for datacenters. LGC-ShQ relies on ECN feedback from a Shadow Queue, and it uses ECN not only to decrease the rate, but it also increases the rate in relation to this signal. Real-life tests in a Linux testbed show that LGC-ShQ keeps the real queue at low levels while achieving good link utilization and fairness.

• Islam, Md Safiqul; Welzl, Michael & Fladby, Tobias (2022). Real-Life Implementation and Evaluation of Coupled Congestion Control for WebRTC Media and Data Flows. IEEE Access. 10, s. 95046–95066. doi: 10.1109/ACCESS.2022.3206041. Fulltekst i vitenarkiv Vis sammendrag

  WebRTC enables users to simultaneously transfer media (over the Real-Time Transport Protocol (RTP)) and data (over the Stream Control Transmission Protocol (SCTP)) between web browsers, multiplexed onto a single UDP port pair. This design choice of using two different transport protocols, each with their own congestion control mechanism, can lead to competition between the flows, resulting in undesirable spikes in queuing delay and packet loss. In this paper, we investigate solutions to the harmful effects WebRTC flows cause on each other by having the different congestion controllers of the flows collaborate. Using implementations in the Chromium browser, we show that our mechanism can combine a set of heterogeneous congestion control mechanisms, fairly allocate the available bandwidth between the flows, and reduce overall delay and losses.

• Welzl, Michael; Teymoori, Peyman; Islam, Md Safiqul; Hutchison, David & Gjessing, Stein (2022). Future Internet Congestion Control: The Diminishing Feedback Problem. IEEE Communications Magazine. ISSN 0163-6804. 60(9), s. 87–92. doi: 10.1109/MCOM.006.2200008. Fulltekst i vitenarkiv Vis sammendrag

  It is increasingly difficult for Internet congestion control mechanisms to obtain the feedback that they need. This lack of feedback can have severe performance implications, and it is bound to become worse. In the long run, the problem may only be fixable by fundamentally changing the way congestion control is done in the Internet. We substantiate this claim by looking at the evolution of the Internet's infrastructure over the past 30 years, and by examining the most common behavior of Internet traffic. Considering the goals that congestion control mechanisms are intended to address, and taking into account contextual developments in the Internet ecosystem, we arrive at conclusions and recommendations about possible future congestion control design directions. In particular, we argue that congestion control mechanisms should move away from their strict “end-to-end” adherence. This change would benefit from avoiding a “one size fits all circumstances” approach, and moving toward a more selective set of mechanisms that will result in a better performing Internet. We also discuss how this future vision differs from today's use of performance enhancing proxies.

• Ciko, Kristjon; Welzl, Michael & Teymoori, Peyman (2021). PEP-DNA: A Performance Enhancing Proxy for Deploying Network Architectures. I Gunes, Mehmet & Lee, Patrick P.C. (Red.), 2021 IEEE 29th International Conference on Network Protocols (ICNP 2021). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9781665441315. doi: 10.1109/ICNP52444.2021.9651953. Fulltekst i vitenarkiv
• Welzl, Michael; Islam, Safiqul; Gundersen, Michael & Fischer, Andreas (2021). Transport Services: A Modern API for an Adaptive Internet Transport Layer. IEEE Communications Magazine. ISSN 0163-6804. 59(4), s. 16–2242. doi: 10.1109/MCOM.001.2000870. Fulltekst i vitenarkiv Vis sammendrag

  Transport services (TAPS) is a working group of the the Internet Engineering Task Force (IETF). TAPS defines a new recommended API for the Internet's transport layer. This API gives access to a wide variety of services from various protocols, and it is protocol-independent: the transport layer becomes adaptive, and applications are no longer statically bound to a particular protocol and/or network interface. We give an overview of the TAPS API, and we demonstrate its flexibility and ease of use with an example using a Python-based open source implementation.

• Teymoori, Peyman; Hayes, David Andrew; Welzl, Michael & Gjessing, Stein (2021). Estimating an Additive Path Cost With Explicit Congestion Notification. IEEE Transactions on Control of Network Systems. 8(2), s. 859–871. doi: 10.1109/TCNS.2021.3053179. Fulltekst i vitenarkiv Vis sammendrag

  Network utility maximization is a well-accepted theoretical framework that describes how congestion controls cooperate to achieve an ideal sending rate allocation, for given utility functions of senders and constraints of the network. These network constraints are expressed as a “cost” in the framework. In practice, most congestion control mechanisms obtain feedback that is different from the framework's “cost.” This article focuses on explicit congestion notification (ECN), which has been shown to be advantageous when it is available, e.g., with the popular datacenter transmission control protocol mechanism. However, different from the framework's cost, ECN marks are not additive. We present a practical solution to this problem; it changes how end hosts interpret the ECN signal, while for routers, a special configuration of random early detection is used.

• Hayes, David Andrew; Welzl, Michael; Ferlin, Simone; Ros, David & Islam, Safiqul (2020). Online Identification of Groups of Flows Sharing a Network Bottleneck. IEEE/ACM Transactions on Networking. ISSN 1063-6692. 28(5), s. 2229–2242. doi: 10.1109/TNET.2020.3007346. Fulltekst i vitenarkiv Vis sammendrag

  Most Internet hosts today support multiple access technologies and network interfaces. Multipath transport protocols, like MPTCP, are being deployed (e.g., in smartphones), allowing transparent simultaneous use of multiple links. Besides providing increased resilience to link failures, multipath transports may better exploit available (aggregate) capacity across all interfaces. The safest way to ensure fairness is to assume that any subflows of a multipath end-to-end connection may share bottleneck links, but knowledge of non-shared bottlenecks could allow multipath senders to exploit more capacity without being unfair to other flows. The problem of reliably detecting the existence of (non)-shared bottlenecks is not trivial and is compounded by the fact that bottlenecks may change due to traffic dynamics. In this paper we focus on practical methods to reliably group flows that share, possibly dynamic, bottlenecks online and in a passive manner (i.e., without injecting measurement traffic). We introduce a novel dynamic clustering algorithm that we apply to update our previous shared bottleneck flow grouping (SBFG) method standardized by the IETF, based on delay statistics. We also adapt an offline SBFG method based on wavelet filters to enable it for online operation. These SBFG methods are evaluated by a simple testbed, rigorous simulation and real-world Internet experiments in a testbed comprised of multihomed hosts. Our results suggest that there is no clear winner, and selection of the “best” SBFG method will have to consider tradeoffs regarding accuracy, lag, and application requirements.

• Welzl, Michael; Teymoori, Peyman; Gjessing, Stein & Islam, Safiqul (2020). Follow the Model: How Recursive Networking Can Solve the Internet's Congestion Control Problems, 2020 International Conference on Computing, Networking and Communications (ICNC). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9781728149059. s. 518–524. doi: 10.1109/ICNC47757.2020.9049648. Fulltekst i vitenarkiv Vis sammendrag

  The Recursive InterNetworking Architecture RINA describes a new way to look at networking; it offers a point of view that is fundamentally different from today's networks. This paper explains how designing congestion control strictly in line with this model almost automatically leads to a conceptually cleaner, and quite possibly altogether better design than what we have in the Internet today. We give an overview of how far the OCARINA research project has come with the development of the RINA congestion control elements that follow from this design. Then, we conclude with an explanation of how the shift in thinking that RINA suggests can be applied to more gradual Internet developments related to congestion control.

• Barik, Runa; Welzl, Michael; Fairhurst, Gorry; Elmokashfi, Ahmed; Dreibholz, Thomas & Gjessing, Stein (2020). On the Usability of Transport Protocols other than TCP: A Home Gateway and Internet Path Traversal Study. Computer Networks. ISSN 1389-1286. 173. doi: 10.1016/j.comnet.2020.107211. Fulltekst i vitenarkiv Vis sammendrag

  Network APIs are moving towards protocol agility, where applications express their needs but not a static protocol binding, and it is up to the layer below the API to choose a suitable protocol. The IETF Transport Services (TAPS) Working Group is standardizing a protocol-independent transport API and offering guidance to implementers. Apple’s recent “Network.framework” is specifically designed to allow such late and dynamic binding of protocols. When the network stack autonomously chooses and configures a protocol, it must first test which protocols are locally available and which work end-to-end (“protocol racing”). For this, it is important to know the set of available options, and which protocols should be tried first: Does it make sense to offer unchecked payload delivery, as with UDP-Lite? Is a UDP-based protocol like QUIC always a better choice, or should native SCTP be tried? This paper develops answers to such questions via (i) a NAT study in a local testbed, (ii) bidirectional Internet tests, (iii) a large scale Internet measurement campaign. The examined protocols are: SCTP, DCCP, UDP-Lite, UDP with a zero checksum and three different UDP encapsulations.

• Barik, Runa; Welzl, Michael; Teymoori, Peyman; Islam, Safiqul & Gjessing, Stein (2020). Performance Evaluation of In-network Packet Retransmissions using Markov Chains, 2020 International Conference on Computing, Networking and Communications (ICNC). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9781728149059. doi: 10.1109/ICNC47757.2020.9049757. Fulltekst i vitenarkiv Vis sammendrag

  We consider the retransmission of packets inside the network on a segment of an end-to-end path. Using a Markov chain formulation of the problem, we evaluate the effect of the loss probability in the network segment that is responsible for retransmission as well as the ratio between the local RTT and the end-to-end timeout. We also obtain the optimal cache size required for retransmission depending on the packet loss and cache blocking probabilities. Our study reveals that the local RTT of the path segment for retransmission as a function of the end-to-end timeout significantly influences the packet caching time. We also observe that a small increment in the packet loss probability above a threshold can severely affect the percentage of cache filling irrespective of the cache size. Use cases to which our findings apply include a recent IETF proposal called “LOOPS”, recursive network architectures and proposals that use hop-by-hop retransmissions in special network scenarios, e.g. for Information-Centric Networking (ICN).

• Ciko, Kristjon; Welzl, Michael & Marek, Marcel (2020). TAPS and RINA: Do they fit together? I Chatras, Bruno & Lopez, Diego (Red.), 2020 23rd Conference on Innovation in Clouds, Internet and Networks and Workshops (ICIN). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9781728151274. s. 35–40. doi: 10.1109/ICIN48450.2020.9059406. Fulltekst i vitenarkiv
• Hiorth, Kristian Andreas & Welzl, Michael (2019). Design Considerations for RINA Congestion Control over WiFi Links. I Galis, Alex; Guillemin, Fabrice; Noldus, Rogier; Secci, Stefano; Idzikowski, Filip & Sayit, Müge Fesci (Red.), 2019 22nd Conference on Innovation in Clouds, Internet and Networks and Workshops (ICIN). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9781538683361. s. 54–59. doi: 10.1109/ICIN.2019.8685874. Fulltekst i vitenarkiv
• Marek, Marcel; Teymoori, Peyman; Gjessing, Stein & Welzl, Michael (2019). High-Rate Data Transfer with Congestion-Aware Multipath Routing. I Unknown, Person (Red.), IEEE GLOBECOM 2019 Workshops. IEEE (Institute of Electrical and Electronics Engineers). ISSN 9781728109619. doi: 10.1109/GCWkshps45667.2019.9024564. Fulltekst i vitenarkiv
• Ciko, Kristjon & Welzl, Michael (2019). First Contact: Can Switching to RINA save the Internet? I Galis, Alex; Guillemin, Fabrice; Noldus, Rogier; Secci, Stefano; Filip, Idzikowski & Sayit, Müge Fesci (Red.), 22nd Conference on Innovation in Clouds, Internet and Networks and Workshops, ICIN 2019. IEEE (Institute of Electrical and Electronics Engineers). ISSN 9781538683361. s. 37–42. doi: 10.1109/ICIN.2019.8685912. Fulltekst i vitenarkiv
• Bianchi, Giuseppe; Welzl, Michael; Tulumello, Angelo; Gringoli, Francesco; Belocchi, Giacomo & Faltelli, Marco [Vis alle 7 forfattere av denne artikkelen] (2019). XTRA: Towards portable transport layer functions. IEEE Transactions on Network and Service Management. 16(4), s. 1507–1521. doi: 10.1109/TNSM.2019.2948467. Fulltekst i vitenarkiv Vis sammendrag

  XTRA (XFSM for Transport) aims at providing a first attempt towards a “code-once-port-everywhere” platform-agnostic programming abstraction tailored to the deployment of transport layer functions. XTRA's programming abstraction not only fits SW platforms, but is specifically designed to harness, with no re-coding effort, the offloading opportunities offered by CPU-less HW boards or smart NICs. We demonstrate the viability of XTRA with three completely different implementations of the underlying execution engine (HW proof-of-concept on a NetFPGA board, User-space SW over Linux' Open Data Plane, and NS3 emulator). Flexibility is shown via a number of example applications, ranging from a variety of congestion control algorithms, to a middlebox-type TCP proxy functionality, up to a customized “Timer-Based” (TB) TCP which leverages the native reliance of XTRA on timers, so as to produce a loss recovery operation which, despite being formalized only via a handful of code lines, performs almost comparable with the highly optimized Linux and FreeBSD implementations.

• Barik, Runa; Welzl, Michael; Elmokashfi, Ahmed; Dreibholz, Thomas; Islam, Safiqul & Gjessing, Stein (2019). On the Utility of Unregulated IP DiffServ Code Point (DSCP) Usage by End Systems. Performance evaluation (Print). ISSN 0166-5316. 135. doi: 10.1016/j.peva.2019.102036. Fulltekst i vitenarkiv Vis sammendrag

  DiffServ was designed to implement service provider quality of service (QoS) policies, where routers change and react upon the DiffServ Code Point (DSCP) in the IP header. However, nowadays, applications are beginning to directly set the DSCP themselves, in the hope that this will yield a more appropriate service for their respective video, audio and data streams. WebRTC is a prime example of such an application. We present measurements, for both IPv4 and IPv6, of what happens to DSCP values along Internet paths after an end system has set them without any prior agreement between a customer and a service provider. We find that the DSCP is often changed or zeroed along the path, but detrimental effects from using the DSCP are extremely rare; moreover, DSCP values sometimes remain intact (potentially having an effect on traffic) for several AS hops. This positive result motivates an analysis of the potential latency impact from such DSCP usage, for which we present the first measurement results. We find that routers at approximately 3% of more than 100,000 links differentiate between the WebRTC DSCP values (EF, AF42 and CS1) and consistently reduce delay in comparison with probes carrying a zero value (CS0) under congestion. In contrast, routers at around 2% of these links increase the delay by a comparable amount under congestion, uniformly for EF, AF42 and CS1.

• Pontarelli, Salvatore; Bianchi, Giuseppe & Welzl, Michael (2019). A Programmable Hardware Calendar for High Resolution Pacing, IEEE International Conference on High Performance Switching and Routing (HPSR) 2018. IEEE Sarnoff Symposium. ISSN 9781538655412. s. 1–6. doi: 10.1109/HPSR.2018.8850731. Fulltekst i vitenarkiv Vis sammendrag

  The challenge addressed in this paper consists in offloading packet-based pacing to a hardware Network Interface Card, while retaining the flexibility of software timers. In this direction, we propose, design, implement, and evaluate a hardware calendar, which can be programmed via a simple yet very flexible programming interface leveraging stateful (adaptive) per-packet timers. We show, for both specific examples (exponential, linearly increasing, etc) as well as for the general case, how to derive such a per-packet timer setting from a high-level desired rate envelope. Further, we describe and evaluate an FPGA implementation which relies on a novel insertion strategy for solving collisions in the calendar’s hash table.

• Ijaz, Humaira; Jameel, Bushra & Welzl, Michael (2019). A survey and comparison on overlay‐underlay mapping techniques in peer‐to‐peer overlay networks. International Journal of Communication Systems. ISSN 1074-5351. 32(3), s. 1–26. doi: 10.1002/dac.3872. Fulltekst i vitenarkiv Vis sammendrag

  Peer‐to‐peer (P2P) overlay networks were developed initially for file sharing such as Napster and Gnutella; but later, they have become popular for content sharing, media streaming, telephony applications, etc. Underlay‐unawareness in P2P systems can result in suboptimal peer selection for overlay routing and hence poor performance. In this paper, we present a comprehensive survey of the research work carried out to solve the overlay‐underlay mapping problems up till now. The majority of underlay‐aware proposals for peer selection focus on finding the shortest overlay routes by selecting nearest nodes according to proximity information. Another class of approaches is based on passive or active probing for provision of underlay information to P2P applications. Some other optimizations propose use of P2P middleware to extract, process, and refine underlay information and provide it to P2P overlay applications. Another class of approaches strive to use ISPs or third parties to provide underlay information to P2P overlay applications according to their requirements. We have made a state‐of‐the‐art review and comparison for addressing the overlay‐underlay mismatch in terms of their operation, merits, limitations, and future directions.

• Islam, Safiqul; Welzl, Michael & Gjessing, Stein (2019). How to Control a TCP: Minimally-Invasive Congestion Management for Datacenters. I Ansari, Nirwan & Wang, Honggang (Red.), Proceedings of the 2019 International Conference on Computing, Networking and Communications (ICNC): Green Computing, Networking, and Communications. IEEE Sarnoff Symposium. ISSN 9781538692233. s. 121–125. doi: 10.1109/ICCNC.2019.8685516. Fulltekst i vitenarkiv Vis sammendrag

  In multi-tenant datacenters, the hardware may be homogeneous but the traffic often is not. For instance, customers who pay an equal amount of money can get an unequal share of the bottleneck capacity when they do not open the same number of TCP connections. To address this problem, several recent proposals try to manipulate the traffic that TCP sends from the VMs. VCC and AC/DC are two new mechanisms that let the hypervisor control traffic by influencing the TCP receiver window (rwnd). This avoids changing the guest OS, but has limitations (it is not possible to make TCP increase its rate faster than it normally would). Seawall, on the other hand, completely rewrites TCP's congestion control, achieving fairness but requiring significant changes to both the hypervisor and the guest OS. There seems to be a need for a middle ground: a method to control TCP's sending rate without requiring a complete redesign of its congestion control. We introduce a minimally-invasive solution that is flexible enough to cater for needs ranging from weighted fairness in multi-tenant datacenters to potentially offering Internet-wide benefits from reduced interflow competition.

• Welzl, Michael; Islam, Safiqul; Barik, Runa; Gjessing, Stein & Elmokashfi, Ahmed Mustafa Abdalla (2019). Investigating the Delay Impact of the DiffServ Code Point (DSCP). I Ansari, Nirwan & Wang, Honggang (Red.), Proceedings of the 2019 International Conference on Computing, Networking and Communications (ICNC): Green Computing, Networking, and Communications. IEEE Sarnoff Symposium. ISSN 9781538692233. s. 612–616. doi: 10.1109/ICCNC.2019.8685538. Fulltekst i vitenarkiv Vis sammendrag

  The DiffServ Code Point (DSCP) field in the IP header allows to specify a desired per-hop behavior as packets traverse routers. Setting the DSCP field opportunistically, without prior contractual agreement, has recently become accepted practice for Internet end hosts. Measurement studies find that there is reason to hope for a DSCP setting to have an effect on traffic, and at least configuring this value is not heavily detrimental: systematic drops of packets due to non-zero DSCP values are rare, and the value is often left intact along an end-to-end path. What these studies do not discuss is whether per-hop behaviors truly are honored: what happens to packets in terms of the delay they experience? In this paper, we make an attempt to find a first answer to this by mining a dataset of our own recent large-scale measurement study. Using a deep neural network, we obtain the importance of the factors which help us understand the delay impact of the DSCP.

• Islam, Safiqul; Welzl, Michael & Gjessing, Stein (2018). Lightweight and flexible single-path congestion control coupling, NOMS 2018 - 2018 IEEE/IFIP Network Operations and Management Symposium. IEEE Sarnoff Symposium. ISSN 9781538634165. s. 1–6. doi: 10.1109/NOMS.2018.8406137. Fulltekst i vitenarkiv Vis sammendrag

  Communication between two Internet hosts using parallel connections may result in unwanted interference between the connections. In this dissertation, we propose a sender-side solution to address this problem by letting the congestion controllers of the different connections collaborate, correctly taking congestion control logic into account. Real-life experiments and simulations show that our solution works for a wide variety of congestion control mechanisms, provides great flexibility when allocating application traffic to the connections, and results in lower queuing delay and less packet loss.

• Islam, Safiqul; Welzl, Michael; Hiorth, Kristian Andreas; Hayes, David; Armitage, Grenville & Gjessing, Stein (2018). ctrlTCP: Reducing latency through coupled, heterogeneous multi-flow TCP congestion control, IEEE INFOCOM 2018 - IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS). IEEE Sarnoff Symposium. ISSN 9781538659793. s. 214–219. doi: 10.1109/INFCOMW.2018.8406887. Fulltekst i vitenarkiv Vis sammendrag

  We present ctrlTCP, a method to combine the congestion controls of multiple TCP connections. In contrast to the previous methods such as the Congestion Manager, ctrlTCP can couple all TCP flows that leave one sender, traverse a common bottleneck (e.g., a home user's thin uplink) and arrive at different destinations. Using ns-2 simulations and an implementation in the FreeBSD kernel, we show that our mechanism reduces queuing delay, packet loss, and short flow completion times while enabling precise allocation of the share of the available bandwidth between the connections according to the needs of the applications.

• Barik, Runa; Welzl, Michael; Elmokashfi, Ahmed Mustafa Abdalla; Dreibholz, Thomas & Gjessing, Stein (2018). Can WebRTC QoS Work? A DSCP Measurement Study. I Reichl, Peter; Hossfeld, Tobias; Guiseppe, Bianchi; Zinner, Tobias & Altman, Eitan (Red.), ITC 30 - Teletraffic in a Smart World. IEEE Press. ISSN 9780988304550. s. 167–175. doi: 10.1109/ITC30.2018.00034. Fulltekst i vitenarkiv
• Marek, Marcel; Teymoori, Peyman; Welzl, Michael & Gjessing, Stein (2018). Computer-Aided Reproducibility. I IEEE, . (Red.), Proceedings of International Conference on Computing, Networking and Communications (ICNC 2018). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9781111111113. s. 127–133. doi: 10.1109/ICCNC.2018.8390274. Fulltekst i vitenarkiv Vis sammendrag

  Computer networks research has been notoriously bad at reproducibility – a key aspect of making research results credible and convincing. This has been attributed to a lack of incentive for researchers to share the data underlying scientific results. We conjecture that this can be helped by reducing the amount of work that is required to make results reproducible. This paper introduces CAR – a system for “Computer-Aided Reproducibility”. Similar to other forms of “Computer-Aided- *”, our CAR tool facilitates the process of sharing the necessary data by partially automating it. © 2018 Institute of Electrical and Electronics Engineers (IEEE)

• Bianchi, Giuseppe; Welzl, Michael; Tulumello, Angelo; Belocchi, Belocci; Faltelli, Marco & Pontarelli, Salvatore (2018). A fully portable TCP implementation using XFSMs, Proceedings of ACM SIGCOMM 2018 Conference on Posters and Demos. Association for Computing Machinery (ACM). ISSN 9781450359153. s. 99–101. doi: 10.1145/3234200.3234237. Fulltekst i vitenarkiv
• Papastergiou, Giorgos; Fairhurst, Gorry; Ros, David; Brunstrom, Anna; Grinnemo, Karl-Johan & Hurtig, Per [Vis alle 11 forfattere av denne artikkelen] (2017). De-Ossifying the Internet Transport Layer: A Survey and Future Perspectives. IEEE Communications Surveys and Tutorials. 19(1), s. 619–639. doi: 10.1109/COMST.2016.2626780. Fulltekst i vitenarkiv Vis sammendrag

  It is widely recognized that the Internet transport layer has become ossified, where further evolution has become hard or even impossible. This is a direct consequence of the ubiquitous deployment of middleboxes that hamper the deployment of new transports, aggravated further by the limited flexibility of the application programming interface (API) typically presented to applications. To tackle this problem, a wide range of solutions have been proposed in the literature, each aiming to address a particular aspect. Yet, no single proposal has emerged that is able to enable evolution of the transport layer. In this paper, after an overview of the main issues and reasons for transport-layer ossification, we survey proposed solutions and discuss their potential and limitations. The survey is divided into five parts, each covering a set of point solutions for a different facet of the problem space: (1) designing middlebox-proof transports; (2) signaling for facilitating middlebox traversal; (3) enhancing the API between the applications and the transport layer; (4) discovering and exploiting end-to-end capabilities; and (5) enabling user-space protocol stacks. Based on this analysis, we then identify further development needs toward an overall solution. We argue that the development of a comprehensive transport layer framework, able to facilitate the integration and cooperation of specialized solutions in an application-independent and flexible way, is a necessary step toward making the Internet transport architecture truly evolvable. To this end, we identify the requirements for such a framework and provide insights for its development.

• Khademi, Naeem; Welzl, Michael; Armitage, Grenville; Ros, David; Zander, Sebastian & Fairhurst, Gorry (2017). Alternative Backoff: Achieving Low Latency and High Throughput with ECN and AQM, 16th International IFIP TC6 Networking Conference, Networking 2017. IFIP Open Digital Library. ISSN 9783901882944. doi: 10.23919/ifipnetworking.2017.8264863. Fulltekst i vitenarkiv Vis sammendrag

  A number of recently proposed Active Queue Management (AQM) mechanisms instantiate shallow buffers with burst tolerance to minimise the time that packets spend enqueued at a bottleneck. However, shallow buffering causes noticeable TCP performance degradation as a path’s underlying round trip time (RTT) heads above typical intra-country levels. Using less-aggressive multiplicative backoffs in TCP can compensate for shallow bottleneck buffering. AQM mechanisms may either drop packets or mark them using Explicit Congestion Notification (ECN), depending on whether the sender marked packets as ECN-capable. While a drop may therefore stem from any type of queue, an ECN-mark indicates that an AQM mechanism has done its job, and therefore the queue is likely to be shallow. We propose ABE: "Alternative Backoff with ECN", which consists of enabling ECN and letting individual TCP senders back off less aggressively in reaction to ECN-marks from AQM-enabled bottlenecks. Using controlled testbed experiments with standard NewReno and CUBIC flows, we show significant performance gains in lightly-multiplexed scenarios, without losing the delay-reduction benefits of deploying AQM. ABE is a sender-side-only modification that can be deployed across networks incrementally (requiring no flag-day) and offers a compelling reason to deploy and enable ECN across the Internet.

• Khademi, Naeem; Ros, David; Welzl, Michael; Bozakov, Zdravko; Brunstrom, Anna & Fairhurst, Gorry [Vis alle 13 forfattere av denne artikkelen] (2017). NEAT: A Platform- and Protocol-Independent Internet Transport API. IEEE Communications Magazine. ISSN 0163-6804. 55(6), s. 46–54. doi: 10.1109/MCOM.2017.1601052. Fulltekst i vitenarkiv Vis sammendrag

  The sockets API has become the standard way that applications access the transport services offered by the IP stack. This article presents NEAT, a user space library that can provide an alternate transport API. NEAT allows applications to request the service they need using a new design that is agnostic to the specific choice of transport protocol underneath. This not only allows applications to take advantage of common protocol machinery, but also eases introduction of new network mechanisms and transport protocols. The article describes the components of the NEAT library and illustrates the important benefits that can be gained from this new approach. NEAT is a software platform for developing advanced network applications that was designed in accordance with the standardization efforts on transport services in the IETF, but its features exceed the envisioned functionality of a TAPS system. © 2017 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works

• Barik, Runa; Divakaran, Dinil Mon & Welzl, Michael (2017). Characterizing the Effects of TCP's Initial Window. I Vuran, Mehmet Can & Xie, Linda Jiang (Red.), 2017 IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS). IEEE Sarnoff Symposium. ISSN 9781538627846. s. 886–891. doi: 10.1109/INFCOMW.2017.8116493. Fulltekst i vitenarkiv
• Barik, Runa; Welzl, Michael; Elmokashfi, Ahmed Mustafa Abdalla; Gjessing, Stein & Islam, Safiqul (2017). fling: A Flexible Ping for Middlebox Measurements. I Davoli, Franco & Jain, Nikhil (Red.), Teletraffic Congress (ITC 29), 2017 29th International. IEEE Sarnoff Symposium. ISSN 9780988304536. s. 134–142. doi: 10.23919/ITC.2017.8064349. Fulltekst i vitenarkiv Vis sammendrag

  Middleboxes in private networks have been known to change packets in many ways, making it hard to design protocol extensions that work for the large majority of Internet users. Addressing the need to know what such middleboxes do, we introduce a tool called fling ("flexible ping"). fling can carry out (almost) any kind of protocol dialogue between a server and a client based on a simple specification in a json and a pcap file, and identify what middleboxes do to the packets of the dialogue. This fills a gap in the state of the art, where other tools that control both ends of a path are either limited in some form or have to be updated for every new test. We present results from small-scale tests that prove the flexibility of fling, which is a prerequisite for our next step: development of a large-scale measurement platform.

• Hayes, David A.; Teymoori, Peyman & Welzl, Michael (2016). Feedback in recursive congestion control. Lecture Notes in Computer Science (LNCS). ISSN 0302-9743. 9951 LNCS, s. 109–125. doi: 10.1007/978-3-319-46433-6_8. Fulltekst i vitenarkiv
• Islam, Safiqul & Welzl, Michael (2016). Start Me Up: Determining and Sharing TCP's Initial Congestion Window, Proceedings of the 2016 Applied Networking Research Workshop. Association for Computing Machinery (ACM). ISSN 9781450344432. s. 52–54. doi: 10.1145/2959424.2959440. Fulltekst i vitenarkiv Vis sammendrag

  When multiple TCP connections are used between the same host pair, they often share a common bottleneck – especially when they are encapsulated together, e.g. in VPN scenarios. Then, all connections after the first should not have to guess the right initial value for the congestion window, but rather get the appropriate value from other connections. This allows short flows to complete much faster – but it can also lead to large bursts that cause problems on their own. Prior work used timer-based pacing methods to alleviate this problem; we introduce a new algorithm that “paces” packets by instead correctly maintaining the ACK clock, and show its positive impact in combination with a previously presented congestion coupling algorithm.

• Barik, Runa; Welzl, Michael & Elmokashfi, Ahmed (2016). How to say that you're special: Can we use bits in the IPv4 header? Proceedings of the 2016 Applied Networking Research Workshop. Association for Computing Machinery (ACM). ISSN 9781450344432. s. 68–70. doi: 10.1145/2959424.2959442. Fulltekst i vitenarkiv Vis sammendrag

  The IP header should be the ideal part of a packet that an end sys- tem could use to ask the network for special treatment. Recently, there has been renewed interest in using bits of this header – e.g. the ECN and the DSCP fields. But can we really use these bits? Or should we try to use other bits? We contribute to the body of work that tries to answer these questions by reporting on IPv4 measure- ments regarding the DSCP field and the Evil bit. Our findings show unexpected treatment to packets that set either of these fields and also confirm recent results on IP Options and ECN.

• Teymoori, Peyman; Hayes, David Andrew; Welzl, Michael & Gjessing, Stein (2016). Even Lower Latency, Even Better Fairness: Logistic Growth Congestion Control in Datacenter. I Kanhere, Salil & Mnaouer, Adel Ben (Red.), Proceedings 2016 IEEE 41st Conference on Local Computer Networks (LCN) 2016. IEEE (Institute of Electrical and Electronics Engineers). ISSN 9781509020546. s. 10–18. doi: 10.1109/LCN.2016.12. Fulltekst i vitenarkiv
• Teymoori, Peyman; Welzl, Michael; Gjessing, Stein; Grasa, Eduard; Riggio, Roberto & Rausch, Kewin [Vis alle 7 forfattere av denne artikkelen] (2016). Congestion control in the recursive InterNetworking Architecture (RINA). I Bregni, Stefano & Fonseca, Nelson (Red.), IEEE International Conference on Communications (ICC 2016). IEEE Sarnoff Symposium. ISSN 9781479966646. doi: 10.1109/ICC.2016.7510818. Fulltekst i vitenarkiv
• Islam, Safiqul; Welzl, Michael; Hayes, David Andrew & Gjessing, Stein (2016). Managing Real-Time Media Flows through a Flow State Exchange. I Badonnel, Semma (Red.), Proceedings of NOMS 2016 - 2016 IEEE/IFIP Network Operations and Management Symposium. IEEE (Institute of Electrical and Electronics Engineers). ISSN 9781509002238. s. 112–120. doi: 10.1109/NOMS.2016.7502803. Fulltekst i vitenarkiv
• Barik, Runa; Welzl, Michael; Ferlin, Simone & Alay, Ozgu (2016). LISA: A linked slow-start algorithm for MPTCP. I Bregni, Stefano & Fonseca, Nelson (Red.), IEEE International Conference on Communications (ICC 2016). IEEE Sarnoff Symposium. ISSN 9781479966646. doi: 10.1109/ICC.2016.7510786. Fulltekst i vitenarkiv
• Islam, Safiqul; Welzl, Michael; Gjessing, Stein & You, Jianjie (2016). OpenTCP: Combining congestion controls of parallel TCP connections. I Cheng, Xi'an Guojian (Red.), 2016 IEEE Advanced Information Management, Communicates, Electronic and Automation Control Conference (IMCEC). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9781467396134. s. 194–198. doi: 10.1109/IMCEC.2016.7867199. Fulltekst i vitenarkiv
• Briscoe, Bob; Brunstrom, Anna; Petlund, Andreas; Hayes, David Andrew; Ros, David & Tsang, Ing-Jyh [Vis alle 10 forfattere av denne artikkelen] (2016). Reducing internet latency: a survey of techniques and their merits. IEEE Communications Surveys and Tutorials. 18(3), s. 2149–2196. doi: 10.1109/COMST.2014.2375213. Fulltekst i vitenarkiv
• Welzl, Michael; Gjessing, Stein & Jahre, Håkon Botnmark (2016). Fighting fire with fire: Eliminating standing queues with large UDP packet floods. I Yang, Liuqing (Red.), Proceedings of International Conference on Computing, Networking and Communications (ICNC), 2016. IEEE (Institute of Electrical and Electronics Engineers). ISSN 9781467385794. s. 656–660. doi: 10.1109/ICCNC.2016.7440667. Fulltekst i vitenarkiv
• Ferlin, Simone; Alay, Ozgu; Dreibholz, Thomas; Hayes, David Andrew & Welzl, Michael (2016). Revisiting congestion control for multipath TCP with shared bottleneck detection. IEEE Infocom. Proceedings. ISSN 0743-166X. 2016-July. doi: 10.1109/INFOCOM.2016.7524599. Fulltekst i vitenarkiv
• Rajiullah, Mohammad; Hurtig, Per; Brunstrom, Anna; Petlund, Andreas & Welzl, Michael (2015). An evaluation of tail loss recovery mechanisms for TCP. Computer communication review. ISSN 0146-4833. 45(1), s. 5–11. doi: 10.1145/2717646.2717648. Fulltekst i vitenarkiv
• Islam, Safiqul; Welzl, Michael; Gjessing, Stein & Khademi, Naeem (2015). Coupled congestion control for RTP media. Computer communication review. ISSN 0146-4833. 44(4). doi: 10.1145/2630088.2630089. Fulltekst i vitenarkiv
• Islam, Safiqul; Welzl, Michael; Gjessing, Stein & Khademi, Naeem (2014). Coupled congestion control for RTP media. I Kühlewind, Mirja & Kutscher, Dirk (Red.), Proceedings of the 2014 ACM SIGCOMM workshop on Capacity sharing workshop. Association for Computing Machinery (ACM). ISSN 9781450329910. s. 21–26. doi: 10.1145/2630088.2630089. Fulltekst i vitenarkiv
• Yousaf, Muhammad & Welzl, Michael (2014). Network-Aware HEFT Scheduling for Grid. Scientific World Journal. 2014. doi: 10.1155/2014/317284. Fulltekst i vitenarkiv
• Welzl, Michael; Gjessing, Stein & Khademi, Naeem (2014). Less-than-best-effort service for Community Wireless Networks: Challenges at three layers. I Hui, Pan & Karlson, Gunnar (Red.), 11th Annual Conference on Wireless On-demand Network Systems and Services. IEEE (Institute of Electrical and Electronics Engineers). ISSN 9781479949373. s. 148–153. doi: 10.1109/WONS.2014.6814737. Fulltekst i vitenarkiv
• Elkhatib, Yehia; Tyson, Gareth & Welzl, Michael (2014). Can SPDY Really Make the Web Faster? 2014 IFIP Networking Conference. IEEE (Institute of Electrical and Electronics Engineers). ISSN 9783901882586. s. 100–108. doi: 10.1109/IFIPNetworking.2014.6857089. Fulltekst i vitenarkiv
• Hayes, David Andrew; Ferlin, Simone & Welzl, Michael (2014). Practical passive shared bottleneck detection using shape summary statistics. I Kanhere, Salil & Akkaya, Kernal (Red.), 39th Annual IEEE Conference on Local Computer Networks (LCN 2014). IEEE (Institute of Electrical and Electronics Engineers). ISSN 9781479937806. s. 150–158. doi: 10.1109/LCN.2014.6925767. Fulltekst i vitenarkiv
• Yousaf, Muhammad Murtaza & Welzl, Michael (2013). On the Accurate Identification of Network Paths Having a Common Bottleneck. Scientific World Journal. doi: 10.1155/2013/890578. Fulltekst i vitenarkiv
• Ijaz, Humaira; Saleem, Sadia & Welzl, Michael (2013). Fewest Common Hops (FCH): An Improved Peer Selection Approach for P2P Applications, Parallel, Distributed and Network-Based Processing (PDP), 2013 21st Euromicro International Conference on. IEEE (Institute of Electrical and Electronics Engineers). ISSN 9780769549392. s. 449–453. doi: 10.1109/PDP.2013.73. Fulltekst i vitenarkiv Vis sammendrag

  Underlay-unawareness in P2P systems can result in sub-optimal peer selection for overlay routing and hence poor performance. The majority of underlay aware proposals for peer selection focus on finding the shortest overlay routes by selecting the nearest peers according to proximity. However, in case of multiple and parallel downloads, if the underlay paths between a downloader and its selected nearest peers share a bottleneck, this can cause congestion, leading to performance deterioration instead of improvement. This effect was neglected in previous work because, in today's Internet, the bottleneck is usually not shared as it is the end user's access link. This is no longer the case in more modern scenarios, e.g. with FTTH or with upcoming in-network caching techniques such as DECADE. We propose an improved peer selection approach for P2P applications called Fewest Common Hops (FCH) that ensures proximity based node selection having maximum path disjointness. It is a client based, infrastructure independent heuristic to optimize download time for multiple and parallel downloads in P2P content distribution applications. Simulations show that, even when FCH is implemented in the simplest possible fashion (using only traceroute), it can significantly decrease the download time. © 2012 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other users, including reprinting/ republishing this material for advertising or promotional purposes, creating new collective works for resale or redistribution to servers or lists, or reuse of any copyrighted components of this work in other works.

• Ros, David & Welzl, Michael (2013). Less-than-Best-Effort Service: A Survey of End-to-End Approaches. IEEE Communications Surveys and Tutorials. 15(2), s. 898–908. doi: 10.1109/SURV.2012.060912.00176. Fulltekst i vitenarkiv
• Ros, D. & Welzl, Michael (2013). Assessing LEDBAT's Delay Impact. IEEE Communications Letters. ISSN 1089-7798. 17(5), s. 1044–1047. doi: 10.1109/LCOMM.2013.040213.130137. Fulltekst i vitenarkiv
• Munir, Kashif; Welzl, Michael; Pasin, Marcelo & Vicat-Blanc, Pascale Primet (2012). Combining Explicit Admission Control and Congestion Control for Predictable Data Transfers in Grids. Future Generation Computer Systems. ISSN 0167-739X. 28(7), s. 1121–1132. doi: 10.1016/j.future.2011.06.012. Fulltekst i vitenarkiv
• Khademi, Naeem; Welzl, Michael & Gjessing, Stein (2012). Experimental evaluation of TCP performance in multi-rate 802.11 WLANs. I Banchs, Albert & Prabhakaran, Balakrishnan (Red.), 2012 IEEE international symposium on a world of wireless, mobile, and multimedia networks, WoWMoM. IEEE (Institute of Electrical and Electronics Engineers). ISSN 9781467312387. s. 1–9. doi: 10.1109/wowmom.2012.6263696. Fulltekst i vitenarkiv Vis sammendrag

  The goal of Rate Adaptation (RA) mechanisms in 802.11 WLANs is to provide optimum system throughput under varying channel conditions (e.g. in presence of noise) by carrying out run-time prediction and selection of the most appropriate bit-rate. The cross-layer interaction of TCP, as the major transport protocol in the Internet, with different RA mechanisms and DCF is yet to be thoroughly investigated. Previously reported efforts A) have never included real-life measurements of uplink TCP traffic; B) lack a practical view because they do not consider the RA mechanisms commonly deployed in today's off-the-shelf 802.11 devices; C) miss the study of RA mechanisms in low-noise environments. This paper covers all the above shortages, by conducting real-life measurements in two different test-beds (NDlab and Emulab) alongside with simulations, to study the performance of TCP coupled with different commonly deployed RA mechanisms. Our measurements reveal that 1) most conventional RA mechanisms are unable to distinguish frame errors due to collisions from channel noise/interference, and will respond to them negatively to some extent; 2) different than downlink TCP, uplink TCP can be adversely affected by collision-triggered rate downshifts that some RA schemes exhibit even under perfect channel conditions or in low-noise environments; 3) the relatively recent Minstrel RA mechanism can counter this negative uplink behavior well, yielding almost equal performance as in the downlink case.

• Welzl, Michael & Normann, Rolf Erik Gurandsrud (2012). A Client-side Split-ACK Tool for TCP Slow Start Investigation, 2012 International Conference on Computing, Networking and Communications (ICNC). IEEE Sarnoff Symposium. ISSN 9781467300094. s. 804–808. doi: 10.1109/iccnc.2012.6167535. Fulltekst i vitenarkiv
• Munir, Kashif; Cigno, Renato Lo; Vicat-Blanc, Pascale Primet & Welzl, Michael (2012). Planning Data Transfers in Grids: A Multi-Service Queueing Approach. Concurrency and Computation: Practice and Experience. ISSN 1532-0626. 24(4), s. 407–422. doi: 10.1002/cpe.1742. Fulltekst i vitenarkiv
• Schier, Michael & Welzl, Michael (2012). Optimizing Selective ARQ for H. 264 Live Streaming: A Novel Method for Predicting Loss-Impact in Real Time. IEEE transactions on multimedia. ISSN 1520-9210. 14(2), s. 415–430. doi: 10.1109/TMM.2011.2178235. Fulltekst i vitenarkiv
• Welzl, Michael; Niederbacher, Florian & Gjessing, Stein (2011). Beneficial Transparent Deployment of SCTP: the Missing Pieces. GLOBECOM - conference record / IEEE Global Telecommunications Conference. ISSN 0895-1195. Fulltekst i vitenarkiv
• Welzl, Michael; Niederbacher, Florian & Gjessing, Stein (2011). Beneficial Transparent Deployment of SCTP: the Missing Pieces. I N., N. (Red.), 2011 IEEE Global Telecommunications Conference - GLOBECOM 2011. IEEE Sarnoff Symposium. ISSN 9781424492688. Fulltekst i vitenarkiv
• Khademi, Naeem; Welzl, Michael & Cigno, Renato Lo (2011). On the uplink performance of TCP in multi-rate 802.11 WLANs. I N., N. (Red.), NETWORKING'11 Proceedings of the 10th international IFIP TC 6 conference on Networking. Springer. ISSN 9783642207976. s. 368–378. doi: 10.1007/978-3-642-20798-3_28. Fulltekst i vitenarkiv
• Schier, Michael & Welzl, Michael (2011). Video Streaming using Content-Aware Unequal Error Protection Fountain Codes. I N, N (Red.), CCNC 2011: IEEE Consumer Communications and Networking Conference; 9-12 Jan. 2011, Las Vegas. IEEE (Institute of Electrical and Electronics Engineers). ISSN 9781424487905. s. 525–526. doi: 10.1109/ccnc.2011.5766530. Fulltekst i vitenarkiv
• Schier, Michael & Welzl, Michael (2011). A content-aware rateless error protection scheme for live video streaming systems. I N., N. (Red.), 2011 IEEE International Symposium on a World of Wireless, Mobile and Multimedia Networks (WoWMoM). IEEE Press. ISSN 9781457703522. doi: 10.1109/wowmom.2011.5986437. Fulltekst i vitenarkiv
• Hayes, David A.; Welzl, Michael; Armitage, Grenville & Rossi, Mattia (2011). Improving HTTP performance using 'Stateless' TCP. I N., N. (Red.), Proceedings of the 21st international workshop on Network and operating systems support for digital audio and video. Association for Computing Machinery (ACM). ISSN 9781450307772. s. 57–62. doi: 10.1145/1989240.1989256. Fulltekst i vitenarkiv
• Nguyen, Anh Tuan; Eliassen, Frank & Welzl, Michael (2011). Quality-aware Membership Management for Layered Peer-to-Peer Streaming. I N, N (Red.), CCNC 2011: IEEE Consumer Communications and Networking Conference; 9-12 Jan. 2011, Las Vegas. IEEE (Institute of Electrical and Electronics Engineers). ISSN 9781424487905. s. 720–724. doi: 10.1109/ccnc.2011.5766585. Fulltekst i vitenarkiv
• Nguyen, Anh Tuan; Li, Baochun; Welzl, Michael & Eliassen, Frank (2011). Stir: Spontaneous Social Peer-to-Peer Streaming. I Ni, Lionel & Zhang, WenJun (Red.), 2011 IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS). IEEE Sarnoff Symposium. ISSN 9781457702488. s. 816–821. doi: 10.1109/infcomw.2011.5928925. Fulltekst i vitenarkiv
• Stewart, Lawrence; Hayes, David A.; Armitage, Grenville; Welzl, Michael & Petlund, Andreas (2011). Multimedia-unfriendly TCP congestion control and home gateway queue management. I N., N. (Red.), MMSys '11 Proceedings of the second annual ACM conference on Multimedia systems. Association for Computing Machinery (ACM). ISSN 9781450305181. s. 35–44. doi: 10.1145/1943552.1943558. Fulltekst i vitenarkiv
• Welzl, Michael; Jörer, Stefan & Gjessing, Stein (2011). Towards a Protocol-Independent Internet Transport API. I Uji, Noritaka (Red.), 2011 IEEE International Conference on Communications Workshops (ICC). IEEE Sarnoff Symposium. ISSN 9781612849553. doi: 10.1109/iccw.2011.5963568. Fulltekst i vitenarkiv
• Welzl, Michael; Abfalterer, Armin & Gjessing, Stein (2011). XCP vs. CUBIC with Quick-Start: Observations on Implicit vs. Explicit Feedback for Congestion Control. I Uji, Noritaka (Red.), 2011 IEEE International Conference on Communications (ICC). IEEE Sarnoff Symposium. ISSN 9781612842332. doi: 10.1109/icc.2011.5963377. Fulltekst i vitenarkiv
• Damjanovic, Dragana & Welzl, Michael (2011). An Extension of the TCP Steady-State Throughput Equation for Parallel Flows and its Application in MulTFRC. IEEE/ACM Transactions on Networking. ISSN 1063-6692. 19(6), s. 1676–1689. doi: 10.1109/TNET.2011.2142008. Fulltekst i vitenarkiv
• Welzl, Michael; Abfalterer, Armin & Gjessing, Stein (2011). XCP vs. CUBIC with Quick-Start: Observations on Implicit vs. Explicit Feedback for Congestion Control. IEEE International Conference on Communications. ISSN 1550-3607. doi: 10.1109/icc.2011.5963377. Fulltekst i vitenarkiv
• Schier, Michael & Welzl, Michael (2010). Content-Aware Selective Reliability for DCCP Video Streaming. I Falah, Abdul Aziz (Red.), Multimedia Computing and Information Technology (MCIT), 2010 International Conference On. IEEE Sarnoff Symposium. ISSN 9781424470037. s. 53–56. Fulltekst i vitenarkiv

Se alle arbeider i NVA

• Welzl, Michael; Karlsson, Gunnar; Alay, Ozgu & Peng, Chunyi (2022). 2022 17th Wireless On-Demand Network Systems and Services Conference (WONS). IEEE (Institute of Electrical and Electronics Engineers). ISBN 9783903176461. 150 s. Fulltekst i vitenarkiv

Se alle arbeider i NVA

• Sacco, Alessio; Shiomoto, Kohei; Zhani, Mohamed Faten; Marchetto, Guido; Mumtaz, Shahid & Welzl, Michael [Vis alle 7 forfattere av denne artikkelen] (2025). Guest Editors’ Introduction: Special section on Research Advances Toward Effective and Sustainable Next Generation Networks. IEEE Transactions on Network and Service Management. 22(5), s. 3764–3768. doi: 10.1109/tnsm.2025.3612131. Fulltekst i vitenarkiv
• Welzl, Michael; Oepen, Stephan; Jaskula, Cezary Cezary; Griwodz, Carsten & Islam, Safiqul (2021). Collaboration in the IETF: an initial analysis of two decades in email discussions. Computer communication review. ISSN 0146-4833. 51(3), s. 29–32. doi: 10.1145/3477482.3477488. Fulltekst i vitenarkiv Vis sammendrag

  RFC 9000, published in May 2021, marks an important milestone for the Internet's standardization body, the Internet Engineering Task Force (IETF): finally, the specification of the QUIC protocol is available. QUIC is the result of a five-year effort - and it is also the second of two major protocols (the first being SPDY, which later became HTTP/2) that Google LLC first deployed, and then brought to the IETF for standardization. This begs the question: when big players follow such a "shoot first, discuss later" approach, is IETF collaboration still "real", or is the IETF now being (mis-)used to approve protocols for standardization when they are already practically established, without really actively involving anyone but the main proponents?

• Khademi, Naeem; Welzl, Michael; Armitage, Grenville & Gjessing, Stein (2016). Improving the Fairness of Alternative Backoff with ECN (ABE). Fulltekst i vitenarkiv Vis sammendrag

  Abstract: Explicit Congestion Notification (ECN) lets a bottleneck's Active Queue Management (AQM) mechanism inform an endpoint about congestion without having to drop a packet. A recently proposed sender-side modification called Alternative Backoff with ECN (ABE) enables reduced latency while maintaining good utilization with ECN. However, under certain circumstances ABE can produce a degree of unfair behavior between ABE-enabled TCP senders and conventional TCP senders. We propose specific guidance for configuring bottleneck AQMs to assist in fairness between ABE-enabled and conventional TCP flows. We evaluate our proposal using RED, then describe how it can be applied to other AQM mechanisms and incrementally introduced into the Internet.

• Islam, Safiqul; Welzl, Michael & Gjessing, Stein (2013). One Control to Rule Them All - Coupled Congestion Control for RTP Media. Fulltekst i vitenarkiv
• Trammell, B.; Welzl, Michael; Enghardt, R.; Fairhurst, G.; Kühlewind, M. & Perkins, C. S. [Vis alle 8 forfattere av denne artikkelen] (2025). An Abstract Application Programming Interface (API) for Transport Services. RFC Series. doi: 10.17487/rfc9622. Fulltekst i vitenarkiv Vis sammendrag

  This document describes an abstract Application Programming Interface (API) to the transport layer that enables the selection of transport protocols and network paths dynamically at runtime. This API enables faster deployment of new protocols and protocol features without requiring changes to the applications. The specified API follows the Transport Services Architecture by providing asynchronous, atomic transmission of Messages. It is intended to replace the BSD Socket API as the common interface to the transport layer, in an environment where endpoints could select from multiple network paths and potential transport protocols.

• Brunstrom, A.; Pauly, T.; Enghardt, R.; Tiesel, P.S. & Welzl, Michael (2025). Implementing Interfaces to Transport Services. RFC Series. doi: 10.17487/rfc9623. Fulltekst i vitenarkiv Vis sammendrag

  The Transport Services System enables applications to use transport protocols flexibly for network communication and defines a protocol-independent Transport Services Application Programming Interface (API) that is based on an asynchronous, event-driven interaction pattern. This document serves as a guide to implementing such a system.

• Ciko, Kristjon; Welzl, Michael & Teymoori, Peyman (2022). Improving Internet Performance with a "Clean-Slate" Network Architecture - The Case of RINA. Universitetet i Oslo. Fulltekst i vitenarkiv Vis sammendrag

  Today’s Internet is facing significant challenges as the number of the connected devices, applications, and the amount of traffic grow tremendously. The challenges are particularly related to the insufficient security and performance, complexity in providing mobility and multihoming, and the inability to easily accommodate new protocols and technologies. Over the last two decades, there has been considerable effort by the networking research community to evolve or redesign the Internet architecture. Unlike evolutionary approaches, “clean-slate” designs follow a more drastic path to overcome the fundamental architecture limitations of the current Internet. However, such solutions are facing large deployment barriers, mainly because the Internet does not provide architectural modularity. Among the “clean-slate” initiatives, RINA—the Recursive InterNetwork Architecture—is perhaps the most drastic one which attempts to offer a general solution with a complete networking theory. In addition, the recursive nature of RINA introduces potential benefits in many aspects of networking such as management, routing, security, and congestion control. This dissertation tackles the deployment challenges and investigates several strategies to deploy a new clean-slate architecture such as RINA in the current networks. These strategies are: (1) direct usage with a fall-back, (2) adapting the Application Programming Interface (API), and (3) a gateway to translate between protocols. Real-life experiments of the first congestion control approach that is suitable for a flexible recursive architecture show the performance improvement that can be attained with these strategies.

• Ciko, Kristjon; Welzl, Michael & Teymoori, Peyman (2021). PEP-DNA: a Performance Enhancing Proxy for Deploying Network Architectures. Universitetet i Oslo. ISSN 9788273686008. Fulltekst i vitenarkiv Vis sammendrag

  Deploying a new network architecture in the Internet requires changing some, but not necessarily all elements between communicating applications. One way to achieve gradual deployment is a proxy or gateway which “translates” between the new architecture and TCP/IP. We present such a proxy, called “Performance Enhancing Proxy for Deploying Network Architectures (PEP-DNA)”, which allows TCP/IP applications to benefit from advanced features of a new network architecture without having to be redeveloped. Our proxy is a kernel-based Linux implementation which can be installed wherever a translation needs to occur between a new architecture and TCP/IP domains. We discuss the proxy operation in detail and evaluate its efficiency and performance in a local testbed, demonstrating that it achieves high throughput with low additional latency and low CPU and memory overhead. In our experiments we use the Recursive InterNetwork Architecture (RINA) and Information- Centric Networking (ICN) as examples, but our proxy is modular and flexible, and hence enables realistic gradual deployment of any new “clean-slate” approaches.

• Teymoori, Peyman & Welzl, Michael (2020). LGCC: Food Chain Multi-Hop Congestion Control. Universitetet i Oslo. ISSN 9788273684592. Fulltekst i vitenarkiv Vis sammendrag

  Technological advancements have provided wireless links with very high capacity for 5G mobile networks and WiFi 6, which will be widely deployed by 2025; however, the capacity heavily fluctuates, violating the assumption at the transport layer that the capacity is (almost) fixed. This issue has so far been targeted by mostly ”hack-and-patches", which are still far from an efficient design. In this paper, we present a general and efficient, yet deployable solution to the aforementioned problem through a novel design empowered with a rich theory, allowing a significantly-improved experience in using new technologies, especially mobile cellular services.

• Teymoori, Peyman; Hayes, David Andrew; Welzl, Michael & Gjessing, Stein (2019). Estimating an Additive Path Cost with Explicit Congestion Notification. Universitetet i Oslo. ISSN 9788273684523. Fulltekst i vitenarkiv Vis sammendrag

  Abstract—Network Utility Maximization (NUM) is a well accepted theoretical concept that describes how congestion controls could cooperate to achieve an ideal sending rate allocation, for given utility functions of senders and constraints of the network. These network constraints are expressed as a “cost” in the framework. In practice, most congestion control mechanisms obtain feedback that is different from a “cost”. This paper focuses on Explicit Congestion Notification (ECN), which has been shown to be quite advantageous when it is available, e.g. with the popular Datacenter TCP (DCTCP) mechanism. However, different from a cost, ECN marks are not additive. We present a practical solution to this problem. Our solution changes how end hosts interpret the ECN signal, while the router side can be implemented via an unusual configuration of RED parameters.

• Li, Qian & Welzl, Michael (2017). Vulture – Variable Aggressiveness Ultra Low Impact Transport Using Receiver-based Flow-control Mechanism. Universitetet i Oslo. ISSN 9788273684394. Fulltekst i vitenarkiv Vis sammendrag

  Less than Best Effort (LBE) transports are transport protocols that use spare bandwidth left by Best Effort (BE) transports to fulfill their own data transfer tasks. Such kind of protocols can be used by non-delay/bandwidth sensitive applications such as software updating, peer-to-peer file sharing, prefetching, and replication to realize background transfer. By incorporating BE and LBE protocols into the Internet, we can realize flow-based prioritization, which can ensure that the available bandwidth is divided among different applications wisely and give every application an opportunity to run. We have designed and developed a receiver-side flow-control-based LBE congestion control mechanism. By observing the sign and value of packet interval variations at the receiver side, we can successfully estimate the degree of congestion in the network and distinguish situation (a): the congestion is caused by the LBE flow itself from situation (b): the congestion is caused by the LBE flow as well as other flows. The protocol runs in a conservative mode if it detects other flows, otherwise, it runs in an aggressive mode to quickly reap the available bandwidth. Testbed evaluation results show that our LBE protocol inflicts low throughput impact on BE flows with which it shares the same bottleneck. The average goodput achieved by a BE flow does not change after an LBE flow is started to operate. An LBE flow can add, on average, 40% round trip delay to a BE flow. Bandwidth utilization ratios of an LBE flow when it runs alone and shares a bottleneck with a constant bit rate BE flow are 99% (stable state, LBE only), 93% (overall, LBE only), and 83% (LBE and BE) respectively. When a BE flow exhibits on-off behavior, the LBE flow’s bandwidth utilization ratio decreases with the decrease of the on/off interval. An LBE flow can immediately yield its bandwidth when a BE flow is started. Finally, our LBE flows are fair toward each other. Jain’s fairness index is 0.99 for 4, 16, and 64 LBE flows.

• Islam, Safiqul; Welzl, Michael; Gjessing, Stein & Khademi, Naeem (2014). Coupled Congestion Control for RTP Media. University of Oslo. ISSN 9788273684059. Fulltekst i vitenarkiv Vis sammendrag

  Congestion occurs at a bottleneck along an Internet path; multiple flows between the same sender and receiver pairs can benefit from using only a single congestion control instance when they share the same bottleneck. These benefits include the ability to control the rate allocation between flows and reduced overall delay (multiple congestion control instances cause more queuing delay than one since each has no knowledge of the congestion episodes experienced by the others). We present a mechanism for coupling congestion control for real-time media and show its benefits by coupling multiple congestion controlled flows that share the same bottleneck.

• Khademi, Naeem; Ros, David & Welzl, Michael (2013). The New AQM Kids on the Block: Much Ado About Nothing? Universitetet i Oslo. ISSN 9788273683991. Fulltekst i vitenarkiv Vis sammendrag

  Active Queue Management (AQM) design has again come into the spotlight of network operators, vendors and OS developers. This reflects the growing concern and sensitivity about the end-to-end latency perceived by today’s Internet users. Indeed, delays on the order of seconds have become common due to the deployment of excessively-sized FIFO/DropTail buffers at the edge of many networks. CoDel and PIE are two AQM mechanisms that have recently been presented and discussed at the IRTF and the IETF. However, to the best of our knowledge, they have not yet been thoroughly evaluated or compared against each other except by simulation. We set thus to perform an experimental evaluation using real-world implementations, in both wired and wireless testbeds. We have in addition compared them with a decade-old variant of RED called Adaptive RED, which shares with CoDel and PIE the goal of “knob-free” operation. Surprisingly, in many instances results were much more favorable towards Adaptive RED. We do not call into question the need for new AQMs, however, there are lessons yet to be learned from old designs.

Se alle arbeider i NVA