Alexander Frömmgen

ReMP TCP: Low latency multipath TCP

More and more Internet-enabled devices, such as server instances or smartphones, have multiple network interfaces. Multipath TCP (MPTCP) has proven to increase bandwidth for these devices, while remaining compatible with the existing network infrastructure and applications. For interactive applications and services, however, low latency and low jitter often is more important than bandwidth. In this paper, we rethink the MPTCP approach, focusing on end-to-end latency and jitter. We propose ReMP TCP, an MPTCP extension that sends data redundantly over multiple paths in the network. Exchanging bandwidth for latency, this approach guarantees the lowest possible latency in existing best-effort networks. The integration into the MPTCP protocol provides benefits such as transparent end-to-end connection establishment, multipath-enabled congestion control, and the prevention of head of line blocking. We discuss end-to-end latency in multipath environments considering both queuing delays and packet drops. Further, we evaluate the performance of our ReMP TCP Linux Kernel implementation for data center and mobile scenarios in Mininet and real world experiments. We show for a real world mobile scenario in a stressed environment that ReMP TCP can halve the average round-trip time and reduce its standard deviation by a factor of 19.

Fossa: Learning ECA Rules for Adaptive Distributed Systems

The development of adaptive distributed systems is complex. Due to a large amount of interdependencies and feedback loops between network nodes and software components, distributed systems respond nonlinearly to changes in the environment and system adaptations. Although Event Condition Action (ECA) rules allow a crisp definition of the adaptive behavior and a loose coupling with the actual system implementation, defining concrete rules is nontrivial. It requires specifying the events and conditions which trigger adaptations, as well as the selection of appropriate actions leading to suitable new configurations. In this paper, we present the idea of Fossa, an ECA framework for adaptive distributed systems. Following a methodology that separates the adaptation logic from the actual application implementation, we propose learning ECA rules by automatically executing a multitude of tests. Rule sets are generated by algorithms such as genetic programming, and the results are evaluated using a utility function provided by the developer. Fossa therefore provides an automated offline learner that derives suitable ECA rules for a given utility function.

Towards the Description and Execution of Transitions in Networked Systems

Today’s distributed systems have to work in changing environments and under different working conditions. To provide high performance under these changing conditions, many distributed systems implement adaptive behavior. While simple adaptation through parameter tuning can only react to a limited range of conditions, a switch between different mechanisms at runtime enables broader adaptivity. However, distributed systems that switch mechanisms at runtime lack a clear abstraction for the adaptive behavior and, thus, usually interleave the adaptation and actual application logic. This leads to complex and error-prone systems that are hard to maintain and not easy to extend.

TARL: modeling topology adaptations for networking applications

Many networking applications implement topology adaptations to cope with network dynamics. Related work focuses on the specific application, lacking a general model for topology adaptations. In this paper, we analyze 14 topology adaptations from two different application domains. Based on the derived characteristics, we propose a general topology adaptation model. We present the Topology Adaptation Rule Language (TARL) to specify topology adaptation logic following this model. We discuss the execution of TARL rules for two application domains as well as how our model enables reasoning and optimizations on topology adaptations. For the evaluation, we developed a TARL runtime environment as a reusable topology adaptation framework. TARL simplifies the development of topology adaptations and is able to express 13 of the analyzed algorithms.

Crowdsourcing Measurements of Mobile Network Performance and Mobility During a Large Scale Event

Cellular infrastructure in urban areas is provisioned to easily cope with the usual daily demands. When facing shockingly high loads, e.g., due to large scale sport or music events, users complain about performance degradations of the mobile network. Analyzing the impact of large scale events on the mobile network infrastructure and how users perceive overload situations is essential to improve user experience. Therefore, a large data set is required to get a detailed understanding of the differences between providers, mobile devices, mobile network access technologies, and the mobility of people.

Where are the Sweet Spots?: A Systematic Approach to Reproducible DASH Player Comparisons

The current body of research on Dynamic Adaptive Streaming over HTTP (DASH) contributes various adaptation algorithms aiming to optimize performance metrics such as the Quality of Experience. Intuitively, the heterogeneity of the streaming environment and the underlying technologies lead many of the developed approaches to possess clear performance affinities denoted here as sweet spots. We observe, however, that systematic comparisons of these algorithms are usually conducted within homogeneous player environments. In this work, we show the substantial impact of player choice and configuration on the streaming performance. To this end, we systematically examine three established open-source DASH players, i.e., DASH.JS, Googles Shaka Player, and AStream, that implement fundamentally different configurations featuring various adaptation algorithms. By establishing a large scale emulation framework we (i) extract player sweet spots and (ii) achieve a direct, reproducible comparison of real-world DASH players and algorithms. We present empirical evidence demonstrating that an isolated analysis of DASH player modules is insufficient to capture the player streaming performance. One of the major observations is that the choice of the target buffer size together with the player implementation dominates the choice of the adaptation algorithms.

A programming model for application-defined multipath TCP scheduling

Multipath TCP enables remarkable optimizations for throughput, load balancing, and mobility in todays networks. The design space of Multipath TCP scheduling, i.e., the application-aware mapping of packets to paths, is largely unexplored due to its inherent complexity. Evidence in this paper suggests that an application-aware scheduling decision, if leveraged right, pushes Multipath TCP beyond throughput optimization and thereby provides benefits for a wide range of applications. This paper introduces a high-level programming model that enables application-defined Multipath TCP scheduling. We provide an efficient interpreter and eBPF-based runtime environment for the Linux Kernel, enabling isolated application-defined schedulers in multi-tenancy environments. In combination with a high-level API, our work closes the gap between scheduler specification and deployment. We show the strength of our programming model by implementing seven novel schedulers tackling diverse objectives. Our real world measurements, for example, of an application- and preference-aware scheduler, show that the programming model enables timely scheduling decisions to retain fine-grained throughput objectives. Further measurements of a novel HTTP/2-aware scheduler show significantly improved interactions with upper-layer protocols, e.g., an optimized dependency resolution, while preserving path preferences.

Fossa: Using genetic programming to learn ECA rules for adaptive networking applications

Due to complex interdependencies and feedback loops between network layers and nodes, the development of adaptive applications is difficult. As networking applications respond nonlinearly to changes in the environment and adaptations, defining concrete adaptation rules is nontrivial. In this paper, we present the offline learner Fossa, which uses genetic programming to automatically learn suitable Event Condition Action (ECA) rules. Based on utility functions defined by the developer, the genetic programming learner generates a multitude of rule sets and evaluates them using simulations to obtain their utility. We show, for a concrete example scenario, how the genetic programming learner benefits from the clear model of the ECA rules, and that the methodology efficiently generates ECA rules which outperform nonadaptive and manually tuned solutions.

Poster: Use your Senses: A Smooth Multipath TCP WiFi/Mobile Handover

The handover from WiFi to mobile networks is known to lead to TCP connection drops due to changing IP addresses. Multipath TCP (MPTCP), a recent TCP extension, enables a transparent mobile handover by combining subflows on multiple interfaces, such as WiFi and LTE, to one logical connection. MPTCP provides multiple handover modes, which differ in their energy consumption and the performance during the handover. The Full-MPTCP mode uses permanently both WiFi and the mobile network, which increases energy consumption. The Single-Path mode establishes the mobile network connection after the WiFi connection broke, which leads to a short performance degradation. In this paper, we argue that this trade-off is not necessary. We propose to use the available (sensor) information to forecast the mobile handover. This allows switching to the Full-MPTCP mode before the WiFi connection breaks, providing both low energy consumption and high performance during the handover. For a first experimental evaluation, we use a declining WiFi link quality to forecast a handover. Our real world measurements show that both low energy consumption and high performance during the handover are possible at the same time.

QoE Analysis of DASH Cross-Layer Dependencies by Extensive Network Emulation

With the rising importance of video streaming in the Internet, dynamic adaptive streaming over HTTP (DASH) has been established as a key technology for video delivery. Yet, variable network conditions often result in a limited quality of experience (QoE)—with the interrelation of cross-layer network factors and DASH mechanisms widely unexplored. To understand the complex dependencies between DASH configurations and network conditions, we propose a systematic extensive large-scale emulation approach with state-of- the-art QoE metrics. Using this approach with a real DASH player in Mininet, we emulated more than 10, 000 combinations of static and dynamic network conditions and DASH configurations to derive their QoE. The obtained results show that no single DASH configuration provides the highest achievable QoE. Depending on the network conditions, specific combinations of the TCP congestion control, segment sizes and the DASH adaptation algorithm provide higher QoE—showing the possibility of performance improvements in practice. Furthermore, the ex- tensive emulations show a linear relation between delay, loss and QoE that is mostly independent of bandwidth.