Fixed buffer sizing in computer networks, especially the Internet, is a compromise between latency and bandwidth. A decision in favor of high bandwidth, implying larger buffers, subordinates the latency as a consequence of constantly filled buffers. This phenomenon is called Bufferbloat. Active Queue Management (AQM) algorithms such as CoDel or PIE, designed for the use on software based hosts, offer a flow agnostic remedy to Bufferbloat by controlling the queue filling and hence the latency through subtle packet drops.In previous work, we have shown that the data plane programming language P4 is powerful enough to implement the CoDel algorithm. While legacy software algorithms can be easily compiled onto almost any processing architecture, this is not generally true for AQM on programmable data plane hardware, i.e., programmable packet processors. In this work, we highlight corresponding challenges, demonstrate how to tackle them, and provide techniques enabling the implementation of such AQM algorithms on different high speed P4-programmable data plane hardware targets. In addition, we provide measurement results created on different P4-programmable data plane targets. The resulting latency measurements reveal the feasibility and the constraints to be considered to perform Active Queue Management within these devices. Finally, we release the source code and instructions to reproduce the results in this paper as open source to the research community.

QoS requirements of current network control and management applications require the ability to conduct precise measurements of network elements, including switches, routers and Virtual Network Functions (VNFs). State-of-the-art network switches have a forwarding delay of 1µs and below and offer high bandwidths of hundreds Gigabits per second. This imposes high time accuracy and loss-detection requirements on measurement equipment that are not met by existing, software-based measurement tools. The use of specialized tools, meeting these requirements, is restricted by limited flexibility and high cost.In this work, we introduce P4STA, an open source frame-work that combines the flexibility of software-based traffic load generation with the accuracy of hardware packet timestamping. Our evaluation results, obtained using an off-the-shelf P4-programmable switch, show that a time resolution up to 1ns can be achieved on these programmable data plane platforms. Moreover we show how to combine the traffic load of multiple software-based load generators to achieve a measurement load of up to 100Gbit/s per port. Experiments on further programmable platforms, specifically on P4-SmartNICs and FPGAs, show similar results. With this work, we make P4STA available for the research community to advance high performance experiment measurements at nanosecond accuracy.

Large-scale telecommunications providers have to continuously challenge and evolve their network infrastructure to efficiently serve growing markets demands. They must increase performance, lower time-to-market, provide new services, and lower the cost of the infrastructure and its operation. Network Functions Virtualization (NFV) on commodity hardware offers an attractive, low-cost platform to establish innovations much faster than with purpose-built hardware products. Unfortunately, implementing NFV on commodity processors does not match the performance requirements of the high-throughput data plane components in large carrier access networks. In this article, we propose a way to offer residential network access with programmable packet processing architectures. Based on the highly flexible P4 programming language, we present a design and open source implementation of a BNG data plane that meets the challenging demands of Broadband Network Gateways in carrier-grade environments. The proposed evaluation results show the desired performance characteristics and our proposed design together with upcoming P4 hardware can offer a giant leap towards highest performance NFV network access.

Large Distributed Denial-of-Service (DDoS) attacks pose a major threat not only to end systems but also to the Internet infrastructure as a whole. Remote Triggered Black Hole filtering (RTBH) has been established as a tool to mitigate inter-domain DDoS attacks by discarding unwanted traffic early in the network, e.g., at Internet eXchange Points (IXPs). As of today, little is known about the kind and effectiveness of its use, and about the need for more fine-grained filtering.

Bare-metal switches (BMS) are switches sold without an operating system (OS), i.e. including only the switch hardware. For network operators this may not only provide a cost advantage, it also allows them to customize the OS behavior to their own needs. The aim of this demonstration is to show the applicability of BMS in a carrier-grade environment. To this end, a representative use case has been implemented on a BMS, which includes the healing of RTP packet loss by automatically selecting a stream with the least packet loss from redundant uplinks. This use case demonstrates the flexibility of BMS to be adaptable to even such specific needs in a network.

Crowd sensing exploits users' smart devices and human mobility to collect information on a large scale. To realize a crowd sensing campaign, sensing tasks with spatio-temporal requirements are distributed to the devices that can provide the requested information. Typically, the distribution of sensing tasks relies on a centralized communication infrastructure such as cloud servers. However, such an approach will be unsuitable if access to communication infrastructure is restricted, for example in disaster relief scenarios. To fill this gap, we propose a distributed context-aware framework for disseminating sensing tasks, based on the Named Data Networking (NDN) paradigm. By adding context attributes to Interest packets, we allow a device to utilize this information to make forwarding decisions autonomously, thus guiding the sensing tasks towards the suitable sensing devices. Through intensive evaluation, we show that our framework achieves a timely delivery of sensing tasks, while keeping the communication overhead to a minimum compared to pure geo forwarding and flooding approaches.

The current ecosystem of network elements, such as switches and appliances, is largely dominated by devices supplied and sold with a bundled operating system, and software dedicated to manage the device's forwarding hardware, however, these platforms are not open-source and cannot be arbitrarily customized, and there is no cost transparency or flexibility in choosing software different to the bundled components.,,,,In this paper, we explore the capabilities of bare-metal switches, which are equipped with commodity switching hardware components, but shipped without an operating system. We evaluate the feasibility of these commonly lower-cost devices to meet the requirements of a customized, carrier-grade network function. Therefore, we have implemented a prototype on generic hardware, re-using as much open-source software as possible. Our Broadband Remote Access Server (BRAS) prototype can lower the cost compared to proprietary network appliances, and, known to have a hardware backplane capacity of 720 Gbps, the merchant-silicon / ASIC approach can highly outperform the state of the art of current x86-based virtualized network functions, while implementing the most important BRAS features.

Multi-layer resilience is one of the prominent new concepts for modern carrier networks; it efficiently combines the advantages of the optical and the packet layer. However, not all features offered by modern optical transport networks can be fully used by the packet layer yet. In case of a fiber cut, an optical protection mechanism restores the original IP topology after a short transient time. But such an optical restoration is expected to use a new, longer light-path, which in turn might affect the optical capacity of the link. Bit-rate flexible optical transceivers are able to utilize the remaining optical capacity efficiently by adapting the network link capacity accordingly. However, the packet layer is not able to cope with fluctuating link capacities; often the policy is to rather shut down the link completely instead of using the remaining capacity. Consequently, this paper proposes a Segment Routing-based approach with superimposed Software-defined Networking (SDN) to allow the IP network to benefit from these new features. The minimally invasive, gradual deployment of the system is investigated, while keeping other proven and resilient technologies and systems unmodified. Using the topology and traffic matrix of a large German carrier the feasibility of such a deployment is evaluated.

A number of today's over-the-top (OTT) services could greatly benefit from a flexible, efficient, and scalable network-layer multicast support on the Internet. IP multicast showed not to meet these requirements, hindering its adoption in more than individual network islands. As a consequence, content delivery networks (CDNs) emerged as alternative and implemented one-to-many multicasting at the application layer. Yet, due to practical reasons and conflicting stakeholder interests, CDN deployments usually end at the border of Internet service provider networks, making one-to-many deliveries at an Internet-scale challenging up to date. In a previous work, the authors presented software-defined multicast (Sdm) to mitigate this problem by unlocking network-layer multicast support for OTT services inside ISP networks. In this paper, DYNSDM is proposed to complement Sdm in key aspects, allowing for a more holistic support of relevant scenarios aspects, and enabling a fine-granular traffic engineering of multicast traffic inside ISP networks. For this, DYNSDM presents a detailed design of the ISP-internal traffic and service management process and introduces a set of novel SDN-based network-layer mechanisms for traffic load balancing as well as handling of group and network dynamics. DYNSDM was implemented as prototype and extensively evaluated, showing its flexibility, high traffic efficiency, good scalability, and superior ability to balance traffic load inside ISP networks.

Dynamic network service chaining allows network operators to apply network services to customer traffic on demand and in a highly flexible manner. SDN and particularly OpenFlow-based approaches have been presented that exploit the flexibility and feature richness of OpenFlow for service chaining in data center settings. These systems often use network architectures similar to MPLS, where network traffic is processed at the network edges and the network core conducts simple packet forwarding only. This approach is suitable for many use cases due to the complementary characteristics of the hardware devices used in the core and the virtual switches used at the edge on virtualization hosts. Yet, this leads to a concentration of processing flow updates solely at the network edges, which can cause lowered QoS during flow update peaks, e.g. when a virtualization host fails. To tackle this problem, in this paper a concept is presented and evaluated that offloads OpenFlow rule updates from software-based edge switches to exploit hitherto unused resources on hardware switches in the core of the network. Furthermore, an analytical model is presented that allows describing the expected gain of the approach based on characteristics of the used OpenFlow devices.

A number of today's over-the-top (OTT) services could greatly benefit from a scalable and efficient network-layer multicast support on the Internet. IP multicast showed to not meet these requirements and, thus, is not available for this purpose. Content Delivery Networks emerged as global alternative but usually end at the border of ISP networks. Software-Defined Multicast (SDM) is proposed in a previous work by the authors, enabling ISP-internal network-layer multicast delivery of OTT traffic. While it coins fundamental concepts, it does not detail the ISP-internal traffic and service management and leaves important questions unanswered. To this end, DYNSDM is proposed in this paper, detailing the multicast planning and management, proposing a novel network-layer multi-tree mechanism to distribute traffic on links inside the ISP network, and introducing mechanisms to handle group and network dynamics. DYNSDM was prototypically evaluated, showing its high traffic efficiency, good scalability, and superior traffic distribution characteristics.

Energy efficiency of networking equipment has become an important factor for private consumers as well as business customers in the recent years. This trend is driven by two factors: the proliferation of new mobile devices as well as energy prices as cost driver. Battery capacity as well as dynamic distribution of workloads adapted to energy costs are an active field of research. This demo contributes to this discussion by presenting a framework allowing the assessment of energy costs along the content distribution chain.

Internet Service Providers (ISPs) need to cope with a growing amount of over-the-top (OTT) traffic, often without a share in the high revenues of the content providers. To achieve an efficient global delivery of content, today content providers usually employ content delivery networks (CDNs) located at the edge of ISP networks from where content is delivered to end users via IP unicast. Many OTT services could benefit from a better support within the ISP's network, e.g. by packet duplication to deliver OTT video streams. While traditional solutions like IP multicast did not prevail, SDN-based alternatives have started to gain attention recently. In contrast to traditional approaches, SDN enables ISPs to support network services in a more manageable and flexible manner. However, the approaches proposed so far are quite rigid and keep state at every network device, independent of the multicast group size. To alleviate this problem, this paper proposes a new approach termed ASDM enabling ISPs to dynamically adjust the tradeoff between bandwidth and state for any multicast service. It is shown that, given ISP-defined bandwidth and state cost functions, the optimal parameter for ASDM can be derived and applied for a transparent multicast-to-unicast conversion achieving the desired characteristics. The proposed approach results in up to 30% bandwidth reduction compared to unicast while using only a seventh of the network state compared to traditional multicast.

Network management systems (NMS) need to support a wide range of application requirements while ensuring an efficient use of networking resources. Application control of Software Defined Networking (SDN) promises to drastically reduce OPEX and increase flexibility of NMS by automatically translating application requirements into network policies and subsequently into network behavior. The current research in this area focuses mainly on top-down approaches aiming at expressiveness, generality, and control plane resource efficiency. However, the expressive power and flexibility of Open Flow, a widely used SDN implementation, comes at the cost of high complexity when high-level policies are automatically translated into Open Flow rules, which can lead to an inefficient use of data plane resources. To counter this problem, this paper proposes a bottom-up approach to application-controlled SDN which aims to leverage scenario- and application-specific information for optimization. Thereby, the proposed approach is able to satisfy diverse application requirements and use the available data plane resources more efficiently than comparable approaches.

Internet video streaming causes the second largest transfer volume and is the second fastest growing application class in Internet traffic analysis [3]. In this context, also the streaming of live content becomes increasingly relevant as more traditional broadcasters start delivering content over the Internet. Today, live video streaming services rely on IP-unicast delivery or closed IP-multicast systems inside single administrative domains. Approaches such as Content Delivery Networks (CDNs) are used to improve the unicast delivery of content. They usually end at the edge of the residential broadband access Internet Service Provider (ISP) networks that connect end users to the Internet. For live content delivery, network layer multicast would be desirable for ISPs as well as content providers to reduce the load due to parallel unicast connections for the same content. Because of the well-known drawbacks and limitations of IP-multicast [2], however, network layer multicast support is usually not available.

Mobile Web services lie on the intersection of two big IT trends, namely Service-oriented Architectures (SOA) and mobile applications. So, their usage is likely to expand dramatically in the next years. However, the heavyweight nature of service-orientation in terms of the messaging-overhead that is necessary in order to achieve interoperability and loose-coupling comes in contrast with the lightweight nature of mobile devices and with the need to transmit wirelessly as few data as possible. This study categorizes the mechanisms that have been designed in order to bridge this gap, provides comparisons, discusses the results of related experiments, and introduces the future scenario in which the insights of the study can be exploited.