Liang LI Hamid FARHADY Ping DU Akihiro NAKAO
In most cases, the programmability of Software Defined Network (SDN) refers to the flexibility existing in northbound interface that enables network managers to control the behaviors of the networks. However, the lack of flexibility in data plane conversely results in wasting potentially usable information for controlling flows, especially from network services and applications point of view. For example, OpenFlow switches only deal with L2-L4 headers and ignore the other parts of packet. We propose Ouroboros as a programmable switch logic to increase the flexibility of SDN southbound interface. Ouroboros switches not only remove the limitation of regular OpenFlow switches using packet headers as the reference for packet switching, but also provides a highly flexible interface for network managers to conduct application-specific flow control according to packet content at any arbitrary offsets. Ouroboros can penetrate deeply into packet (e.g., RTP or SIP) protocol headers, or further into packet payload, to process user-defined switching protocol. Our evaluations of Ouroboros on 10Gbps traffic indicates the effectiveness of proposed method.
In this paper, we examine a new P2P traffic localization approach that exploits peer selection adaptation (i.e., preferring peers who are likely to provide better performance), called Netpherd. Netpherd enables peers to communicate with local domain peers by manipulating networking performance across network domains (i.e., adding an artificial delay to inter-domain traffic). Our feasibility study shows that Netpherd reduces the inter-domain traffic by influencing peer selection adaptation. Netpherd also improves download performance of the peers who know many local domain peers. We discuss one guideline to improve Netpherd based on the feasibility study and verify the guideline with evaluation results.
As one innovative research that heavily depends on the network virtualization for its realization and deployment on an Internet-scale, we propose an approach to utilize user resources in information-centric network (ICN). We try to fully benefit from the in-network cache that is one attractive feature of ICN by expanding the in-network cache indirectly based on the user resources. To achieve this, in this paper, we focus on how to encourage users to contribute their resources in ICN. Through simulations, we examine a feasibility of our approach and an effect of user participation on the content distribution performance in ICN. We also briefly discuss how the network virtualization technique can be utilized for our research in terms of its evaluation and deployment.
Akihiro NAKAO Ping DU Takamitsu IWAI
In this paper, we apply the concept of software-defined data plane to defining new services for Mobile Virtual Network Operators (MVNOs). Although there are a large number of MVNOs proliferating all over the world and most of them provide low bandwidth at low price, we propose a new business model for MVNOs and empower them with capability of tailoring fine-grained subscription plans that can meet users' demands. For example, abundant bandwidth can be allocated for some specific applications, while the rest of the applications are limited to low bandwidth. For this purpose, we have recently proposed the concept of application and/or device specific slicing that classifies application and/or device specific traffic into slices and applies fine-grained quality of services (QoS), introducing various applications of our proposed system [9]. This paper reports the prototype implementation of such proposal in the real MVNO connecting customized smartphones so that we can identify applications from the given traffic with 100% accuracy. In addition, we propose a new method of identifying applications from the traffic of unmodified smartphones by machine learning using the training data collected from the customized smartphones. We show that a simple machine learning technique such as random forest achives about 80% of accuracy in applicaton identification.
Masahiro YOSHIDA Akihiro NAKAO
BitTorrent is one of the most popular P2P file sharing applications worldwide. Each BitTorrent network is called a swarm, and millions of peers may join multiple swarms. However, there are many unreachable peers (NATed (network address translated), firewalled, or inactive at the time of measurement) in each swarm; hence, existing techniques can only measure a part of all the peers in a swarm. In this paper, we propose an improved measurement method for BitTorrent swarms that include many unreachable peers. In essence, NATed peers and those behind firewalls are found by allowing them to connect to our crawlers by actively advertising our crawlers' addresses. Evaluation results show that the proposed method increases the number of unique contacted peers by 112% compared to the conventional method. Moreover, the proposed method increases the total volume of downloaded pieces by 66%. We investigate the sampling bias among the proposed and conventional methods, and we find that different measurement methods yield significantly different results.
This paper introduces a methodology for engineering best-effort P2P algorithms into dependable P2P-based network control mechanism. The proposed method is built upon an iterative approach consisting of improving the original P2P algorithm by appropriate mechanisms and of thorough performance assessment with respect to dependability measures. The potential of the methodology is outlined by the example of timely routing control for vertical handover in B3G wireless networks. In detail, the well-known Pastry and CAN algorithms are enhanced to include locality. By showing how to combine algorithmic enhancements with performance indicators, this case study paves the way for future engineering of dependable network control mechanisms through P2P algorithms.
In this paper, we posit that, in future mobile network, network softwarization will be prevalent, and it becomes important to utilize deep machine learning within network to classify mobile traffic into fine grained slices, by identifying application types and devices so that we can apply Quality-of-Service (QoS) control, mobile edge/multi-access computing, and various network function per application and per device. This paper reports our initial attempt to apply deep machine learning for identifying application types from actual mobile network traffic captured from an MVNO, mobile virtual network operator and to design the system for classifying it to application specific slices.
In this letter, we argue that user resources will be still useful in the information-centric network (ICN). From this point of view, we first examine how P2P utilizing user resources looks like in ICN. Then, we identify challenging research issues to utilize user resources in ICN.
Yuichiro HEI Akihiro NAKAO Tomohiko OGISHI Toru HASEGAWA Shu YAMAMOTO
The current interdomain routing protocol, BGP, is not resilient to a path failure due to its single-path and slowly-converging route calculation. This paper proposes a novel approach to improve the resilience of the interdomain communication by enabling a set of ASes to form an alliance for themselves. The alliance members cooperatively discover a set of disjoint paths using not only the best routes advertised via BGP but also the ones unadvertised. Since such a set of disjoint paths are unlikely to share a link failure, a member AS can provide a pair of the other members with a transit to circumvent the failure. We evaluate how many disjoint paths we could discover from both advertised and hidden (unadvertised) routes by analyzing publicly available BGP route data. Our feasibility study indicates that an alliance of ASes can establish a set of disjoint paths between arbitrary pair of its alliance members with high probability to improve the resilience of interdomain routing among the members.
In this paper, we posit that extension of SDN to support deeply and flexibly programmable, software-defined data plane significantly enhance SDN and NFV and their interaction in terms of (1) enhanced interaction between applications and networks, (2) optimization of network functions, and (3) rapid development of new network protocols. All of these benefits are expected to contribute to improving the quality of diversifying communication networks and services. We identify three major technical challenges for enabling software-defined data plane as (1) ease of programming, (2) reasonable and predictable performance and (3) isolation among multiple concurrent logics. We also promote application-driving thinking towards defining software defined data-plane. We briefly introduce our project FLARE and its related technologies and review four use cases of flexible and deeply programmable data plane.
Due to limitations of today's widely-deployed commercial networks, some end-user applications are only possible through, or greatly improved by execution on virtualized networks that have been enhanced or idealized in a way which specifically supports the application. This paper describes US Ignite and the advantages provided to US Ignite end-user applications running on virtual networks which variously: (a) minimize latency, (b) minimize jitter, (c) minimize or eliminate packet drops, (d) optimize branch points for multicast packet duplication, (e) provide isolation for sensitive information flows, and/or (f) bundle network billing with application use. Examples of US Ignite applications in these categories are provided.
Akihiro NAKAO Kengo SASAKI Shu YAMAMOTO
We observe that P2P traffic has peculiar characteristics as opposed to the other type of traffic such as web browsing and file transfer. Since they exploit swarm effect -- a multitude of end points downloading the same content piece by piece nearly at the same time, thus, increasing the effectiveness of caching -- the same pieces of data end up traversing the network over and over again within mostly a short time window. In the light of this observation, we propose a network layer packet-level caching for reducing the volume of emerging P2P traffic, transparently to the P2P applications -- without affecting operations of the P2P applications at all -- rather than banning it, restricting it, or modifying P2P systems themselves. Unlike the other caching techniques, we aim to provide as generic a caching mechanism as possible at network layer -- without knowing much detail of P2P application protocols -- to extend applicability to arbitrary P2P protocols. Our preliminary evaluation shows that our approach is expected to reduce a significant amount of P2P traffic transparently to P2P applications.
We introduce a new kind of P2P traffic localization technique, called Netpherd, benefiting from the network virtualization technique for its successful deployment. Netpherd exploits one feature of P2P applications, a peer selection adaptation (i.e., preferring peers who are likely to provide better performance) for the traffic localization. Netpherd tries to enable local peers (i.e., peers in target network domain) to communicate with each other by affecting the peer selection adaptation. To affect the peer selection adaptation, Netpherd adds artificial delay to inter-domain traffic going to local peers. Our experiment conducted over Internet testbed verifies that Netpherd achieves the traffic localization and also improves the content download performance with the network delay insertion. In addition, we show that how the network virtualization technique can be utilized for efficient and graceful implementation of Netpherd.
John Russell LANE Akihiro NAKAO
Multipath routing and the ability to simultaneously use multiple network paths has long been proposed as a means for meeting the reliability and performance improvement goals of a next generation Internet. However, its use causes out-of-order packet delivery, which is well known to hinder TCP performance. While next-generation transport protocols will no doubt better cope with this phenomenon, a complete switch to these new protocols cannot be made on all devices "overnight"; the reality is that we will be forced to continue using TCP on such multipath networks well after deployment of a future Internet is complete. In this paper, we investigate the use of best-effort packet reordering -- an optional network layer service for improving the performance of any TCP session in the presence of out-of-order packet delivery. Such a service holds the promise of allowing unmodified TCP to take advantage of the reliability and performance gains offered by a future multipath-enabled Internet without suffering the adverse performance effects commonly associated with out-of-order packet delivery. Our experiments test the performance of two common TCP variants under packet dispersion with differing numbers of paths and amounts of inter-path latency variance. They were conducted using multipath network and packet reorderer implementations implemented within the Emulab testbed. Our results demonstrate that a simple best-effort reordering service can insulate TCP from the type of reordering that might be expected from use of packet dispersion over disjoint paths in a wide-area network, and is capable of providing significant performance benefits with few ill side-effects.
We propose TagFlow, a data plane mechanism for classification in Software-Defined Networking (SDN). We first argue that simple field-matching proposals of current SDN APIs are not efficient and flexible enough and then propose a tag based classification mechanism as an alternative. Moreover, we propose user-defined actions as an improvement over current hardcoded actions in SDN APIs. Our experiments show TagFlow forwarding is almost 40% faster than OpenFlow. Furthermore, our user-defined actions at SDN southbound are thousands of times faster that equivalent northbound implementations in the literature.
Network virtualization has become a common research topic that many researchers consider a basis for defining a new generation network architectures. In this paper, we attempt to clarify the concept of network virtualization with its brief history, to introduce the benefit of network virtualization for the future network, to posit our strong belief in that the future network should adopt a form of a meta-architecture that accommodates multiple competing multiple architectures, and to identify challenges to achieving this architecture.
In cloud computing, a cloud user pays proportionally to the amount of the consumed resources (bandwidth, memory, and CPU cycles etc.). We posit that such a cloud computing system is vulnerable to DDoS (Distributed Denial-of-Service) attacks against quota. Attackers can force a cloud user to pay more and more money by exhausting its quota without crippling its execution system or congesting links. In this paper, we address this issue and claim that cloud should enable users to pay only for their admitted traffic. We design and prototype such a charging model in a CoreLab testbed infrastructure and show an example application.
HyunYong LEE Masahiro YOSHIDA Akihiro NAKAO
Despite its great success, BitTorrent suffers from the content unavailability problem where peers cannot complete their content downloads due to some missing chunks, which is caused by a shortage of seeders who hold the content in its entirety. The multi-swarm collaboration approach is a natural choice for improving content availability, since content unavailability cannot be overcome by one swarm easily. Most existing multi-swarm collaboration approaches, however, suffer from content-related limitations, which limit their application scopes. In this paper, we introduce a new kind of multi-swarm collaboration utilizing a swarm as temporal storage. In a nutshell, the collaborating swarms cache some chunks of each other that are likely to be unavailable before the content unavailability happens and share the cached chunks when the content unavailability happens. Our approach enables any swarms to collaborate with each other without the content-related limitations. Simulation results show that our approach increases the number of download completions by over 50% (26%) compared to normal BitTorrent (existing bundling approach) with low overhead. In addition, our approach shows around 30% improved download completion time compared to the existing bundling approach. The results also show that our approach enables the peers participating in our approach to enjoy better performance than other peers, which can be a peer incentive.
Yufeng WANG Akihiro NAKAO Jianhua MA
As a concept stemmed from social field, we argued that, in P2P networks, peers' recommendation behaviors and functional behaviors should be explicitly separated, thus we propose the HopRec scheme which uses hop-based recommendation ability to improve the accuracy of reputation ranking in P2P networks. Our contributions lie in the following aspects: firstly, we adopt the simple but effective idea to infer peer's recommendation ability (RA): the farer away that peer is from the initial malicious seeds, the higher RA that peer should have; Then, the computation of reputation rankings appropriately reflects peer's different RA. The simulation results show that, in comparison with Eigentrust-like algorithms, HopRec can be robust to sybils and front peers attacks, and achieve significant performance improvement. Moreover, we compare HopRec with two related schemes, Poisonedwater and CredibleRank, and found that: in hospitable P2P environment, HopRec can obtain better performance than Poisonedwater, and can achieve the comparable performance as CredibleRank, with less computation overhead then CredibleRank. Finally, we also show that, if the initial good and malicious seeds could be selected based on peers' degrees, then HopRec and CredibleRank can achieve perfect performance.