acm sigcomm

The late noughties have seen an influx of work in different scientific disciplines, all addressing the question of 'design' and 'architecture'. It is a battle between those advocating the theory of 'emergent properties' and others who strive for a 'theory for 'architecture'. We provide a particular insight into this battle, represented in the form of a story that focuses on the role of a possibly unusual protagonist and his influence on computer science, the Internet, architecture and beyond. We show his relation to one of the great achievements of system engineering, the Internet, and the possible future as it might unfold. Note from the writer: The tale is placed in a mixture of reality and fiction, while postulating a certain likelihood for this fiction. There is no proof for the assertions made in this tale, leaving the space for a sequel to be told.

In Named Data Networking (NDN) architecture, packets carry data names rather than source or destination addresses. This change of paradigm leads to a new data plane: data consumers send out Interest packets, routers forward them and maintain the state of pending Interests, which is used to guide Data packets back to the consumers. NDN routers' forwarding process is able to detect network problems by observing the two-way traffic of Interest and Data packets, and explore multiple alternative paths without loops. This is in sharp contrast to today's IP forwarding process which follows a single path chosen by the routing process, with no adaptability of its own. In this paper we outline the design of NDN's adaptive forwarding, articulate its potential benefits, and identify open research issues.

Desynchronization is useful for scheduling nodes to perform tasks at different time. This property is desirable for resource sharing, TDMA scheduling, and collision avoiding. Inspired by robotic circular formation, we propose DWARF (Desynchronization With an ARtificial Force field), a novel technique for desynchronization in wireless networks. Each neighboring node has artificial forces to repel other nodes to perform tasks at different time phases. Nodes with closer time phases have stronger forces to repel each other in the time domain. Each node adjusts its time phase proportionally to its received forces. Once the received forces are balanced, nodes are desynchronized. We evaluate our implementation of DWARF on TOSSIM, a simulator for wireless sensor networks. The simulation results indicate that DWARF incurs significantly lower desynchronization error and scales much better than existing approaches.

In many wireless systems, it is desirable to precede a data transmission with a handshake between the sender and the receiver. For example, RTS-CTS is a handshake that prevents collisions due to hidden terminals. Past work, however, has shown that the overhead of such handshake is too high for practical deployments. We present a new approach to wireless handshake that is almost overhead free. The key idea underlying the design is to separate a packet's PLCP header and MAC header from its body and have the sender and receiver first exchange the data and ACK headers, then exchange the bodies of the data and ACK packets without additional headers. The header exchange provides a natural handshake at almost no extra cost. We empirically evaluate the feasibility of such lightweight handshake and some of its applications. Our testbed evaluation shows that header-payload separation does not hamper packet decodabilty. It also shows that a light handshake enables hidden terminals, i.e., nodes that interfere with each other without RTS/CTS, to experience less than 4% of collisions. Furthermore, it improves the accuracy of bit rate selection in bursty and mobile environments producing a throughput gain of about 2x.

This paper introduces libtrace, an open-source software library for reading and writing network packet traces. Libtrace offers performance and usability enhancements compared to other libraries that are currently used. We describe the main features of libtrace and demonstrate how the libtrace programming API enables users to easily develop portable trace analysis tools without needing to consider the details of the capture format, file compression or intermediate protocol headers. We compare the performance of libtrace against other trace processing libraries to show that libtrace offers the best compromise between development effort and program run time. As a result, we conclude that libtrace is a valuable contribution to the passive measurement community that will aid the development of better and more reliable trace analysis and network monitoring tools.

In spite of the tremendous amount of measurement efforts on understanding the Internet as a global system, little is known about the 'local' Internet (among ISPs inside a region or a country) due to limitations of the existing measurement tools and scarce data. In this paper, empirical in nature, we characterize the evolution of one such ecosystem of local ISPs by studying the interactions between ISPs happening at the Slovak Internet eXchange (SIX). By crawling the web archive waybackmachine.org we collect 158 snapshots (spanning 14 years) of the SIX website, with the relevant data that allows us to study the dynamics of the Slovak ISPs in terms of: the local ISP peering, the traffic distribution, the port capacity/utilization and the local AS-level traffic matrix. Examining our data revealed a number of invariant and dynamic properties of the studied ecosystem that we report in detail.

There are many deployed approaches for blocking unwanted traffic, either once it reaches the recipient's network, or closer to its point of origin. One of these schemes is based on the notion of traffic carrying capabilities that grant access to a network and/or end host. However, leveraging capabilities results in added complexity and additional steps in the communication process: Before communication starts a remote host must be vetted and given a capability to use in the subsequent communication. In this paper, we propose a lightweight mechanism that turns the answers provided by DNS name resolution - which Internet communication broadly depends on anyway - into capabilities. While not achieving an ideal capability system, we show the mechanism can be built from commodity technology and is therefore a pragmatic way to gain some of the key benefits of capabilities without requiring new infrastructure.

Operators have deployed Multiprotocol Label Switching (MPLS) in the Internet for over a decade. However, its impact on Internet topology measurements is not well known, and it is possible for some MPLS configurations to lead to false router-level links in maps derived from traceroute data. In this paper, we introduce a measurement-based classification of MPLS tunnels, identifying tunnels where IP hops are revealed but not explicitly tagged as label switching routers, as well as tunnels that obscure the underlying path. Using a large-scale dataset we collected, we show that paths frequently cross MPLS tunnels in today's Internet: in our data, at least 30% of the paths we tested traverse an MPLS tunnel. We also propose and evaluate several methods to reveal MPLS tunnels that are not explicitly flagged as such: we discover that their fraction is significant (up to half the explicit tunnel quantity) but most of them do not obscure IP-level topology discovery.

Time tends to pass more quickly than we would like. Sometimes it is helpful to reflect on what you have accomplished, and to derive what you have learned from the experiences. These "lessons learned" may then be leveraged by yourself or others in the future. Occasionally, an external event will motivate this self reflection. For me, it was the 50th anniversary reunion of the St. Walburg Eagles, held in July 2011. The Eagles are a full-contact (ice) hockey team I played with between 1988 and 1996 (the Eagles ceased operations twice during this period, which limited me to four seasons playing with them), while attending university. What would I tell my friends and former teammates that I had been doing for the past 15 years? After some thought, I realized that my time as an Eagle had prepared me for a research career, in ways I would never have imagined. This article (an extended version with color photos is available in [1]) shares some of these similarities, to motivate others to reflect on their own careers and achievements, and perhaps make proactive changes as a result.

It has become a truism that innovation in the information and communications technology (ICT) fields is occurring faster than ever before. This paper posits that successful innovation requires three essential elements: a need, know-how or knowledge, and favorable economics. The paper examines this proposition by considering three technical areas in which there has been significant innovation in recent years: server virtualization and the cloud, mobile application optimization, and mobile speech services. An understanding of the elements that contribute to successful innovation is valuable to anyone that does either fundamental or applied research in fields of information and communication technology.

I'd like to devote this editorial to a description of the process we use to select and publish papers submitted to CCR. CCR publishes two types of papers: technical papers and editorials.  I'll first describe the process for technical papers then for editorials.

Internet traffic has Zipf-like properties at multiple aggregation levels. These properties suggest the possibility of offloading most of the traffic from a complex controller (e.g., a software router) to a simple forwarder (e.g., a commodity switch), by letting the forwarder handle a very limited set of flows; the heavy hitters. As the volume of traffic from a set of flows is highly dynamic, maintaining a reliable set of heavy hitters over time is challenging. This is especially true when we face a volume limit in the non-offloaded traffic in combination with a constraint in the size of the heavy hitter set or its rate of change. We propose a set selection strategy that takes advantage of the properties of heavy hitters at different time scales. Based on real Internet traffic traces, we show that our strategy is able to offload most of the traffic while limiting the rate of change of the heavy hitter set, suggesting the feasibility of alternative router designs.

Unsolicited one-way Internet traffic, also called Internet background radiation (IBR), has been used for years to study malicious activity on the Internet, including worms, DoS attacks, and scanning address space looking for vulnerabilities to exploit. We show how such traffic can also be used to analyze macroscopic Internet events that are unrelated to malware. We examine two phenomena: country-level censorship of Internet communications described in recent work, and natural disasters (two recent earthquakes). We introduce a new metric of local IBR activity based on the number of unique IP addresses per hour contributing to IBR. The advantage of this metric is that it is not affected by bursts of traffic from a few hosts. Although we have only scratched the surface, we are convinced that IBR traffic is an important building block for comprehensive monitoring, analysis, and possibly even detection of events unrelated to the IBR itself. In particular, IBR offers the opportunity to monitor the impact of events such as natural disasters on network infrastructure, and in particular reveals a view of events that is complementary to many existing measurement platforms based on (BGP) control-plane views or targeted active ICMP probing.

Long-term historical analysis of captured network traffic is a topic of great interest in network monitoring and network security. A critical requirement is the support for fast discovery of packets that satisfy certain criteria within large-scale packet repositories. This work presents the first indexing scheme for network packet traces based on compressed bitmap indexing principles. Our approach supports very fast insertion rates and results in compact index sizes. The proposed indexing methodology builds upon libpcap, the de-facto reference library for accessing packet-trace repositories. Our solution is therefore backward compatible with any solution that uses the original library. We experience impressive speedups on packet-trace search operations: our experiments suggest that the index-enabled libpcap may reduce the packet retrieval time by more than 1100 times.

ADSL and cable connections are the prevalent technologies available from Internet Service Providers (ISPs) for residential Internet access. Asymmetric access technologies such as these offer high download capacity, but moderate upload capacity. When the Transmission Control Protocol (TCP) is used on such access networks, performance degradation can occur. In particular, sharing a bottleneck link with different upstream and downstream capacities among competing TCP flows in opposite directions can degrade the throughput of the higher speed link. Despite many research efforts to solve this problem in the past, there is no solution that is both highly effective and easily deployable in residential networks. In this paper, we propose an Asymmetric Queueing (AQ) mechanism that enables full utilization of the bottleneck access link in residential networks with asymmetric capacities. The extensive simulation evaluation of our design shows its effectiveness and robustness in a variety of network conditions. Furthermore, our solution is easy to deploy and configure in residential networks.

This editorial was motivated by a panel on the relationship between academia and industry at the SIGCOMM 2011 conference that was moderated by Bruce Davie. I can claim some familiarity with the topic having spent roughly ten years each in academia and industry during the last twenty years.

My thesis is that although industry can make incremental gains, real technical breakthroughs can only come from academia. However, to have any impact, these academic breakthroughs must be motivated, at some level, by a real-world problem and the proposed solutions should be feasible, even if implausible. Therefore, it is in the self-interest of industry to fund risky, longterm, curiosity-driven academic research rather than sure-shot, short-term, practical research with welldefined objectives. Symmetrically, it is in the self-interest of academic researchers to tackle real-world problems motivated by the problems faced in industry and propose reasonable solutions.

There are many underlying reasons why technological revolutions today can only come from academia. Perhaps the primary reason is that, unlike most industrial research labs of today (and I am purposely excluding the late, great, Bell Labs of yore), academia still supports long-term, curiosity-driven research. This is both risky and an inherently `wasteful’ use of time. Yet, this apparently wasteful work is the basis for many of today’s technologies, ranging from Google search to the World Wide Web to BSD Unix and Linux. On closer thought, this is not too surprising. Short-term, practical research requires the investigator to have well-defined goals. But revolutionary ideas cannot be reduced to bullet items on Powerpoint slides: they usually arise as unexpected outcomes of curiosity-driven research. Moreover, it takes time for ideas to mature and for the inevitable missteps to be detected and corrected. Industrial funding cycles of six months to a year are simply not set up to fund ideas whose maturation can take five or even ten years. In contrast, academic research built on the basis of academic tenure and unencumbered by the demands of the marketplace is the ideal locus for long-term work.

Long-term, curiosity-driven research alone does not lead to revolutions. It must go hand-in-hand with an atmosphere of intellectual openness and rigour. Ideas must be freely exchanged and freely shot down.

The latest work should be widely disseminated and incorporated into one’s thinking. This openness is antithetical to the dogma of `Intellectual Property’ by which most corporations are bound. Academia, thankfully, has mostly escaped from this intellectual prison. Moreover, industry is essentially incompatible with intellectual rigour: corporate researchers, by and large, cannot honestly comment on the quality of their own company’s products and services.

A third ingredient in the revolutionary mix is the need for intense thinking by a dedicated group of researchers. Hands-on academic research tends to be carried out by young graduate students (under the supervision of their advisors) who are unburdened by either responsibilities or by knowing that something just cannot be done. Given training and guidance, given challenging goals, and given a soul-searing passion to make a difference in the world, a mere handful of researchers can do what corporate legions cannot.

These three foundations of curiosity-driven research, intellectual openness, and intense thinking set academic research apart from the industrial research labs of today and are also the reason why the next technological revolution is likely to come from academia, not industry.

In the foregoing, I admit that I have painted a rather rosy picture of academic research. It is important to recognize, however, that the same conditions that lead to breakthrough research also are susceptible to abuse. The freedom to pursue long-term ideas unconstrained by the marketplace can also lead to work that is shoddy and intellectually dishonest. For instance, I believe that it may be intellectually honest for a researcher to make assumptions that do not match current technology, but it is intellectually dishonest to make assumptions that violate the laws of physics. In a past editorial, I have written in more depth about these assumptions, so I will not belabour the point. I will merely remark here that it is incumbent on academic researchers not to abuse their freedom.

A second inherent problem with academic research, especially in the field of computer networking, is that it is difficult, perhaps impossible, to do large-scale datadriven research. As a stark example, curiosity-driven work on ISP topology is impossible if ISPs sequester this data. Similarly, studying large-scale data centre topology is challenging when the largest data centre one can build in academia has only a few hundred servers.

Finally, academic research tends to be self-driven and sometimes far removed from real-world problems.  These real-world problems, which are faced daily by industrial researchers, can be intellectually demanding and their solution can be highly impactful. Academic researchers would benefit from dialogue with industrial researchers in posing and solving such problems.

Given this context, the relationship between academia and industry becomes relatively clear. What academia has and industry needs is committed, focussed researchers and the potential for long-term, revolutionary work. What industry has and academia needs is exposure to real-world problems, large-scale data and systems, and funding. Therefore, it would be mutually beneficial for each party to contribute to the other. Here are a few specific suggestions how.

First, industry should fund academic research without demanding concrete deliverables and unnecessary constraints. Of course, the research (and, in particular, the research assumptions) should be adequately monitored. But the overall expectation should be that academic work would be curiosity-driven, open, and long-term.

Second, industry should try to expose academic researchers to fundamental real-world problems and put at their disposal the data that is needed for their solution. If necessary, academic researchers should be given access to large-scale systems to try out their solutions. This can be done without loss of intellectual property by having students and PIs visit industrial research labs as interns or during sabbaticals. It could also be done by having industrial researchers spend several weeks or months as visitors to university research labs.

Third, industry should spend resources not only on funding, but on internal resources to match the output of academic research (papers and prototypes) to their own needs (products and systems).

Fourth, academic researchers should choose research problems based not just on what is publishable, but (also) based on the potential for real-world impact. This would naturally turn them to problems faced by industry.

Fifth, academic researchers should ensure that their solutions are feasible, even if implausible. For instance, a wireless system for cognitive radio built on USRP boards is implausible but feasible. In contrast, a wireless system that assumes that all radio coverage areas are perfectly circular is neither plausible nor feasible. This distinction should be emphasized in the academic review of technical papers.

Finally, academic researchers should recognize the constraints under which industry operates and, to the extent possible, accommodate them. For instance, they should encourage students to take on internships, fight the inevitable battles with the university office of research to negotiate IP terms, and understand that their points of contact will change periodically due to the nature of corporate (re-)organizations.

The SIG can also help this interaction. Industry-academic fora such as the panel at SIGCOMM, industryspecific workshops, and industry desks at conferences allow academic researchers to interact with representatives from industry. SIGCOMM could have tutorials focussed on topics of current interest to industry. These two efforts would certainly make deep collaboration between academia and industry more likely.

I hope that these steps will move our community towards a future where academic research, though curiosity-driven, continues to drive real-world change because of its symbiotic relationship with industrial partners.

This editorial benefited from comments from Bruce Davie and Gail Chopiak. 

The Internet has changed dramatically in recent years. In particular, the fundamental change has occurred in terms of who generates most of the content, the variety of applications used and the diverse ways normal users connect to the Internet. These factors have led to an explosion of the amount of user-specific meta-information that is required to access Internet content (e.g., email addresses, URLs, social graphs). In this paper we describe a foundational service for storing and sharing user-specific meta-information and describe how this new abstraction could be utilized in current and future applications.

Should a new "platform" target a functionality-rich but complex and expensive design or instead opt for a bare-bone but cheaper one? This is a fundamental question with profound implications for the eventual success of any platform. A general answer is, however, elusive as it involves a complex trade-off between benefits and costs. The intent of this paper is to introduce an approach based on standard tools from the field of economics, which can offer some insight into this difficult question. We demonstrate its applicability by developing and solving a generic model that incorporates key interactions between platform stakeholders. The solution confirms that the "optimal" number of features a platform should offer strongly depends on variations in cost factors. More interestingly, it reveals a high sensitivity to small relative changes in those costs. The paper's contribution and motivation are in establishing the potential of such a cross-disciplinary approach for providing qualitative and quantitative insights into the complex question of platform design.