Guidance for Authors of Position Papers

For the third year in a row, SIGCOMM is soliciting submissions of position papers in addition to regular full-length papers. Position papers are five to eight (5-8) pages long.  The main aim of a position paper is to shape research direction, draw attention to an unexplored (or overexplored!) research topic, argue a controversial opinion, or otherwise stretch our thinking as researchers.  This is in sharp contrast to a work-in-progress paper or a shortened conference paper, neither of which would fare well if submitted as a position paper.

Position papers should contribute perspective rather than performance numbers, wisdom rather than knowledge, and guidance rather than results.  We expect them to be broad in scope, forward-looking, and provocative, rather than narrowly focused on a detailed solution a specific technical problem.  We also expect them to be more controversial than full papers and not necessarily present quantitative results, though they must remain well-argued and justified in terms of existing work. For examples of position papers, see the conference programs for SIGCOMM 2003 and SIGCOMM 2002. Also, see other venues such as HotNets'2003 and HotNets'2002, especially the papers that are broad in scope.

Here are some hypothetical examples:

• An argument that [network troubleshooting, formal testing of implementations, application-level protocols, economic incentives] are important but undervalued considerations in networking research, along with a promising approach for addressing them.
• A blue-sky redesign of how [IP, DNS, congestion control, Internet routing, Internet measurement] would be architected today that identifies the key differences with existing protocols and argues how these should alter our research.
• An argument that conventional wisdom in the area of [QoS, active networking, IP multicast, congestion control, packet switching, overlay networks] is misguided that uses its insights to lay out an alternative strategy to address the same underlying problems.
• A discussion of how emerging technologies for [quantum networking, untethered sensors, network processors, large-scale packet simulation] will change networking or network research and why these changes matter.

• "Case For" papers that examine an old problem in a new light (or shed the first light on an entirely new problem), and propose a new research agenda.
• Case studies that synthesize examples that show how to (or how not to) design future architectures, systems, or protocols.
• Papers that put forth architectural principles or frameworks for the design or analysis of communication systems, or that re-examine well-known principles or frameworks in light of changes in the world of networking since they were originally conceived.
• Papers that apply techniques or organizing principles from outside of networking to a networking problem, and demonstrate how this is likely to lead to new approaches or solutions.
  
• Papers that argue that "Research in Topic X Considered Harmful", where X is a topic that the community is actively pursuing (e.g., congestion control, overlay networks, traffic measurement, sensor networks, peer-to-peer systems, etc.).