COMM

Welcome to the 2012 ACM SIGCOMM Conference in Helsinki, Finland, which is a wonderful location and incidentally is the farthest north that the conference has ventured to in its long history! It is our pleasure to introduce the technical program of the conference. This year's program continues SIGCOMM's tradition of presenting outstanding work that suggests new areas of research in networking. The program committee accepted 31 papers selected from 235 submissions. We felt it was important to have a broad view of networking; this allowed us to choose papers on such diverse topics as on-chip networking, key-value stores, and measuring click spam. While this year's program continues to reflect a recent renaissance in data centers and wireless, traditional topics such as routing, QoS, and measurement are also well represented.

The papers came from a rich mix of authors from both academia and industry across the globe. Authors from organizations in around 25 countries submitted papers, and the final program includes authors from over 30 organizations in 8 countries. It was a delight to work with a terrific technical program committee of 49 members from all over the world and including distinguished academic and industrial researchers. We chose the PC with care, trying to balance experience with youth, and recognizing authors who have made consistent technical contributions in SIGCOMM and other top conferences, while ensuring broad technical area coverage and a wide geographic representation. Members of the PC were drawn from 9 countries in 4 continents, with a quarter of the PC comprising first-time members, both relatively young networking researchers and experienced researchers from sister areas such as operating systems and mobile systems. We are grateful for the insight and care that our PC displayed, working hard and responding to our every request. We are also delighted to report that the first-time PC members were a wonderful addition, enriching the PC with their insights and exhibiting a high degree of diligence and professionalism.

We continued the tradition of a two-round review process, with a first round of three reviews for each paper. Around half the submitted papers were chosen for the second round, which then generated three more reviews. After the second round and a subsequent online discussion phase, around 60 papers were chosen for discussion at the PC meeting. As chairs we went less by numerical score in order to decide which paper to advance than by the actual reviews, realizing that numerical scores vary widely with reviewers. In a few cases, we sought external reviewers to obtain an expert opinion. The PC meeting was held in San Jose, CA, USA, over a day and a half. While all of the papers in the discussion set were discussed and had tentative decisions made on day 1, the half-day on day 2 allowed plenty of time for the PC to revisit, and in some cases reverse, the decisions made on day 1, and even bring up papers for discussion papers that hadn't been shortlisted for discussion on day 1.

Despite following the traditional process, this year's review process was marked by two new mechanisms. First, we reintroduced an author response phase, which had last been tried in SIGCOMM over a decade ago. To make the author response effective, we formulated a careful set of instructions, including illustrative examples, to help make the author responses more factual and effective. The author response phase was intended partly to reduce the possibility of mistaken reviews, but also to promote an approximation to the healthy dialog between authors and reviewers inherent in the classical journal process. We were delighted with the quality of the author responses we saw even with (eventually) rejected papers: respectful, factual, and insightful. The responses were all considered as part of the PC discussion, and some indeed changed the reviewers' minds. Second, the entire PC made an attempt to provide constructive reviews, emphasizing the positive, and suggesting where the paper could be improved. To make explicit the value we placed on good reviews and recognize exceptional contributions in this regard by PC members, we instituted 3 Outstanding Reviewer Awards, which will be presented at the conference.

The selection of the award papers started with nominations made by PC members. We as the PC chairs shortlisted 5 of the nominated papers and passed these on to an awards committee chaired by David Wetherall and also comprising T.V. Lakshman, Dave Levin, and Ratul Mahajan, all members of the PC. We are grateful to the members of the awards committee for conducting a thorough process and sending us their selections in quick time. It is a pleasure to present the Best Paper Award for ACM SIGCOMM 2012 to the paper titled "Multi-Resource Fair Queueing for Packet Processing" by Ali Ghodsi, Vyas Sekar, Matei Zaharia, and Ion Stoica, and the Best Student Paper Award to the paper titled "Picasso: Flexible RF and Spectrum Slicing" by Steven Hong, Jeff Mehlman, and Sachin Katti. For the Outstanding Reviewer Awards, we, the PC chairs, made the selections based on the quality and timeliness of reviewing, participation in the online and PC meeting discussions, and the willingness to go the extra mile when needed. It was challenging for us to make a selection from among the many PC members who contributed so strongly, but after careful thought, it is our pleasure to recognize Aditya Akella, Matt Caesar, and Dave Levin with the Outstanding Reviewer Awards for SIGCOMM 2012. Congratulations to all of the award winners, authors as well as reviewers.