The main purpose of this seminar is to function as a "curated reading group." Each week everyone will read either a single paper or a few papers loosely related by topic area, and we will discuss the papers and common themes together. This helps us be more aware of what other researchers are doing, and see examples of successful papers that get published in top venues. Papers in this reading group generally make use of, or enhance, embedded networked systems research.

To ensure quality we will restrict ourselves to the top conferences such as SenSys, IPSN, MobiSys, and ASPLOS. We will also look at interesting papers from related conferences such as SIGMETRICS, PLDI, UbiComp, SOSP/OSDI, Mobicom, SIGCOMM, OOPSLA, UIST, Oakland, etc.

Expectations

All seminar participants are required to read "How to Read a Paper" before the seminar starts.This invaluable text details a practical and efficient three-pass method for reading research papers, saving time and helping with comprehension.

It is expected that each person will have read each paper enough to grasp core concepts / contributions, identify weaknesses and strengths, be able to place the paper in the literature, and compare against other papers read. Come with questions and comments on the paper's insights, novelty, significance, assumptions, methodology, flaws, etc. You specifically want to look for the following:

• One positive insight into each work; for example a novel evaluation method that could be used in your own research.
• One negative insight into each work; for example pointing out a flaw in methodology.
• One general observation about overall topic area of the papers; this could be identifying major problems, talking about holes in the literature, or forecasting the future of the field.

To help you organize your thoughts on the paper, please use this review form (similar to what you would use if you were reviewing for a conference). If you're taking the seminar for credit, please submit your review form by 8 AM the day before we meet.