From: "Xinghua Shi" Date: Sunday, May 05, 2002 11:04 PM Paper1: Next Century Challenges: Scalable Coordination in Sensor Networks Contribution: The paper studies the principles underlying the design if services and applications in sensor networks. The authors also describe directed diffusion, a promising model for describing localized algorithms. Rate: a. 3 in significance: This paper has significantmoderate contribution because it deals with the study of sensor networks but doesn't provide much valuable information. b. 3 in convincing of methodology: The argument is based on some experiments but not well designed. There are two many arguments coming from intuition but not fully explained. c. The most important limitation of the approach is that there're few sound proof, some of their ideas are not so clear as they lack proofs. The strongest and/or most interesting ideas in the paper: a. They hypothesized that sensor network coordination applications are better realized using localized algorithms and then then proposed such a algorithm and illustrate an application of this. Questions: a. What's the current development of sensor networks? b. How about other localized algorithms? Paper2: System Architecture Directions for Networked Sensors. Contribution: This paper provides an initial exploration of system architectures for networked sensors based on a prototype. The authors also developed a tiny microthreaded OS, TinyOS, to deal with the key missing technology in their project. They addressed the problems of concurrency intensive and efficient modularity with their TinyOS. Through the analysis of the results, there comes out valuable directions for future architectural innovation.They also studied the design requirements that characterize the networked sensor regime and guide their microthreading approach. Rate: a.4 in significance: This paper has significant importance because it provides a prototype which is not only important as the base of their system design study, but also has much value for future research in this field. b.4 in convincing of methodology: The argument is based on the results of their small, flexible networked sensor platform. Their evaluation is based on quantitative analysis of the results which is convincing. The strongest and/or most interesting ideas in the paper: a. Drive the networked sensor down to microscopic scale by taking advantage of advances in semiconductor processes. b. Make the task scheduler be power aware by putting the processor to sleep when the task queue is empty but leaving the peripherals operating, so that any of them can wake the sytem. c.It's possible to maintain multiple flows of data with a single microcontroller, thus it's not a requirement to uitilize individual microcontrollers per device. The most striking weaknesses in the paper: a. They can develop more other applications to illustrate the evaluation of their system. Questions: a. What's about other research in this feild? Are there some reusable software addressing this issue? b. Are there any benchmarks to simulate and evaluate the sensor networks? ################################################## ################################################## From: "Yu Hu" Date: Sunday, May 05, 2002 11:56 PM YU HU Next Century Challenges: Scalable Coordination in Sensor Network Main contributeion of the paper: This paper showed us a brief view of sensor network and mentioned some sensor network's different charaters. Based on these , authors trusted sensor network will organize in different way. Author introdued localized algorithms and discussed directed diffusions model. This paper leaded us into the sensor network anyway. Critique the main contribution: Significance : 3 Methodology:This paper just give a brief view and theoretical analysis on sensor network. It only showed some ideas to us . I think maybe should try to make these ideas to implement. Of course , the intuition that authors gave us make sense somewhat. The most important limitation of the approach: Author depicted the diffusion model. But this model to me is not very clear after reading. Maybe authors should give me details and analysis of this model. Most interesting ideas: The charater of sensor network made nodes just need to communicate with near neighbours, which can work well. It is really interesting and can make sensor network scalable. Weakness&Questions: Firstly, maybe authors should give more details about the model ; And I want toknow more about sensor network's scalability. System Architecture Directions for Networked Sensors Main contributeion of the paper: The paper showed us a tiny event-driven operating system( TinyOs)and using some tests to prove that it provided support for efficient modularity and concurrency-intensive operation. Critique the main contribution: Significance : 3 Methodology:Authors began from sensor network charaters and some design issues to TinyOs design rules and its tests.It sounds it a good event-driven operating system. Maybe authors should give more details about it and give some rationales of this idea. Most interesting ideas: The paper produced us an overall system architecture and a methodology for systematic advance. If the idea is right, it is really of significance to sensor network design. Weakness&Questions: Maybe we need more details of sensor network's difference between nowday's network; Does some standard about sensor networks exist?