tag:blogger.com,1999:blog-9186457242428335144.post7310480983415051658..comments2024-03-28T00:36:13.790-07:00Comments on Volatile and Decentralized: SenSys 2009, Day TwoMatt Welshhttp://www.blogger.com/profile/04255792550910131960noreply@blogger.comBlogger6125tag:blogger.com,1999:blog-9186457242428335144.post-90190527247793954912010-11-22T10:30:46.137-08:002010-11-22T10:30:46.137-08:00What a great company Dust Networks is, my older br...What a great company Dust Networks is, my older brother works there and he tells me only wonderful things about the company and how they treat their employees.viagra onlinehttp://iservepharmacy.com/noreply@blogger.comtag:blogger.com,1999:blog-9186457242428335144.post-60594471378367948702010-11-22T07:42:43.172-08:002010-11-22T07:42:43.172-08:00Thank you for the information.
Why this guy Kris P...Thank you for the information.<br />Why this guy Kris Pister don't do that kind fo bussnies with microsft or apple ?<br />I think it would be better, well It's just an idea?<br /><br />Thanksviagra onlinehttp://www.safemeds.com/noreply@blogger.comtag:blogger.com,1999:blog-9186457242428335144.post-40938951360310491252009-11-07T15:00:57.269-08:002009-11-07T15:00:57.269-08:00James, 802.15.4 actually specifies TDMA and CSMA w...James, 802.15.4 actually specifies TDMA and CSMA working together. TDMA is in fact exciting for one major reason and that is power savings. In WSN, that is almost everything. You can duty cycle the whole network so that all devices save power while remaining synchronized to each other's wake up cycle.<br /><br />The frequency hopping serves a different purpose which is robustness to channel interference. Believe it or not, the reason why WSNs haven't penetrated more is because people are still tryin to figure out the power issues to keep them on for years at a time without replacing batteries.Akibahttp://www.freaklabs.orgnoreply@blogger.comtag:blogger.com,1999:blog-9186457242428335144.post-33690158937840645762009-11-06T08:30:55.043-08:002009-11-06T08:30:55.043-08:00Steve - the issue in most WSN scenarios is that we...Steve - the issue in most WSN scenarios is that we don't have the luxury of completely decoupling the communication and application layers, due to limited buffer capacity and the need to tightly coordinate communication schedules with sampling, computation, etc. The dependence issue is a big part of it; once you have signed up to use TDMA it's hard to back out.<br /><br />James - check out the full program; I'm only blogging on a small subset of the papers. It's true that the community is dominated by networking research so the question is, as we start to actually solve those problems, how do we get the community to evolve to work on the next set of hard problems. One problem is that papers that branch out often get out of the comfort zone of the program committee so having the right people on the PC is essential.Matt Welshhttps://www.blogger.com/profile/04255792550910131960noreply@blogger.comtag:blogger.com,1999:blog-9186457242428335144.post-5777108960585087982009-11-06T05:26:40.418-08:002009-11-06T05:26:40.418-08:00TDMA vs. CSMA? This is what gets sensor nets peop...TDMA vs. CSMA? This is what gets sensor nets people excited? My god! The panel sounded like the only useful thing there. Wake up guys. You need to figure out what to use this stuff for! And being limited by quant. analysis is a big problem in other communities also. You need to figure out a way around it. I think the HCI community (my community) is suffering from this same problem, but it is just a "looser" level of rigor. Forcing everything to look like a tight experiment limits what you can ever build and explore. There should be room for both styles of work (tight experiments about something small and looser explorations of big ideas).Anonymoushttps://www.blogger.com/profile/15776616183969942139noreply@blogger.comtag:blogger.com,1999:blog-9186457242428335144.post-42678226054689286572009-11-06T01:01:42.069-08:002009-11-06T01:01:42.069-08:00I'm not sure how much I agree with your conten...I'm not sure how much I agree with your contention that TDMA schemes muddy the water between communication substrate and application logic. Nobody's claimed that about token-ring, and IP runs over lots of TDMA links (GSM,LTE,WIMAX, even the GTS parts of 802.15.4). IP doesn't have any particular considerations for this to work… Furthermore, it's simply a fact that TDMA has higher spectral efficiency and can make QoS guarantees. It's not that I'm particularly in favor of TDMA, I'm just agreeing with KJP that it often seems overlooked.<br /><br />Of course there are scheduled or pseudo-scheduled MACs around- SCP-MAC or synchronous LPL/B-MAC are pretty well known. I think a better objection is a sort of wariness to baking in a dependence on consistent schedules to your protocol, especially for ad-hoc, multi-hop applications. <br /><br />FYI 802.15.4e is on the horizon that has a lot of elements of TSMP-- there's even a (new,open) implementation out of Kris's group-- see http://wsn.eecs.berkeley.edu/projects/openwsn/wiki<br />I'm definitely going to be taking a close look at it.stevedhhttps://www.blogger.com/profile/16915672376249457629noreply@blogger.com