Dear all, i heard several times.. some of our current researches in wireless comm doesnt not reflect the real world scenario. but... i asked some professionals.. most comment on channel state assumptions such as Rayleigh and its counterparts... not others.... i would like to know and ask those industrial guru in our group........... are there other gaps in current wireless comm researches and its industrial applications... how much severe.... with regards,
which current researches of wireless comm doesnt not reflect the reality
Started by ●December 1, 2008
Reply by ●December 1, 20082008-12-01
On 1 Des, 11:33, "euler.shan...@gmail.com" <euler.shan...@gmail.com> wrote:> Dear all, > > i heard several times.. some of our current researches in wireless > comm doesnt not reflect the real world scenario. but... i asked some > professionals.. most comment on channel state assumptions such as > Rayleigh and its counterparts... not others....Don't know much about communications, but any real-world application of maths, physics or statistics can be said to 'not reflect reality.' The reason is that one can only do computations and calculations with the simplest, idealized models. There will always be some factors left out of the models, and those that are included are idealized. One doesn't use the linear model because it is 'true', but because it is possible to do computations with it. One doesn't use Gaussian models because they are 'true' but because they are described by merely two moments, and thus are tractable. So whenever we can, we push the simple idealized linear models as far as possible (and all too often a bit and a half further), and don't include nonlinear terms or non-Gaussian statistics unless we really have to. So physical or statistical modeling is not about 'truth' or 'reality', but about finding a model that is useful and yet tractable. Rune
Reply by ●December 1, 20082008-12-01
Rune Allnor wrote:> On 1 Des, 11:33, "euler.shan...@gmail.com" <euler.shan...@gmail.com> > wrote: >> Dear all, >> >> i heard several times.. some of our current researches in wireless >> comm doesnt not reflect the real world scenario. but... i asked some >> professionals.. most comment on channel state assumptions such as >> Rayleigh and its counterparts... not others.... > > Don't know much about communications, but any real-world > application of maths, physics or statistics can be said > to 'not reflect reality.' > > The reason is that one can only do computations and > calculations with the simplest, idealized models. There > will always be some factors left out of the models, and > those that are included are idealized. > > One doesn't use the linear model because it is 'true', > but because it is possible to do computations with it. > One doesn't use Gaussian models because they are 'true' > but because they are described by merely two moments, > and thus are tractable. > > So whenever we can, we push the simple idealized linear > models as far as possible (and all too often a bit and > a half further), and don't include nonlinear terms > or non-Gaussian statistics unless we really have to. > > So physical or statistical modeling is not about 'truth' > or 'reality', but about finding a model that is useful > and yet tractable. > > RuneAnd when we _do_ make a model more realistic by including nonlinear terms, or by removing the Gaussian assumption from our random processes, we do so using highly simplified models, and only enough to get a good enough match with reality so that we may proceed. And it's for the reason that Rune states -- because you need a mathematical model that is tractable enough that you can get something useful out of it in less time and for less effort than you would just building a system and giving it a whirl. -- Tim Wescott Wescott Design Services http://www.wescottdesign.com Do you need to implement control loops in software? "Applied Control Theory for Embedded Systems" gives you just what it says. See details at http://www.wescottdesign.com/actfes/actfes.html
Reply by ●December 1, 20082008-12-01
Tim Wescott wrote:> Rune Allnor wrote: > >> On 1 Des, 11:33, "euler.shan...@gmail.com" <euler.shan...@gmail.com> >> wrote: >> >>> Dear all, >>> >>> i heard several times.. some of our current researches in wireless >>> comm doesnt not reflect the real world scenario. but... i asked some >>> professionals.. most comment on channel state assumptions such as >>> Rayleigh and its counterparts... not others.... >> >> >> Don't know much about communications, but any real-world >> application of maths, physics or statistics can be said >> to 'not reflect reality.' >> >> The reason is that one can only do computations and >> calculations with the simplest, idealized models. There >> will always be some factors left out of the models, and >> those that are included are idealized. >> >> [SNIP *MAJOR* portion of Rune's and all of Tim's comments]I do not know if I am missing the OP's intended point [but I *DO* think that his post could refer to a *VALUABLE* point ;] Rune and Tim assume the question revolves around: "Do current academic models well answer certain questions?" I would say OP's post raises a more fundamental question: "Are academics attempting to answer the _right_/_useful_ questions?"
Reply by ●December 1, 20082008-12-01
On Dec 1, 10:11�pm, Rune Allnor <all...@tele.ntnu.no> wrote:> On 1 Des, 11:33, "euler.shan...@gmail.com" <euler.shan...@gmail.com> > wrote: > > > Dear all, > > > i heard several times.. some of our current researches in wireless > > comm doesnt not reflect the real world scenario. but... i asked some > > professionals.. most comment on channel state assumptions such as > > Rayleigh and its counterparts... not others.... > > Don't know much about communications, but any real-world > application of maths, physics or statistics can be said > to 'not reflect reality.' > > The reason is that one can only do computations and > calculations with the simplest, idealized models. There > will always be some factors left out of the models, and > those that are included are idealized. > > One doesn't use the linear model because it is 'true', > but because it is possible to do computations with it. > One doesn't use Gaussian models because they are 'true' > but because they are described by merely two moments, > and thus are tractable. > > So whenever we can, we push the simple idealized linear > models as far as possible (and all too often a bit and > a half further), and don't include nonlinear terms > or non-Gaussian statistics unless we really have to. > > So physical or statistical modeling is not about 'truth' > or 'reality', but about finding a model that is useful > and yet tractable. > > RuneThanks a lot Mr Rune and Tim Wescott, as you mentioned above, i understand that we need to devise a simple model ( not simpler ) enough to track the approximate behavior of the system without much math or computational complexity. then we need to add some more real world reflected parameters to make our prototype more concrete. may be.. if i am not wrong... thats the way... Shannon and all other engineers or applied scientists attack the problems... thanks for ur nonlinear and Gaussian information. i think that's kind of problems can be found in Power Amplifier and Channel Assumption Problem. with wideband applications' popular, nonlinear and memory features of Power Amplifier becomes more apparent and we need more sophisticated linearization and memory compensation techniques ( including PA's nonlinear and memory characterization techniques ). for Guassian Channel Assumption, we have this Rayleigh complex Gaussian Assumption which have been already readdressed to other more realistic model such as Nakagami model... let me know if i miss out some in my propositions..
Reply by ●December 1, 20082008-12-01
On Dec 2, 5:12�am, Richard Owlett <rowl...@atlascomm.net> wrote:> Tim Wescott wrote: > > Rune Allnor wrote: > > >> On 1 Des, 11:33, "euler.shan...@gmail.com" <euler.shan...@gmail.com> > >> wrote: > > >>> Dear all, > > >>> i heard several times.. some of our current researches in wireless > >>> comm doesnt not reflect the real world scenario. but... i asked some > >>> professionals.. most comment on channel state assumptions such as > >>> Rayleigh and its counterparts... not others.... > > >> Don't know much about communications, but any real-world > >> application of maths, physics or statistics can be said > >> to 'not reflect reality.' > > >> The reason is that one can only do computations and > >> calculations with the simplest, idealized models. There > >> will always be some factors left out of the models, and > >> those that are included are idealized. > > >> [SNIP *MAJOR* portion of Rune's and all of Tim's comments] > > I do not know if I am missing the OP's intended point > [but I *DO* think that his post could refer to a *VALUABLE* point ;] > > Rune and Tim assume the question revolves around: > � "Do current academic models well answer certain questions?" > > I would say OP's post raises a more fundamental question: > � "Are academics attempting to answer the _right_/_useful_ questions?"Dear Mr Richard Owlett, thanks a lot for your clarification, i do mention like you do! actually i am always wondering what changes that academic researches should follow ......to reflect the real world development?
Reply by ●December 1, 20082008-12-01
euler.shannon@gmail.com wrote: ...> Dear Mr Richard Owlett, > > thanks a lot for your clarification, i do mention like you do! > actually i am always wondering what changes that academic researches > should follow ......to reflect the real world development?I don't mean to imply that there are none, but this discussion coulld get to the point more readily if you showed us an example where you think that academia and practice are out of step. A hint about what might bring them into closer alignment would also help. Jerry -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
Reply by ●December 1, 20082008-12-01
>euler.shannon@gmail.com wrote: > > ... > >> Dear Mr Richard Owlett, >> >> thanks a lot for your clarification, i do mention like you do! >> actually i am always wondering what changes that academic researches >> should follow ......to reflect the real world development? > >I don't mean to imply that there are none, but this discussion coulld >get to the point more readily if you showed us an example where you >think that academia and practice are out of step. A hint about what >might bring them into closer alignment would also help.OK, lets inject an example.... If you ask the question "What can I expect from ZigBee in a real world situation, like an electricity meter in a cupboard talking with appliances around the house" the answer will be "Dunno. Suck it and see.". Whilst this kind of setup is obvious highly variable, and clearly tough to analyse, academic work currently provides no help at all in trying to get a grip on the problem. This total inability to have a clue how a radio network might behave in any complex environment is probably the biggest headache people have today in planning systems. Regards, Steve
Reply by ●December 1, 20082008-12-01
On Mon, 01 Dec 2008 21:08:51 -0600, steveu wrote:>>euler.shannon@gmail.com wrote: >> >> ... >> >>> Dear Mr Richard Owlett, >>> >>> thanks a lot for your clarification, i do mention like you do! >>> actually i am always wondering what changes that academic researches >>> should follow ......to reflect the real world development? >> >>I don't mean to imply that there are none, but this discussion coulld >>get to the point more readily if you showed us an example where you >>think that academia and practice are out of step. A hint about what >>might bring them into closer alignment would also help. > > OK, lets inject an example.... > > If you ask the question "What can I expect from ZigBee in a real world > situation, like an electricity meter in a cupboard talking with > appliances around the house" the answer will be "Dunno. Suck it and > see.". Whilst this kind of setup is obvious highly variable, and clearly > tough to analyse, academic work currently provides no help at all in > trying to get a grip on the problem. This total inability to have a clue > how a radio network might behave in any complex environment is probably > the biggest headache people have today in planning systems. >How do you know that's what the answer will be? I can't imagine _all_ academics being that unhelpful -- particularly considering that many profs ping-pong between academia and industry. The best answer you'd probably ever get would be "collect a ton of data, crank it through this algorithm, and see". There's got to be at least one thesis in there, if not an instrumentation-and-analysis startup to boot. -- Tim Wescott Control systems and communications consulting http://www.wescottdesign.com Need to learn how to apply control theory in your embedded system? "Applied Control Theory for Embedded Systems" by Tim Wescott Elsevier/Newnes, http://www.wescottdesign.com/actfes/actfes.html
Reply by ●December 2, 20082008-12-02
On 2 Des, 04:08, "steveu" <ste...@coppice.org> wrote:> academic work currently provides no help at all in trying to get a grip on > the problem. This total inability to have a clue how a radio network might > behave in any complex environment is probably the biggest headache people > have today in planning systems.It's depressing. I've met people who had decades of 'experience' with underwater acoustics but got very, very quiet when they saw images like this one: http://www.shipsandoil.com/PicoftheDay/Picof%20the%20day%20images/END%20heavy%20weatherS.JPG I have a set of similar images taken on one of my trips, where one sees a companion vessel to ours riding out a storm. Whenever I give a presentation, I have that sequence of images run as a slideshow on the 'white-screen' (what's the proper name of the screen where slideshows are projected?) while the audience enter the room and find their places. The brighter among the audience are actually able to figure out that since the images show a companion vessel to ours, *we* suffered from the same conditions at the time... Getting the audience's attention is usually not a problem after that. Rune
