Wireless Networks In-the-Loop

Video thumbnail (Frame 0) Video thumbnail (Frame 5276) Video thumbnail (Frame 16520) Video thumbnail (Frame 17772) Video thumbnail (Frame 18498) Video thumbnail (Frame 19205)
Video in TIB AV-Portal: Wireless Networks In-the-Loop

Formal Metadata

Wireless Networks In-the-Loop
gr-winelo - A GNU Radio Network Emulator
Title of Series
CC Attribution 2.0 Belgium:
You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor.
Release Date

Content Metadata

Subject Area
This talk introduces gr-winelo, an in-the-loop simulation framework for communication networks which are based on the GNU Radio software radio toolkit. gr-winelo mimics the behavior of common RF frontends such as the USRP, but instead of sending the signal over the air, a central server plays the role of the wireless communication channel. Arbitrary channel models can be simulated, by passing their respective GNU Radio processing block to the server. Since this whole setup is completely transparent to GNU Radio applications, it is at any moment possible to switch between simulations and real-world tests. For the Demo a frequency hopping network will be simulated and analyzed using gr-winelo.
Complex (psychology) Game controller Code Multiplication sign Real number Execution unit Public domain Student's t-test Scalability Element (mathematics) Computer configuration Internetworking Computer hardware Software testing Software development kit Form (programming) Simulation Theory of relativity Software developer Signed measure Debugger Special unitary group Cartesian coordinate system Measurement Software Personal digital assistant Blog Order (biology) Video game Iteration Wireless LAN Freezing
Group action Code Ferry Corsten Multiplication sign Execution unit Numbering scheme Real-time operating system Mereology Disk read-and-write head Mathematics Different (Kate Ryan album) Computer configuration Oval Cuboid Extension (kinesiology) Physical system Simulation Theory of relativity Block (periodic table) Moment (mathematics) Gradient Sampling (statistics) Computer simulation Special unitary group Instance (computer science) Lattice (order) Benchmark Measurement Flow separation Cognition output Writing Resultant Laptop Point (geometry) Slide rule Implementation Functional (mathematics) Open source Ultraviolet photoelectron spectroscopy Student's t-test Streaming media Event horizon Rule of inference Number Revision control Frequency Computer hardware Software testing Data structure Task (computing) Noise (electronics) Standard deviation Information Interface (computing) Weight Counting Incidence algebra Cartesian coordinate system System call Symbol table Transmitter Subject indexing Computer animation Integrated development environment Mixed reality Blog Video game Absolute time and space
Simulation Computer animation Inheritance (object-oriented programming) Cellular automaton Execution unit Cartesian coordinate system Bilinear transform Address space
Theory of relativity Computer animation Maxima and minima Cartesian coordinate system
Frequency Lecture/Conference Cartesian coordinate system
Group action Building Confidence interval Multiplication sign Equaliser (mathematics) Execution unit Water vapor Mereology Food energy Fraction (mathematics) Impulse response Medical imaging Inference Mathematics Component-based software engineering Bit rate Different (Kate Ryan album) Velocity Core dump Software framework Position operator Physical system Social class Thumbnail Theory of relativity Block (periodic table) Sampling (statistics) Computer simulation Bit Sequence Measurement Oscillation Connected space Category of being Message passing Process (computing) Vector space Telecommunication Order (biology) Phase transition Computer science Summierbarkeit Quicksort Ranking Arithmetic progression Reading (process) Resultant Point (geometry) Dataflow Functional (mathematics) Markup language Information and communications technology Virtual machine Event horizon Rule of inference Scattering Number Frequency Doppler-Effekt Cross-correlation Propagator Root Energy level Data structure Analytic continuation Form (programming) Shift operator Graph (mathematics) Prisoner's dilemma Cellular automaton Projective plane Line (geometry) Cartesian coordinate system System call Symbol table Transmitter Uniform resource locator Word Radius Integrated development environment Personal digital assistant Charge carrier Hydraulic motor Transmissionskoeffizient Object (grammar) Pressure Wireless LAN Marginal distribution Window
it is shown that the KIT and it's we were presented in wireless this is the new so that's basically the idea comes from mountain fall from a time and he was my advisor to the left and so we implemented there is research and and Gerald among was also a student of the Mount was my wasn't masterpieces so again that's now some of of and the 1st about the idea of wireless and so as soon as simplifies radio new element I think you all know it is it's not that's not true for clarity so there come new problems in and so on then and by the great thing about it is that we can apply all workflow between all forms of development of the the workflow radio development is all the debugging tools automatic test and tested in development and the idea of the question to answer is why just use enforcing a blog or a signal we want we could use voice an were sold to simulate internet of all nodes and the channel in between them this would be great if we could allow as seamless switching so that we can work with the same corpus for our real application which you run on the use of peace order and as hardware as well as follows simulation so that we could improve in every single iteration very can switch when we switch from testing all real-world measurements assimilation so the main idea is to say the softer domain as long as possible and to go back into the softer demand as soon as possible if you have to do real measurements the the benefits of this approach are for sure flexibility and scalability if you want to test and really big networks life 20 or 30 nodes and you want to learn get a feeling about how the network will behave in this case it's far from those less or um for most people to be of the possible in in the rear of the a half yeah benefit is that we have only 1 called so you don't need a Matlab simulation code are you don't need to run and his freezing relations right before you could it's depending on how complex unit for review but you don't need to Sun great benefits I think the greatest benefit of its cities debugging so the possibility to set breakpoints on the yelling is really nice if you if you develop complex verbs and also the control of the realism so you can turn on and off impairments so everyone knows that it's uh easy to to run a simulation on the PC revolves something novel or real impairments or something like that and here you have the option to come 2 and on a half impairments but not sold determine impairments and this all leads to a reduced development time and increase his who character so I come to the basic
principles and implementation of those we've chosen time so structure so that you could come you could run really computationally expensive tasks like the channels relation on a call for the long and you also could around is there repeat cease as a single nodes and connect them to those of us who would do with the firm opposite of so yeah 5 laptops each of you is a P a group and the a channel is then simulated by also and standard story over TCP-IP so here are some points that you that we had to implement so 1st of all is still pending rejected from and then all over the and the timing is a really problem and want to simulate a tensile Decius question this what do you do work uh and if the node stops sending if you applications ups and what would you during assimilation because if 1 mole of substance meeting all of the other nodes still get noise scenario in real time and in a real environment so we have to simulated and the big problem about this is that you have to get the exactly correct amount of samples which are my samples so we had to implement and something like time commands all support for technical was that you would use use appeared true to enable it to transmit that's several times or if you have a the hopping scheme yeah we also support for an absolute time if over it's the model after 2 kiss you all to support for GPS time Don Quixote had a hotter and channel models so all we have different models and all of them are modular design so you can easily and once and last but not least this simulation interface because to allows you seamless switch to the uh you don't want to write different code for the simulation because we want use the same code as we which we would run in our relation to allow the switching between the measurement and simulation yeah so let's come to zero-padding detail when you run slotted so assurance that applications with the use of the the 1 possibility is to use the start of the 1st times mn and of 1st and feature is called stream text and radio far it uses the index from where you and the idea is that you could count the number of zeros you have 2 incidents because you know all the start of of standard times and then you know how much they have to go through simulation then you have end of 1st and if the next time so you can calculate this this difference and you know this a all of symbols and then you can it is when the problem is that you know this was not at this moment so you don't know how much they would have to exit so we implemented a method that's starts and reducing noise policy rules or anything and distributes them to all the receivers and until it gets a times and then it calculates how much its head produce already and how much a test will in the future and if it produces too much it gives you a bonding and instance stop his simulation era so then you have to run the simulation of its so we have implemented the free free versions it's same accurate it's a simple accuracy of and which I explained right now and generic inaccuracy repeating what the sources it simply weights and if there's nothing on the input of the function itself and it produces reduces noise and there's also the initial zero-padding which is important if you have transceivers systems which holes all put depends on the input so day so is the chicken egg problem at the beginning when there's no input the the chance you approach result that year and so therefore we implemented the initial zero-padding reproduces noise at the start if if the function didn't don't get an input so we follow the all and modular designed to allow you to move to easy general models hardware models while I changed everything in between them is and also this simulation the face of change right now we're using TCP UDP Sing radio because there there are known and they're working good and but you really easily could extension so In the past years to for the exchange of certain but still student performance so we just use it was the nominal and yeah as you can see here this which use reuse so the blocks of the same options all because I could be president from the huge books and have some of those Malaysian Palm through so it is support for existing applications than my muscles pieces I I wanted to show the performance of in the sun PRN and Diodorus and therefore I use the standard achieves benchmark at which are part of that yardage to package and it worked out of the box size of change tied to broader use the blog and added the winner all sorts of and never got out of the box so and I was and task is to write status information that used to this slide which frequency of knowledge which the rate of and also give support for the application to change the frequency of things therefore I've created infrastructure hoping it talk application which here is the prior which describe the grade and uh possible possibilities for fair use when you want to simulate because unless you model this that the mixing between the frequencies and do you all a talk is really we um the really depends on our that's the very well if you if you fail life yeah failures in your all of a New Simulation all your genome on so as you can see here the differences right now are you don't see the see also which would be very easy to implement YDR it's the characteristic which is implemented already and suggests that a big event at 314 units it's also an opportunity to push in the froze into the whole simulation so to test you network if in the presence of an intifada of perhaps in the presence of secondary user primary user anything about cognitive radios so I want to show you a day
1 on so what you can see here is
the year frequency warping and ad-hoc applications old at the left there's 1 or 2 features them addresses to serve dynamically and yeah also the 2nd also they have to address this 1 has a this 1 and this 1 of those women I yeah I'm able to transmit ligatures from 1 took to be like this of from 2 to 1 like this parents this is a real application which also runs on the use of you so I've it and what I found he had a side of taking all the units deployed and inserted you with overlook so and we
can have a look at that so it's not a feature we can easily and connects new applications for the simulation of so now I connect us huge so the cell is running you on the left you can see that it makes some new connections and
then we could maximizes so now the old flowgraph stops and it runs because we need the initial
zero-padding here on foot all applications in relation to stop so you can see it being
package and so so I could
transmit some packages so that you can see it's really does application running the so
yeah packages so and it's 1 possible
workflows which helped me a lot when I and
developed is frequency hopping it took more because there's so many uh possibilities to make failures and it's easy to to debug biographer
framework like this that so and no 1 guaranteed to continue on going to talk a little bit about China Morrison because so far we talk about it was set up of the system is working with this over the somewhat of a and a denotes the connectedness of an exchange accuse samples of the complete disregard the father models OK and I think because most of you are computer science guys so the idea of and are often quick introduction and what the problem is if for an ingenious with thus wireless communication when it's trying to transmit data audio or hidden so the problem is that ah wireless channel of which are using to transmit data behaves very differently differently and that depends on various factors of positive depends on frequency band so that you and transmitting had 100 megahertz are at 2 . 4 yards it's a lot of that makes the difference of all sample of your although also endorsed but that's was a huge difference you know how long does the for of along the thing for a signal to get from the transmitter to receiver just 1 example it also depends on your movement speed so if you're talking on your cell phone if you're moving around and on foot on deriving in the continuity you also have to take that into account other things and also you can for example the data rate of the so was that the transmitting shot and you therefore you can't into the role of data on the wrong it also makes a difference I'm getting to know what I expected more details later and it also depends of course on the position of objects and well surrounded by a lot of buildings or if you have just in the free space somewhere out that was of a and so here is a reasonable scenario we've got the base station of call cellphone and then we've got of course the line of sight path directly but we also there multiple parts which are we start the levels of of was buildings of order and all of those parts they have different units because the other from the long and then all of the moles of a different phase so well they can either up constructively all you have to have different depending on the face and it changes the signal level but that's the see and moving a reasonable and also if you walk along for exhaustive so then you change from also some from interference and the value of probably all of you in high school and you know that that was that the experiments so there of this maximum mark on your wall and well this is same thing the use of 1 of them it is of a from if you are a lot of cell phones and moving also affects love the perceived carrier frequency of your of your receiver you also some of which you have to correct that so that's been just after class that you have to deal with the open and all that set of of that line for Ulysses of some of you have something like this these these clubs in between and that's just what and some of the depends of course if can move around and world In this talk of time and and the idea all that's money and this is a nice little example any in that the signal assigned Signal nails and interference and if the injury interference decimate make then you can't be anything is very much a same just over long you will be useless if not at the level of crops and you receive in any as you do not receive anything just watch and then well close the unveiling wanted to cancer movement speed so how fast and does the change for example all also sorts position from only but the genes with all the receiver so on problems Sweden and or large trucks in front of you dying of causal models monoscopic maybe on your signal of people's and it has been that of course also if you make this John model you want to take your environment in which you're communicating and take effect you're OK good so 1 thing about shows that despite the channel impulse response while it's just 1 way to describe them and flowers simple imaginable transmit and the inference within a short paths and receive all you bang on harmony parts we have you guys problem may something like this 1 so this is all flows thought that receive and then the but later the the up OK and so on and this is what magic or the channel impulse response and you can you have to keep that in mind if we want to design a system talking been so open just as an example John impulses responses all important but in both cases we have the same samples impulse-response these short passes over here and I'm measuring transmitting a very low rate Journal of URL data right so in this example would be I margins that that's signal signals among the it's a just a 1 bit and of course you receive on all past but we can see that all about because they aren't really salt and along in time now and retransmitting and with little with a higher data rates so all pass this of course some of some of social and and foster some of us about on the agreement 1 and the men enter 1 preconceived group by not transmitting but them again to everyone we also get interference on the other parts of the symbols we receive these just 50 who want every vector that are so and this phenomenon is the sum interference and that's the reason why we actually need an equalizer stenosis images of Martin mentioned so as in the what this is the same job but just because we're using a different system growth at that time and after that because this is like a time was so important and that's not what we wanted to do as a model not just major models where it should be no analytical dreaming some lectures about the environment it will expecting but it can also signed the channel which means measuring the channel impulse response which is something across describes a is general something and uses the Israel you transmit lack of Newton's windowful Poseidon will detect all the echoes the receiving and then you would much that shown in and problem as you just using very shortly also done nothing so that's not very efficient because but we also know the messages that were know you it's small to transmit all the time so the idea is that the correlation channel summer and it transmits the sequence repeatedly and then the sequence the number a nice properties so if you use the core of the correlation function on them and their got this large peak when are perfectly aligned but if this shift to the little bit know close to 0 the so and this is from Grand Prix renowned transmits all 6 of this sequence at the receiver and a transmitter then we can use this correlation to the that the the pieces on the friends do mantra and then we're about became the different kinds of at the receiver OK so when I can just flew from the 5th just so that we can make some measurements uh more young so we place the transmitter and thanks to the bombings regards from multiple yeah know that multipath environments we get different but you can actually take different paths and just move is of hundred so here this was all set up so we have the transmitter you receive on just for the month of thoughts and 1 Egyptian processing and where your and knows that energy versus the velocity of this that you need to measure the Doppler shift of the markup of components you have to speed up a little so this so now for example of the results that you get different moving along this is actually all we see all this again z is a structure it's not quite a lot of and here we lots of the body it profile just them the kind of the day time marriage of the channel impulse response and rank again 0 here with the line of sight peaks at the beginning and end up being bills on the also we see from the buildings there and then all and only got nothing that's just must movie and then from these Eastern measurements then we can also make a channel model I think I skipped the Dimona due to time when all the question is well and what kind of models can be used in radio so I'm going to make some very simple and this and with the models like and WHEN some general just my already fading channel all online we also implemented costs on him some general models that should I use for the same GSM but and the design of them most of its move you can understand what this busy running acuity flow from which simulates the channel and so you can block in pretty much any new idea progress the channel model and as of the point 7 there's a project called G channels from 2 machine and spoken he implemented a couple of genome was that you can also use the 10 channels on our part of the the 1 7 yet here I think of region of the world it is the Audi and the what is a kind of experimental feature film also original models from on the measurements so if you don't want to think too hard about the environment and to work on and take a model on just as a general rule of thumb the measurement and then take from that of the regional level this is the very basic and then think that if the last 2 minutes for a what the outlook on to encourage a lot of them deferral of new features that may exist available prison and we want to reason that I had in 2014 the public's and you can get it and well I then goes from the wrong here on my communications we want also onsets and general models for foreign communications and is also working on this graph you and interface for managing the whole application so am yeah this is um what what we are looking for new looking on another got time for pressure based on if on can albeit not is on on what he it's um of these models uh um you most them on is the I know know I've known right them H I mentioned that I think this the the main events of something on and then abandoned enough because what we have to take into account the a I understand troposphere and that's school and no cell of AWGN just that's nice and radius may just basically when you've got there are completely surrounded by scatterers which is not the case in major so we tried to provide a framework to build up a frame of so there's really simple models in in their octanol water uh as a proposal and it's the same or similar ideas so I think we have to push it as soon as possible to work together on the whole thing and not to have 5 different the readers in relation to what channels on into and that the use of the word the vision of In all of the root of the sum of m this is the fraction off those studio also 1st that you had you don't have any of but frequency of so that the channel sample you know you mean but find if you assimilating the same aspects in our half a model it is known that it's related we have also the comfort of the the people who are part of the reason is that of the yeah um you know when you nothing about its that depends on the optimal you can pretty much like in any other model but since when the infancy of the project that we just have a little bit of most that there is no radio channel of the people here here in Canada so if you are a reading of 2nd and if the economy is probably also if you put this on the bottom line is really know the fact that they could be produced by the end of the day and then you have your own words fail models that you add world and I would like to warn the timing of the words that you know and we also have a higher with the use of the way of writing for the U and the use of I because what I have PPLR perhaps they'll legislators and this was something I actually really writing a journal of assigning the channel model and 1 of the the the the the genome posters forms is changing what can happen for the Doppler frequency but can also say well that's happening as we saw that the the oscillations of the course that's reason why on the use use GPS deals and wanted to measure the Doppler shift but you have to know what you have to shift based on the block block of because all the labels of of of of the year when the yet to be here it was really that kind of thing there's a lot of confidence in you yes if the you would do it so in the coming of the year that you you all of the all the so I think you are the constant in some John once again basically that L for um propagation environments I think this 1 foot up the overlying than that of and he majoring in rural areas so you have those falls in general motors and then depending on summation you want 1 then you can block in 1 of these I the yeah exactly working then Mexican that can together so after the situation so to to get some ideas from you will develop a look at what 1 could lead all of equal to implement statistical the things that you want it and