When will the stable VEINS ver 4 will be available tentatively and what will be its features? Will it have new features like vehicle density, security models and others?
Thank you.
A stable version of Veins 4 is available for download.
Aside from bug fixes, it offers enhanced visualization options, enhanced phy layer models, improved extensibility, and a screenshot/screencast recorder.
Note that, as the TraCI API was changed to report all angles in a different format, Veins 4.3 requires SUMO version 0.25.0.
In contrast, keeping the number of vehicles in the simulation constant (that is, using a fixed vehicle density) has been included since Veins 3 alpha 2 (see the comment by user4786271). If this is the only feature you require, there is no need to upgrade.
Regarding security models, several publications are using Veins for security and privacy research, for example for Privacy assessment in vehicular networks using simulation. However, due to the vast number of different approaches and lack of standard, no "default" security-centric models are likely to be included in Veins anytime soon.
Related
I would like to use Veins with functionality that was only recently added to SUMO and that is still being further developed. The work that I have in mind is a PhD project that has recently started, and I suppose the simulation part will not start before 2021. Given that Veins currently uses SUMO version 1.2.0, and that, IIRC, the TraCI protocol has been updated after that, how difficult would it be for us to update the system so that Veins works with SUMO 1.4.x or even better with the future 1.5.x? Which skills would be needed?
The current Veins master works with Sumo 1.3.1 already. Afaik, neither version 1.4.0 nor the current progress on Sumo's master branch have been tested, but looking at the changelogs, it seems that there were no changes that look like they should break the current implementation, as mostly new functions have been added that are not yet implemented in Veins. I'd suggest simply running the Veins example against a Sumo-master, if this works without issues that's a good indicator that also more sophisticated simulations will work.
How do I set up a backend for Deepwater (h2o) on Ubuntu 16.04? The GPUs I am using are AMD Radeon RX Vega. Anyone experienced in this topic? Do you need further information? Most explanations and procedures described here and elsewhere seem to refer to NVIDIA cards.
Deep Water is a legacy project (as of December 2017), which means that it is no longer under active development. The H2O.ai team has no current plans to add new features, however, contributions from the community (in the form of pull requests) are welcome.
Having said that, Deepwater was never built for AMD because we used Tensorflow and MXNet as backends and they did not support AMD.
I want to use different maps (for example map of an urban area like in manhattan mobility model) in my simulations. Is it possible to do it in my omnetpp.ini file as some initialization?
I use the mobility module in inet framework.
Thanks.
OMNeT++ 4.6 does not contain any support for showing maps except displaying a background image in the module, but that must be carefully aligned to match your mobility model.
OMNeT++ 5.0 will support osgEarth to display 3D models + display maps. You may experiment with the 5.0 Technical Preview (available in the test folder) Download. Note that this is Linux only for the moment, and you still have to implement your visualization if you want to use it in INET. Check the osg-earth example in the samples folder.
I'm writing a Direct3D application, using DirectX 9. Although it works on my PC, I need to make it work on a wide range of systems. I need to know what capabilities I can expect to see on other systems. Is there a list of the DirectX capabilities that graphics cards support?
I've found one site, which I'll post as an answer, but it's a bit out of date.
Edit #1: Of course I will test for all the capabilities before I use them. But there are two different approaches to a missing capability: either workaround it or just fail to start. I need to know how many people will be affected before I decide which approach to take.
Edit #2: By "capabilities" I mean the values in D3DCAPS9 returned by IDirect3D9::GetDeviceCaps. These can (and do) differ amongst different graphics cards.
There's an excel spreadsheet of capabilities of all major graphics cards in DXSDK. Find it in the SDK's Samples\C++\Direct3D\ConfigSystem\CardCaps.xls. It also gives the lowest common denominator that all cards can support.
The newer SDKs come with the spreadsheet of newer cards, while the older versions of DXSDK come with the spreadsheet of older cards. It seems MS is keeping updating it. You can choose based on your needs.
Chris Dragon's Direct3D and OpenGL Device Capabilities Database is extremely well-done, but a bit out of date.
DirectX 9 was last updated (DirectX 9.0C) in 2004, and that was a fairly minor update. The original release was in 2002.
IMO, it's not worth even trying to work around missing DirectX 9 capabilities -- anybody whose card isn't capable of DirectX 9 is long-since accustomed to only ancient games working at all.
Which GUIs would be used in vehicles for radios, gps devices, infotainment systems?
They use proprietary ones.
Here is an article about a system based on Java, which is already a few years old. It was/is used in high-end systems of BMW and VW.
In my company, as far as I know, we have used four types of GUI framework (flash/Qt/HTML5/Internal proprietary HMI framework) for projects.
HTML5 is adopted for Linux based projects which are GENIVI compliance. http://investor.harman.com/releasedetail.cfm?releaseid=817096