I'd like to clarify and respond to some concerns expressed about the software side of the GNU Radio/USRP SDR combination.
First, we remain committed to maintaining Free Software Foundation copyrighted, GPLv3 licensed host code that provides the C++, Python and GNU Radio Companion interfaces to the USRP1 and USRP2 hardware within the GNU Radio framework. Today, this comes in two forms. The first is the two C++ "libusrp" and "libusrp2" shared libraries that provide low-level configuration and sample based access to the USRP hardware over their respective hardware interfaces. This is the kind of framework-independent driver code that is typically provided by hardware vendors and allows its use with a variety of software packages. Analogous drivers are the libasound driver for audio hardware or libhamlib for communication with various radios. Built on top of libusrp and libusrp2 are the GNU Radio gr-usrp and gr-usrp2 signal processing source and sink blocks. These provide the streaming signal flowgraph wrappers around libusrp(2) for C++ and Python applications, and the further metadata needed for use in the graphical design tool GRC. All the above are part of GNU Radio, are copyrighted by the FSF, are distributed in our software tarballs, source code repository, and binary packages, and are licensed for use and redistribution using the GPLv3. This code, with some further integration of the some in-progress development, will become part of our 3.3.x stable series, and will be maintained with bug fixes and/or updates should they become necessary. Ettus Research has announced their intent to release a "Universal Hardware Driver", or UHD, that will be a single driver approach to current and future hardware from them, and has also stated the UHD will be available under both GPL and non-GPL licenses. The UHD will perform essentially the same functions that libusrp and libusrp2 do now, and we will be able to have GNU Radio "gr-uhd" source and sink blocks to allow use within GNU Radio applications. This is a good thing for GNU Radio users for a variety of reasons: * It has proven tedious to develop GNU Radio applications that can transparently use either the USRP1 or USRP2. The UHD will abstract these differences so applications can adapt to the available hardware capabilities that are found. Currently we have some inelegant code to accomplish this but it's not an effective solution. * Future products from Ettus Research will get GNU Radio software support much more rapidly than in the past, as most of the differences will have already been addressed in the UHD. * The UHD makes it easier to expand support for the USRP into other, non-GNU Radio, frameworks or development environments. We see this as a good thing--it both promotes the use of low-cost SDR technology in general and provides an easier transition between those environments and GNU Radio. We have already been collaborating with Ettus Research for some time on the overall design of the UHD to ensure continuity of features and capabilities between the current GNU Radio libusrp(2)/gr-usrp(2) modules and the future Ettus Research UHD/GNU Radio gr-uhd modules supporting their hardware. While this intentionally won't extend to source code compatibility, we do think it will be straightforward to port applications. In addition, we will be coordinating distribution, installation, and maintenance of the software to keep things synchronized between the projects. The acquisition of Ettus Research by National Instruments is clearly recognition that the use of low-cost SDR hardware and open source SDR software has gained acceptance and use across a broad spectrum of academic, corporate, and government users. While it hasn't been explored in detail, there is also an opportunity to bring GNU Radio support to other NI products. It's an exciting time. Johnathan Corgan Corgan Enterprises LLC _______________________________________________ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org http://lists.gnu.org/mailman/listinfo/discuss-gnuradio
