| 41 | | Software Defined Radios (SDR) offer several advantages over traditional, hardware-based radios, most notably flexibility and reconfigurability. Developing SDR applications can be a difficult process however, for several reasons. First, much of the work involved deals with standards compliance, rather than radio functionality. This portion of the work is very detailed and error-prone, leading to wasted time and effort. Second, there is little to no support available for the debugging and refinement portions of the development cycle. There is no easy way to monitor or control SDR applications at runtime. The Waveform Workshop was created to address these issues. A part of the OSSIE SDR project, the Waveform Workshop drastically reduces the time and effort involved in SDR development. It automatically generates the portions of the software related to standards compliance, letting developers concentrate on radio functionality. In addition, the Waveform Workshop provides tools for both monitoring and controlling SDR applications at runtime, making debugging much easier. |
| 42 | | \keywords{Software Defined Radio, Rapid Prototyping Tools, Education |
| | 41 | Software Defined Radios (SDR) offer several advantages over traditional, hardware-based radios, most notably flexibility and reconfigurability. Developing SDR applications can be a difficult process however, for several reasons. First, much of the work involved deals with standards compliance, rather than radio functionality. This portion of the work is very detailed and error-prone, leading to wasted time and effort. Second, there is little to no support available for the debugging and refinement portions of the development cycle. There is no easy way to monitor or control SDR applications at runtime. The Waveform Workshop was created to address these issues. A part of the OSSIE SDR project, the Waveform Workshop drastically reduces the time and effort involved in SDR development. It automatically generates the portions of the software related to standards compliance, letting developers concentrate on radio functionality. In addition, the Waveform Workshop provides tools for both monitoring and controlling SDR applications at runtime, making debugging much easier. |
| | 42 | \keywords{Software Defined Radio, Rapid Prototyping Tools, Education, OSSIE} |
| 47 | | Software Defined Radio involves software implementation of radio functionality that is traditionally performed by analog hardware and digital application specific integrated circuits (ASICs). SDR employs general purpose processors (GPPs), digital signal processors (DSPs), and field programmable gate arrays (FPGAs) that can be programmed or configured to provide great flexibility and support for multiple radio communications standards or waveforms. Within limits of its hardware, a well designed SDR can support standards that were not conceived at the time the radio was designed \cite{Reed2002}. SDR is of interest for both commercial \cite{Vanu} and military \cite{JTRS} applications. |
| 48 | | |
| 49 | | Unfortunately, developers writing SDR software frequently face problems that are not addressed by traditional development tools. This first problem is ensuring that SDR software conforms to the SCA standard. The standard is very complex, with many details that are necessary for compliance but unrelated to the functionality of the radio itself. Another common problem SDR developers face is the lack of any way to interact with running SDR applications. SDR applications are deployed as collections of CORBA software components running on one or more processors. Software component configurations are recorded in XML files. In this situation, the primary way to change a property value for a given component is to shut the application down entirely, change a value in a file, then restart the application. Finally, developers also have no readily available tools to monitor a running SDR application as it operates. |
| 50 | | |
| 51 | | To address these problems, we have developed the Waveform Workshop a set of software tools. The Waveform Workshop was designed to work with the OSSIE open source SDR project \cite{Aguayo2009b}, \cite{ossieweb}. OSSIE is a partial implementation of the U.S. Defense Department's Software Communications Architecture (SCA) \cite{sca}, \cite{bard}, \cite{Aguayo2009a}. The Waveform Workshop aids developers by automatically generating the parts of the software that relate to SCA compliance. The developer can then concentrate on implementing the radio itself. The Waveform Workshop also provides developers with a means to update application properties at runtime directly, without editing configuration files or restarting the application. This greatly reduces the time needed to debug applications. Finally, the Waveform Workshop also provides an interface displaying block diagrams of running SDR applications. It also allows developers to take the output from an application and play it through speakers or plot it on a graph. This allows the developer to easily change application properties and observe the effects of those changes in real time, making it much easier to debug and refine SDR software. |
| | 47 | Software Defined Radios involve software implementation of radio functionality that is traditionally performed by analog hardware and digital application specific integrated circuits (ASICs). An SDR employs general purpose processors (GPPs), digital signal processors (DSPs), and field programmable gate arrays (FPGAs) that can be programmed or configured to provide great flexibility and support for multiple radio communications standards or waveforms. Within the limits of its hardware, a well-designed SDR can support standards that were not conceived at the time the radio was designed \cite{Reed2002}. SDRs are of interest for both commercial \cite{Vanu} and military \cite{JTRS} applications. |
| | 48 | |
| | 49 | Unfortunately, developers writing SDR software frequently face problems that are not addressed by traditional development tools. This first problem is ensuring that SDR software conforms to the Software Communications Architecture (SCA) standard. The standard is very complex, with many details that are necessary for compliance but unrelated to the functionality of the radio itself. Another common problem SDR developers face is the lack of any way to interact with running SDR applications. SDR applications are deployed as collections of CORBA software components running on one or more processors. Software component configurations are recorded in XML files loaded by the CORBA components on start up. In this situation, the primary way to change a property value for a given component is to shut the application down entirely, change a value in one of the XML files, then restart the application. Finally, developers also have no readily available tools to monitor a running SDR application as it operates. |
| | 50 | |
| | 51 | To address these problems, we have developed a set of software tools that constitute the Waveform Workshop. The Waveform Workshop was designed to work with the OSSIE open source SDR project, a partial implementation of the U.S. Department of Defense's Software Communications Architecture (SCA) \cite{sca}, \cite{bard}, \cite{Aguayo2009a}, a standard architectural specification for SDRs. The Waveform Workshop aids developers by automatically generating the parts of the software that relate to SCA compliance. The developer can then concentrate on implementing the radio itself. The Waveform Workshop also provides developers with a means to update application properties at runtime directly, without editing configuration files or restarting the application. This greatly reduces the time needed to debug applications. Finally, the Waveform Workshop provides an interface displaying block diagrams of running SDR applications. This interface allows developers to take the output from an application, or any component within an SDR, and play it through speakers or plot it on a graph. This allows the developer to easily change application properties and observe the effects of those changes in real time, making it much easier to debug and refine SDR software. |
| 54 | | Section~\ref{waveformworkshoptools} introduces the tools that are part of the Waveform Workshop and the problems they solve. Sections~\ref{sdrapplicationdevelopment-oef}--\ref{sdrapplicationdevelopment-alf} describe |
| 55 | | their use in developing, configuring, and monitoring a waveform application. Finally, Section~\ref{conclusion} concludes and presents future work. |
| | 54 | Section~\ref{waveformworkshoptools} introduces the tools that are part of the Waveform Workshop and the problems they solve. Sections~\ref{sdrapplicationdevelopment-oef} through \ref{sdrapplicationdevelopment-alf} describe |
| | 55 | their use in developing, configuring, and monitoring a waveform application. Finally, Section~\ref{conclusion} concludes and presents future work. |
| 63 | | The SCA is an open standard associated with the U.S. Department of Defense Joint Tactical Radio System (JTRS) \cite{JTRS}. The SCA is a component and service based architecture that delineates and provides guidance for implementation of SDR infrastructure (e.g., the SCA core framework) and application software (e.g., waveform applications and signal processing components). The SCA makes extensive use of software design patterns \cite{Gamma}, and of other standards that include POSIX \cite{POSIX}, CORBA \cite{CORBA}, and XML \cite{XML}. Components in the SCA communicate using Common Object Resource Broker Architecture (CORBA) middleware, which provides interprocess communication and remote procedure call capability, allowing methods associated with remote components implemented in a variety of languages to appear as local function calls. XML is used to store profile or configuration information. |
| 64 | | |
| 65 | | The extensive nature of the SCA and its use of CORBA and XML both motivate and facilitate development of tools for SCA-based SDR development. In most cases it is desirable to abstract the interface code required by CORBA from communications engineers who develop signal processing components and waveform applications. Fortunately, use of CORBA also allows SCA-based SDR software tools to interact with SCA core frameworks and applications that are running either locally or remotely, while tools can parse XML files to obtain information needed to perform a variety of functions. Commercial tools for SCA-based SDR development are available \cite{SpectraCX}, but the The Waveform Workshop fills a need for an open source Software Communications Architecture (SCA)-based SDR toolset. These tools allow prototyping of SDR waveform applications that can be ported to work with commercial SCA software \cite{singh}, such as the Harrisâ Domain Manager Tool Kit (dmTK) \cite{dmtk}, the SCA Reference Implementation (SCARI) tools from the Canadian Communications Research Centre \cite{jianxin}, and Prism Techâs Spectra CX. The latter retails for as low as \$15,000 per team member based on a 10-member team \cite{prismtech}. The GNU Radio Companion (GRC) \cite{gnuradio} is an open source tool for SDR that provides many capabilities similar to those of the tools described in this paper, but is not based on the SCA. |
| | 63 | The SCA is an open standard associated with the U.S. Department of Defense Joint Tactical Radio System (JTRS) \cite{JTRS}. The SCA is a component- and service-based architecture that delineates and provides guidance for implementation of SDR infrastructure (e.g., the SCA core framework) and application software (e.g., waveform applications and signal processing components). The SCA makes extensive use of software design patterns \cite{Gamma}, and of other standards that include POSIX \cite{POSIX}, CORBA \cite{CORBA}, and XML \cite{XML}. Components in the SCA communicate using Common Object Resource Broker Architecture (CORBA) middleware, which provides interprocess communication and remote procedure call capability, allowing methods associated with remote components implemented in a variety of languages to appear as local function calls. XML is used to store profile or configuration information. |
| | 64 | |
| | 65 | The extensive nature of the SCA and its use of CORBA and XML both motivate and facilitate development of tools for SCA-based SDR development. In most cases it is desirable to abstract the interface code required by CORBA from communications engineers who develop signal processing components and waveform applications. Fortunately, use of CORBA also allows SCA-based SDR software tools to interact with SCA core frameworks and applications that are running either locally or remotely, while tools can parse XML files to obtain information needed to perform a variety of functions. |
| | 66 | |
| | 67 | Commercial tools for SCA-based SDR development are available \cite{SpectraCX}, but the Waveform Workshop fills a need for an open-source SCA-based SDR tool set. These tools allow prototyping of SDR waveform applications that can be ported to work with commercial SCA software \cite{singh}, such as the Harris Domain Manager Tool Kit (dmTK) \cite{dmtk}, the SCA Reference Implementation (SCARI) tools from the Canadian Communications Research Centre \cite{jianxin}, and Prism Tech's Spectra CX. The latter retails for as low as \$15,000 per team member based on a 10-member team \cite{prismtech}. The GNU Radio Companion (GRC) \cite{gnuradio} is an open source tool for SDR that provides many capabilities similar to those of the tools described in this paper, but is not based on the SCA. |
| 68 | | The Open Source SCA Implementation::Embedded (OSSIE) project |
| 69 | | |
| 70 | | The OSSIE software, including the Waveform Workshop, is a widely used and distributed software package. There have been over 20,000 downloads of the OSSIE source code, ready-to-run live DVDs, and VMware images. We have confirmed that students and engineers at over 20 different universities, companies, nonprofit research centers, and government laboratories have used OSSIE. |
| | 70 | The Open Source SCA Implementation::Embedded (OSSIE) project \cite{Aguayo2009b}, \cite{ossieweb} was started in 2003 by postdoctoral fellow Max Robert and a team of students at Virginia Tech \cite{Robert2004}. OSSIE software consists of a core framework and a library of application software (signal processing components and waveform applications). The project also provides materials for SDR education, training, and self study. |
| | 71 | |
| | 72 | The OSSIE software, including the Waveform Workshop, is a widely used and distributed software package. There have been over 20,000 downloads of the OSSIE source code, ready-to-run live DVDs, and VMware images. Students and engineers at over 20 different universities, companies, nonprofit research centers, and government laboratories have used OSSIE. |
| 88 | | The OSSIE Eclipse Feature (OEF) is the first tool in the Waveform Workshop and was created to automate as much of the SDR development process as possible. OEF is a port of the OSSIE Waveform Developer (OWD), an open source python tool for rapid component and waveform prototyping. OWD automatically generated the necessary XML descriptors and CORBA-enabled skeleton implementations for SCA-based signal processing components and XML files that specify instantiation, interconnection, and deployment of these components to form waveform applications that implement end-to-end communications waveforms. In the case of signal processing components, the SDR developer who used OWD then had to fill in the implementation using his or her editor of choice. Finally, the component or waveform files had to be built and installed from the command line using automake tools. The motivation for OEF was to duplicate the automatic generation functionality of OWD that ensured SCA compatibility, while adding support for the rest of the development process. To do so, OEF was developed as a plug-in for the Eclipse IDE. Along with providing the automatic generation from OWD, OEF also takes care of building and installing components and waveforms, eliminating many opportunities for introducing user error. It also brings all the advantages of a modern IDE. This includes syntax highlighting of source code in multiple programming languages, a simple to use interface for version control systems such as CVS and SVN, and project management tools. |
| | 90 | The OSSIE Eclipse Feature (OEF) is the first tool in the Waveform Workshop and was created to automate as much of the SDR development process as possible. OEF is a port of the OSSIE Waveform Developer (OWD), an open-source Python tool for rapid component and waveform prototyping. OEF automatically generates the necessary XML descriptors and CORBA-enabled skeleton implementations for SCA-based signal processing components, as well as XML files that specify instantiation, interconnection, and deployment of these components to form end-to-end communications waveforms. |
| | 91 | |
| | 92 | In the case of signal processing components, the SDR developer who used the earlier OWD tool still needed to fill in the implementation by using his or her editor of choice to complete each component's skeleton code. Finally, the component or waveform files had to be built and installed from the command line using automake tools. In addition to duplicating the automatic generation functionality of OWD that ensured SCA compatibility, OEF also adds support for the rest of the development process. To do so, OEF was developed as a plug-in for Eclipse, a popular open-source development environment. Along with providing the automatic generation from OWD, OEF also takes care of building and installing components and waveforms, eliminating many opportunities for introducing user error. It also brings all the advantages of a modern IDE. This includes syntax highlighting of source code in multiple programming languages, a simple-to-use interface for version control systems such as CVS and SVN, and project management tools. |
| 92 | | The second tool in the Waveform Workshop is the Waveform Dashboard (WaveDash) that was created to allow users to control SDR applications in real time. WaveDash is an interactive and reconfigurable graphical user interface (GUI) tool that acts as a control panel for working with multiple SDR waveform applications and their components. This tool abstracts the SCA interfaces and the implementation of the OSSIE core framework by providing a direct manipulation interface to control and configure SDR waveforms interactively. Users can install, start, stop and uninstall a running waveform application. More importantly, it allows users to configure the properties of a waveform's components in real time. Prior to WaveDash, all component properties were set at runtime and could not be changed. Finally, the Wavedash interface is customizable on a per-waveform and per-component basis. Users can choose the type of widget they prefer to represent a particular property of a component and can also hide or unhide specific components or properties. |
| | 96 | The second tool in the Waveform Workshop is the Waveform Dashboard (WaveDash), which allows users to control SDR applications in real time. WaveDash is an interactive and reconfigurable graphical user interface (GUI) tool that acts as a control panel for working with multiple SDR waveform applications and their components. This tool abstracts the SCA interfaces and the implementation of the OSSIE core framework by providing a direct manipulation interface to control and configure SDR waveforms interactively. |
| | 97 | |
| | 98 | Users can install, start, stop, and uninstall a running waveform application using WaveDash. More importantly, it allows users to configure the properties of a waveform's components in real time. Prior to WaveDash, all property settings were loaded by a component when it started execution, and no interface was readily available to change them other than editing XML files and restarting the application. |
| | 99 | |
| | 100 | In addition, the WaveDash interface is customizable on a per-waveform and per-component basis. Users can choose the type of widget they prefer to represent a particular property of a component and can also hide or unhide specific components or properties. This allows developers to tailor WaveDash to create a custom GUI for manipulating any waveform in the way that best suits their personal workflow. |
| 97 | | The Waveform Workshop monitors running SDR applications with a tool called ALF. The initial ALF tool was contributed to the project by SAIC in 2007, and Virginia Tech has maintained and enhanced the tool since that time. Implemented in Python, ALF interacts with a local or remote OSSIE Domain Manager via CORBA, and allows users to install, start, stop, and uninstall OSSIE waveforms, and to display the waveforms in block-diagram form. ALF includes plug-in tools that can be used to inject signals into signal processing components and to monitor signals at various points in a waveform application by playing those signals through speakers or by plotting them on a graph. |
| | 105 | The Waveform Workshop monitors running SDR applications with a tool called ALF. The initial version of ALF was contributed to the OSSIE project by SAIC in 2007, and Virginia Tech has maintained and enhanced the tool since that time. Implemented in Python, ALF interacts with a local or remote OSSIE Domain Manager via CORBA, and allows users to install, start, stop, and uninstall OSSIE waveforms, and to display a waveform in the form of a block diagram of its CORBA components and their interconnections. ALF includes plug-in tools that can be used to inject signals into signal processing components and to monitor signals at various points in a waveform application by playing those signals through speakers or by plotting them on a graph. |
| 126 | | To create a signal processing component, the user first opens Eclipse and selects Window $\rightarrow$ Open Perspective $\rightarrow$ Other. The user then chooses OSSIE from the list. This opens the OSSIE perspective, a special Eclipse layout designed for SDR development. Next, the user chooses File $\rightarrow$ New $\rightarrow$ OSSIE Component to open the New OSSIE Component Project wizard. The user then fills in the name for the project, in this case ``amplifierIQ'' and clicks on finish. This creates a new component project called amplifierIQ with a makefile and a .ocd file which is opened in the Component Editor (CE) as seen in Figure~\ref{fig:oef:generatenewcomponent}. |
| 127 | | The CE has the following four sections: |
| | 134 | To create a signal processing component, the user first opens Eclipse and selects Window $\rightarrow$ Open Perspective $\rightarrow$ Other. The user then chooses OSSIE from the list. This opens the OSSIE perspective, a special Eclipse layout designed for SDR development. Next, the user chooses File $\rightarrow$ New $\rightarrow$ OSSIE Component to open the ``New OSSIE Component Project'' wizard. The user then fills in the name for the project, in this case ``amplifierIQ'' and clicks on finish. This creates a new component project called \texttt{amplifierIQ} with a \texttt{makefile} and a component definition file (\texttt{.ocd}), which is opened in the Component Editor as seen in Figure~\ref{fig:oef:generatenewcomponent}. |
| | 135 | The component editor has the following four sections: |
| 129 | | \item Description - for giving the component a textual description |
| 130 | | \item Generation Options - to specify what type of component to generate |
| 131 | | \item Ports - to specify the ports the component will have |
| 132 | | \item Properties - to specify which properties the component will have |
| | 137 | \item Description---to provide a textual description for the component. |
| | 138 | \item Generation Options---to specify what type of component to generate. |
| | 139 | \item Ports---to specify the ports the component will have. |
| | 140 | \item Properties---to specify which properties the component will have. |
| 134 | | The user enters the following descriptions for the component, in this case, ``Amplifier with independently adjustable gains for in- phase and quadrature signal components.'' Under generation options, the user selects basic\_ports from the template drop-down menu and leaves both ACE Support and Timing Port Support unchecked. Selecting basic\_ports will create a C++ component. The other options in the template drop-down, custom\_ports and py-comp, are for components that use an alternative set of interfaces and for python components, respectively. Checking ACE support would let the component use certain features of ACE \cite{ACE} such as ACE threads, and checking Timing Port Support adds profiling support when using the custom ports or python component options. |
| | 142 | The user enters the following descriptions for the component, in this case, ``Amplifier with independently adjustable gains for in-phase and quadrature signal components.'' Under generation options, the user selects ``basic\_ports'' from the template drop-down menu and leaves both ``ACE Support'' and ``Timing Port Support'' unchecked. Selecting ``basic\_ports'' will create a C++ component. The other options in the template drop-down, ``custom\_ports'' and ``py-comp'', are for components that use an alternative set of interfaces and for Python components, respectively. Checking ``ACE Support'' would let the component use certain features of ACE \cite{ACE} such as ACE threads, and checking ``Timing Port Support'' adds profiling support when using the custom ports or Python component options. |
| 149 | | The next step is to add ports to the component. Ports allow components to communicate with one another. The user clicks on the Add button to bring up the Add Port dialog, as shown in Figure~\ref{fig:oef:addport}. Next, the user enters ``amp\_output'' for the port name. As defined by the SCA, a uses port is an output port while a provides port is an input port. Since this port will be an output port, the user selects uses from the type drop-down menu. To choose the type of interface, the user clicks the arrow next to Standard Interfaces and selects complexShort from the choices that appear. Finally, the user clicks ok and sees the new port in the Ports table. All three fields in the Ports table are editable in place, so the user can change the name, interface, or type of any port after it has been added. Selecting any port and clicking the Remove button will delete the port from the component. This component also requires an input port called amp\_input that is also a complexShort, so the user adds another port using the same process |
| 150 | | |
| 151 | | \subsection{Adding Properties to a Component} |
| | 157 | The next step is to add ports to the component. Ports allow components to communicate with one another. The user clicks on the ``Add'' button to bring up the ``Add Port'' dialog, as shown in Figure~\ref{fig:oef:addport}. Next, the user enters ``amp\_output'' for the port name. As defined by the SCA, a \textit{uses} port is an output port while a \textit{provides} port is an input port (a ``backwards'' naming convention due to the client/server orientation of the CORBA remote procedure call implementation of ports). Since this port will be an output port, the user selects ``Uses'' from the ``Port Type'' drop-down menu. To choose the type of interface, the user clicks the arrow next to ``Standard Interfaces'' and selects ``complexShort'' from the choices that appear. Finally, the user clicks ``OK'' and sees the new port in the ``Ports'' table. All three fields in the ``Ports'' table are editable in place, so the user can change the name, interface, or type of any port after it has been added. Selecting any port and clicking the ``Remove'' button will delete the port from the component. This component also requires an input port called \texttt{amp\_input} that is also a \texttt{complexShort}, so the user adds another port using the same process |
| | 158 | |
| | 159 | \subsection{Adding properties to a component.} |
| 164 | | The final step to configure a component is to add any reconfigurable properties it will need. Our amplifier needs two properties: I\_gain and Q\_gain, which multiply in-phase and quadrature components of the signal, respectively. The user clicks the Add button in the Properties section of the CE to bring up the Add Property dialog as shown in Figure~\ref{fig:oef:addproperty}. The user enters ``I\_gain'' for the name and 1 for the value. The rest of the attributes are left as the defaults, so the user clicks Add Property. The process is repeated for a Q\_gain property. |
| 165 | | |
| 166 | | \subsection{Generating Skeleton Implementation} |
| | 172 | The final step to configure a component is to add any reconfigurable properties it will need. Our amplifier needs two properties: \texttt{I\_gain} and \texttt{Q\_gain}, which multiply in-phase and quadrature components of the signal, respectively. The user clicks the ``Add'' button in the ``Properties'' section of the component editor to bring up the ``Add Property'' dialog as shown in Figure~\ref{fig:oef:addproperty}. The user enters ``I\_gain'' for the name and ``1'' for the value. The rest of the attributes are left as the defaults, so the user clicks ``Add Property''. The process is repeated for a \texttt{Q\_gain} property. |
| | 173 | |
| | 174 | \subsection{Generating a skeleton implementation.} |
| 198 | | Now that the user has successfully created and built a component, that component can be used in a new waveform. The example waveform will be a simulated QPSK communications link that runs entirely on a single computer, but essentially the same approach is used to develop a radio communication waveform application that interfaces with RF hardware. To get started, the user selects File $\rightarrow$ New $\rightarrow$ OSSIE Waveform to open the New OSSIE Waveform Project wizard. The user names the project SimpleWaveform and clicks on Finish. When the wizard finishes, the user is presented with the main Waveform Editor (WE), as seen in Figure~\ref{fig:oef:waveformeditor}. The WE contains 3 main trees: |
| | 206 | Now that the user has successfully created and built a component, that component can be used in a new waveform. The example waveform illustrated in this section will be a simulated QPSK communications link that runs entirely on a single computer, but essentially the same approach is used to develop a radio communication waveform application that interfaces with RF hardware. To get started, the user selects File $\rightarrow$ New $\rightarrow$ OSSIE Waveform to open the ``New OSSIE Waveform Project'' wizard. The user names the project ``SimpleWaveform'' and clicks on Finish. When the wizard finishes, the user is presented with the main Waveform Editor, as seen in Figure~\ref{fig:oef:waveformeditor}. The waveform editor contains 3 main trees: |
| 200 | | \item Available Resources - lists all of the components, devices, and nodes currently installed on the system and available for use |
| 201 | | \item Waveform - shows all of the components to be used in the current waveform |
| 202 | | \item Platform - shows the processing and input/output devices and nodes (collections of devices) and components allocated to each processing device in the current waveform. |
| | 208 | \item Available Resources---lists all of the components, devices, and nodes currently installed on the system and available for use. |
| | 209 | \item Waveform---shows all of the components to be used in the current waveform. |
| | 210 | \item Platform---shows the processing and input/output devices that comprise the platform, together with the components in the current waveform that are allocated to each processing device. |
| 215 | | Note that the component that the user just built, amplifierIQ, is listed under components in the Available Resources tree. To begin assembling the waveform, the user drags the TxDemo, amplifierIQ, ChannelDemo, and RxDemo components he will be using to the Waveform tree. The user then drags a default\_GPP\_node (which corresponds to a set of XML files for a Device Manager that manages a single x86 GPP) from nodes in the Available Resources tree to the Platform tree. The components are not yet allocated to any nodes, so the user expands default\_GPP\_node to show the device on it, GPP1. To allocate the components to this device, the user drags each component from the Waveform tree to GPP1 in the Platform tree. When a component is allocated to GPP1, the component appears under the GPP1 in the Platform tree, and the label on the component in the Waveform tree also changes to reflect where it has been allocated. |
| | 223 | Note that the component that the user just built, \texttt{amplifierIQ}, is listed under components in the ``Available Resources'' tree. To begin assembling the waveform, the user drags the \texttt{TxDemo}, \texttt{amplifierIQ}, \texttt{ChannelDemo}, and \texttt{RxDemo} components he will be using to the ``Waveform'' tree. The user then drags a \texttt{default\_GPP\_node} (which corresponds to a set of XML files for a Device Manager that manages a single x86 GPP) from the ``Available Resources'' tree to the ``Platform'' tree. The components are not yet allocated to any nodes, so the user expands \texttt{default\_GPP\_node} to show the device on it, \texttt{GPP1}. To allocate the components to this device, the user drags each component from the ``Waveform'' tree to \texttt{GPP1} in the ``Platform'' tree. When a component is allocated to \texttt{GPP1}, the component appears under the \texttt{GPP1} in the ``Platform'' tree, and the label on the component in the ``Waveform'' tree also changes to reflect where it has been allocated. |
| 227 | | Next, the user has to connect the ports on the components to one another. The user expands each component in the Waveform tree to show their ports, as seen in Figure~\ref{fig:oef:portsoncomponents}. To connect the ports, the user drags a uses port (orange puzzle piece) on top of a provides port (blue puzzle piece). Each port then changes to a combined puzzle piece and its label changes to show which port it is connected to. Once all of the ports have been connected, the final step is to choose an Assembly Controller. To do so, the user right clicks on TxDemo1 in the Waveform tree and chooses Set Assembly Controller. Once this is done, the waveform is fully configured and ready to be generated. The user saves the project and OEF automatically generates all of the necessary XML descriptors for the waveform and installs them to the filesystem. |
| 228 | | |
| 229 | | \section{SDR Application Development Work Flow Part 2: Configuring and Controlling Waveform Applications using the Waveform Dashboard} |
| | 235 | Next, the user has to connect the ports on the components to one another. The user expands each component in the ``Waveform'' tree to show their ports, as seen in Figure~\ref{fig:oef:portsoncomponents}. To connect the ports, the user drags a uses port (orange puzzle piece) on top of a provides port (blue puzzle piece). Each port then changes to a combined puzzle piece and its label changes to show which port it is connected to. Once all of the ports have been connected, the final step is to choose an Assembly Controller. To do so, the user right clicks on \texttt{TxDemo1} in the ``Waveform'' tree and chooses ``Set Assembly Controller''. Once this is done, the waveform is fully configured and ready to be generated. The user saves the project and OEF automatically generates all of the necessary XML descriptors for the waveform and installs them. |
| | 236 | |
| | 237 | \section{SDR Application Development Work Flow, Part 2: Configuring and Controlling Waveform Applications using WaveDash} |
| 265 | | The ability to configure component properties at runtime was urgently needed and was one of the prime motivations to develop WaveDash. Prior to development of WaveDash, OSSIE users had to either modify property values in OEF and rebuild the waveform or hand-edit the property values in the component or waveform XML files and restart the waveform to configure the component property values. |
| 266 | | |
| 267 | | WaveDash enables users to modify component properties at runtime thus making SDR waveform development more live and interactive. Each property has a GUI widget (e.g. text box, slider) associated with by default based on its data type (e.g. integer, string). The user can change the value of a component property by changing the value of the widget. By default, the property will be updated immediately after the user presses TAB or ENTER. Changes can be immediately seen with the ALF Plot tool (see Section $\ref{sdrapplicationdevelopment-alf:relationwithwavedash}$). In this example, the user changes the value of noise\_std\_dev of ChannelDemo from default value of 1000 to a 1500 and presses ENTER for the change to take effect. |
| | 273 | The ability to configure component properties at runtime was urgently needed and was one of the prime motivations to develop WaveDash. Prior to development of WaveDash, OSSIE users had to either modify property values in OEF and rebuild the waveform or hand-edit the property values in the component or waveform XML files, then restart the waveform to configure the component property values. |
| | 274 | |
| | 275 | WaveDash enables users to modify component properties at runtime, thus making SDR waveform development more live and interactive. Each property has a GUI widget (such as a text box or slider) associated with it by default based on the property's data type (such as integer or string). The user can change the value of a component property by changing the value of the widget. By default, the property will be updated immediately after the user presses TAB or ENTER. Changes can be immediately seen with the ALF Plot tool (see Section $\ref{sdrapplicationdevelopment-alf:relationwithwavedash}$). In this example, the user changes the value of the \texttt{noise\_std\_dev} property of the \texttt{ChannelDemo} component from its default value of 1000 to a new value of 1500 and presses ENTER for the change to take effect. |
| 310 | | The ALF environment provides several capabilities that are helpful for troubleshooting and debugging waveform applications. Like WaveDash, ALF allows users to install/uninstall and start/stop waveform applications. ALF also lets users display running applications in block diagram form as shown in Figure~\ref{fig:alf:block_diagram_plugin_menu}, and inject and monitor signals at various component ports in the application using provided or user-developed plug-in tools by right clicking on the port (small rectangle on either side of the component) and selecting from a menu as shown in Figure~\ref{fig:alf:block_diagram_plugin_menu}. Additional features enable creation and installation of single-component waveform applications, and connections between ports on components in the same or different waveforms at run time. |
| | 318 | The ALF environment provides several capabilities that are helpful for troubleshooting and debugging waveform applications. Like WaveDash, ALF allows users to install, uninstall, start, and stop waveform applications. ALF also lets users display running applications as block diagrams as shown in Figure~\ref{fig:alf:block_diagram_plugin_menu}, and inject or monitor signals at various component ports in the application. Port signals are manipulated using provided or user-developed plug-in tools by right clicking on the port (small rectangle on either side of the component) and selecting from a menu, as shown in Figure~\ref{fig:alf:block_diagram_plugin_menu}. Additional features enable creation and installation of single-component waveform applications, and adding connections between ports on components in the same or different waveforms at run time. |
| 331 | | The plot tool (Figure~\ref{fig:alf:plugins:plot}) accepts signals from uses ports that employ the complexShort, complexLong, or complexFloat interfaces. This tool displays the signal's spectrum on a normalized frequency scale or displays the quadrature vs. in-phase components of the signal in order to render the constellations of received digital signals. In this mode, several symbols are shown that under ideal conditions would typically map to regularly spaced points, e.g. in a circle (phase-shift keying) or in a grid pattern (quadrature amplitude modulation). In a realistic simulation or an operational radio, the symbols appear scattered about the nominal symbol points due to noise, fading, imperfect synchronization, or other anomalies. |
| | 339 | ALF's plot tool (Figure~\ref{fig:alf:plugins:plot}) accepts signals from uses ports (output ports) that employ the \texttt{complexShort}, \texttt{complexLong}, or \texttt{complexFloat} interfaces. This tool displays the signal's spectrum on a normalized frequency scale or displays the quadrature vs. in-phase components of the signal in order to render the constellations of received digital signals. In this mode, several symbols are shown that would typically map to regularly spaced points under ideal conditions, e.g. in a circle (phase-shift keying) or in a grid pattern (quadrature amplitude modulation). In a realistic simulation or an operational radio, the symbols appear scattered about the nominal symbol points due to noise, fading, imperfect synchronization, or other anomalies. |
| 365 | | The Connect Tool, shown in Figure~\ref{fig:alf:connect}, provides the capability |
| 366 | | to connect ports on components that are in the same or different waveform applications. This capability, provided with the ability to create and run single-component applications, allows users to build ad-hoc applications by launching components as waveforms and connecting them at run time to form a temporary, larger composite application. The composite application can be configured by configuring properties of its constituent applications using the WaveDash tool. Further, the Connect Tool provides for automatic instantiation and tear-down of applications, and automated connections and routing of data among standard applications, single-component applications, and plug-in tools, by specifying the applications/components/plug-ins and behavior in an XML file. This enables dynamic reconfiguration of SDR applications at run time, a capability that can be very useful in cognitive radio implementations \cite{Cormier2009}. |
| | 373 | ALF's connect tool, shown in Figure~\ref{fig:alf:connect}, provides the capability |
| | 374 | to connect ports on components that are in the same or different waveform applications. This capability, provided with the ability to create and run single-component applications, allows users to build ad-hoc applications by launching components as waveforms and connecting them at run time to form a temporary, larger composite application. The composite application can be configured by using WaveDash to change the properties of its constituent applications. Further, the connect tool provides for automatic instantiation and tear-down of applications, and automated connections and routing of data among standard applications, single-component applications, and plug-in tools, by specifying the applications/components/plug-ins and behavior in an XML file. This enables dynamic reconfiguration of SDR applications at run time, a capability that can be very useful in cognitive radio implementations \cite{Cormier2009}. |
| 381 | | When used together, ALF and WaveDash allow the user to observe effects of parameter values on waveform operation interactively and in real time, as shown in Figure~\ref{fig:alf:alfwavedash}. This is very useful for education, experimentation, and debugging of waveform applications. ALF currently provides some of the same functionality as WaveDash, specifically the ability to install/uninstall and start/stop applications. This functionality is included in both tools so that WaveDash can be used to run applications without the full ALF environment. WaveDash can be launched from within ALF, and further integration is planned. Planned changes include encapsulating all common functionality of the two tools in a python module or library. |
| | 389 | When used together, ALF and WaveDash allow the user to observe effects of parameter values on waveform operation interactively and in real time, as shown in Figure~\ref{fig:alf:alfwavedash}. This is very useful for education, experimentation, and debugging of waveform applications. ALF currently provides some of the same functionality as WaveDash, specifically the ability to install, uninstall, start, and stop applications. This functionality is included in both tools so that WaveDash can be used to run applications without the full ALF environment. WaveDash can be launched from within ALF, and further integration is planned. Planned changes include encapsulating all common functionality of the two tools in a single Python module or library. |
| 394 | | This paper has described a set of open source SCA-based SDR development tools suitable for education, research, and rapid prototyping. These tools ensure SCA compatibility while hiding much its complexity and exploiting the features provided by the SCA to enable an interactive development process. The development process includes creation of signal processing components and waveform applications as well as execution, visualization, monitoring, testing, and run time reconfiguration of waveform applications. The Waveform Workshop tools also allow developers to interact directly with running waveforms and components, as well as to visualize the output from components in real time. |
| 395 | | |
| 396 | | The current state of the Waveform Workshop allows for the work flow shown in the preceding sections, and also allows applications such as SimpleWaveform to be prototyped as ad-hoc applications and then recreated in OEF. Additional development of the tools will enhance this capability by providing the option to display multiple application block diagrams in the same window, make connections by dragging lines between component ports, and formalize ad-hoc applications by saving connection and deployment information in waveform XML files. |
| 397 | | |
| 398 | | Upgrades to the Waveform Workshop are planned as part of current and pending projects. These include the following tasks: |
| | 402 | This paper describes a set of open-source, SCA-based SDR development tools suitable for education, research, and rapid prototyping. These tools ensure SCA compatibility while hiding much its complexity. They exploit the features provided by the SCA to enable an interactive development process. The development process includes creation of signal processing components and waveform applications as well as execution, visualization, monitoring, testing, and run time reconfiguration of waveform applications. The Waveform Workshop tools also allow developers to interact directly with running waveforms and components, as well as to visualize the output from components in real time. |
| | 403 | |
| | 404 | The current state of the Waveform Workshop allows for the work flow shown in this paper, and also allows applications such as \texttt{SimpleWaveform} to be prototyped as ad-hoc applications and then recreated in OEF. Additional development of the tools will enhance this capability by providing the option to display multiple application block diagrams in the same window, make connections by dragging lines between component ports, and formalize ad-hoc applications by saving connection and deployment information in waveform XML files. |
| | 405 | |
| | 406 | Upgrades to the Waveform Workshop are planned as part of current and pending projects. Future work includes: |
| 400 | | \item Integrate ALF and WaveDash capabilities into Eclipse |
| 401 | | |
| 402 | | \item Enable interactive block-diagram based waveform development |
| 403 | | |
| 404 | | \item Move to full Java implementations of the tools |
| 405 | | |
| 406 | | \item Load waveform and component descriptions by parsing the SCA XML files rather than using dedicated file formats (.ocd and .owd) |
| 407 | | |
| 408 | | \item Add features for collaborative development |
| 409 | | |
| 410 | | \item Add model-based development capability for greater flexibility |
| | 408 | \item Integrating ALF and WaveDash capabilities into Eclipse. |
| | 409 | |
| | 410 | \item Enabling interactive, graphical, block-diagram-based waveform development. |
| | 411 | |
| | 412 | \item Moving to a Java implementations of the tools. |
| | 413 | |
| | 414 | \item Loading waveform and component descriptions by parsing the SCA XML files rather than using dedicated file formats (such as \texttt{.ocd} and \texttt{.owd}). |
| | 415 | |
| | 416 | \item Adding features for collaborative development. |
| | 417 | |
| | 418 | \item Adding model-based development capabilities for greater flexibility. |
| 428 | | C. R. Aguayo Gonz$\acute{a}$lez, et al., Open\-Source SCA\-Based Core Framework and Rapid Development Tools Enable Software\-Defined Radio Education and Research, IEEE Communications Magazine, vol 47, no. 10, October 2009. |
| | 436 | C. R. Aguayo Gonz$\acute{a}$lez, et al., ``Open-Source SCA-Based Core Framework and Rapid Development Tools Enable Software-Defined Radio Education and Research,'' \textit{IEEE Communications Magazine}, vol. 47, no. 10, October 2009. |
| 440 | | C. R. Aguayo Gonz$\acute{a}$lez , C. B. Dietrich, and J. H. Reed, Understanding the software communications architecture, IEEE Communications Magazine, vol. 47, no. 9, September 2009. |
| | 448 | C. R. Aguayo Gonz$\acute{a}$lez , C. B. Dietrich, and J. H. Reed, ``Understanding the Software Communications Architecture,'' \textit{IEEE Communications Magazine}, vol. 47, no. 9, September 2009. |
| 480 | | A.R. Cormier, C.B. Dietrich, J. Price, and J.H. Reed, Dynamic reconfiguration of software defined radios using standard architectures, Physical Communication, June, 2010, doi: 10.1016/j.phycom.2009.09.002 |
| | 488 | A.R. Cormier, C.B. Dietrich, J. Price, and J.H. Reed, ``Dynamic Reconfiguration of Software Defined Radios Using Standard Architectures,'' \textit{Physical Communication}, June, 2010. doi: 10.1016/j.phycom.2009.09.002 |
| 483 | | Singh, S., M. Adrat, M. Antweiler, âNATO RTO/IST RTG on SDR: Demonstrating Portability and Interoperability of SCA-Based Waveforms,â SDR 09 Technical Conference and Product Exposition, Washington, DC, December 1-4, 2009. |
| | 491 | Singh, S., M. Adrat, M. Antweiler, ``NATO RTO/IST RTG on SDR: Demonstrating Portability and Interoperability of SCA-Based Waveforms,'' SDR 09 Technical Conference and Product Exposition, Washington, DC, December 1-4, 2009. |
| 489 | | Guan Jianxin, Ye Xiaohui, Gao Jun, Wang Bo, âThe Flow of Software Defined Radio Waveform Development Based on SCARI,â Wireless Communications, Networking and Mobile Computing, 2009. WiCom '09. 5th International Conference on, Beijing, China, 24-26 September 2009. |
| | 498 | Guan Jianxin, Ye Xiaohui, Gao Jun, Wang Bo, ``The Flow of Software Defined Radio Waveform Development Based on SCARI,'' In \textit{Proceedings of the 5th International Conference on Wireless Communications, Networking and Mobile Computing}, Beijing, China, 24--26 September 2009. doi: 10.1109/WICOM.2009.5302559 |
