- About CREW
- Open Calls
Welcome to the CREW repository. The CREW repository contains several types of data, which are useful to be shared with other experimenters. Following types of data are available in the CREW repository:
- full experiment descriptions: when reading reports containing experimental results, it is often difficult or impossible to verify the claims of the author or to use the experiment as a base for further research. Full experiment descriptions contain all information needed to run a particular experiment. By publishing full experiment descriptions, we aim for more transparent and thus more valuable experimentation results. Furthermore, the possibility to re-use experiment descriptions enables fair comparison (benchmarking) of wireless solutions.
- traces: traces are files describing wireless traffic of heterogeneous technologies, either at packet level (e.g. a timed sequence of Bluetooth packets, a certain profile of Wi-Fi use, the activity of a primary LTE user), or data recorded at spectrum level (e.g. recorded spectrum use in a certain location over time, artificially created patterns containing reference interference on a particular frequency).
- background environments: background environments are configurations that can be used to create repeatable wireless environments and are tied to a particular testbed. The configuration files of these background environments link particular traces (see above) to particular nodes at particular times, inside a particular testbed.
- processing scripts: useful scripts (in multiple scripting languages, matlab, java,...) that, for example, convert the output of a certain type of commercially available tool or CREW component to the CREW common data format.
- performance metrics and benchmarking scores: detailed descriptions of performance metrics and measurement methodologies. Benchmarking scores are abstractions derived by combining different performance metrics. These scores may be useful to when comparing a large set of solutions or for automated decision making during an automated benchmarking process.
The CREW repository is currently being populated by members of the CREW consortium. In this early stage of development, it is currently not possible for external experimenters to directly contribute to the repository. If you want to contribute your data to the repository, please contact us. In a later phase, it will either be possible to directly contribute to the CREW repository, or, CREW may decide to migrate its data with existing open repositories such as CRAWDAD.
The data available on the CREW repository may be used for your research, educational and commercial purposes. Unless specified otherwise, copying and modifying the data is allowed, as long as a reference to the CREW project or CREW repository (www.crew-project.eu/repository) is included. In addition, any other references to the original authors that might be included in the data should not be removed. When data from the repository is used in publications, CREW should also be referenced.