“Nutaq’s PicoSDR will enable CorteXlab users to prototype highly computational PHY layer such as wideband MIMO OFDM and to design cooperative transmission techniques in a multi-node real time environment with perfect interference control.”
Head of CorteXlab project
In anticipation of the “internet of things”, where billions of devices will be interconnected, develop a cognitive radio test bed that will allow the global scientific community to simultaneously test three components critical to the realization of this massive wireless network:
- Software defined radios
- Cognitive radios
- Self organizing networks
The creation of CorteXlab at INSA-Lyon, a laboratory consisting of dozens of software defined radio nodes in a highly shielded environment. Nutaq’s PicoSDR platform will provide researchers around the world with the high performance development environment needed to design, benchmark, and tune their cognitive radio protocols.
The “internet of things” is predicted to link tens of billions of devices together by 2020. Given the current wireless spectrum crunch it is clear that there is an urgent need to better understand how these devices are to be wirelessly connected to one another.
Due to the complexity of the problem, researchers need a large scale platform to deploy and test the radio hardware and software algorithms which will enable this massive network.
This research platform must allow the community to simultaneously test:
- Radios that can reconfigure themselves in real time (Software Defined Radios).
- Radios that are aware of their wireless environment and can respond to it intelligently (Cognitive Radios).
- Self organizing networks (Networks that can plan, configure, manage, optimize, and heal on their own).
Until now, this research platform hasn’t existed. The CorteXlab team at INSA-Lyon aims to change that.
About CorteXlab at INSA-Lyon
The INSA-Lyon CorteXlab team is building a 180 square meter, shielded, heterogeneous wireless testbed in order to provide the global scientific community with the hardware and software necessary to test the physical layer (PHY layer) of radio transmissions.
CorteXlab will consist of dozens of software defined radio (SDR) nodes composed of a mix of “standard performance” and “high performance” nodes.
The high performance SDR nodes, which are in fact Nutaq’s PicoSDR platform, are a combination of 2×2 and 4×4 MIMO nodes.
Access to these high performance PicoSDR nodes will allow researchers using CorteXlab to:
- Access a broad range of experiments such as multi-user communications & cooperative scenarios.
- Research ways to improve communication system performance in areas such as modulation, coding, multiple access protocols, routing algorithms, channel estimation and modeling.
Why CorteXlab Chose Nutaq’s PicoSDR
When selecting the software defined radio platform that would supply the high performance demands of the cognitive radio research community, the team at CorteXlab understandably had a stringent list of requirements.
At a functional level these requirements necessitated the ability for researchers around the world to be able to change physical parameters of the SDR hardware such as operating frequency, channel bandwidth, emitted power, and waveform.
However other factors such as channel growth strategy, radio upgradability, size & weight of the hardware, as well as the software development environment played an important role in the selection process.
Nutaq’s PicoSDR was chosen due to its ability to meet both the current requirements of researchers, as well as providing an upgrade strategy enabling future hardware enhancements.
Key features of the PicoSDR which are critical to enabling this large cognitive radio testbed include:
- Capability of running in either a 2×2 or 4×4 MIMO configuration.
- MIMO OFDM support (PicoSDR includes a MIMO OFDM reference design).
- Auto calibrated, dynamically reconfigurable radios with a high degree of isolation between the Tx and Rx paths.
- Large baseband signal processing capabilities (1 or 2 Virtex-6 FPGAs onboard).
- Tight coupling of distinct FPGAs via AMC, in order to avoid real-time limitations imposed by external cabling.
- Ability to store multiple FPGA images in FLASH memory, allowing, for example, rapid waveform switching in the field.
- Support for multi-user communication.
- Presence of FMC-HPC to provide easy radio card upgrade path.
- PCIe and/or GigE interface to the computer (PicoSDR offers both).
- Time synchronization between independently located radio nodes.
- GNU Radio support.
- Model-based design wasn’t required…but was viewed as a favourable addition to the proposed solution.
- Small form factor and as light as possible as the radio nodes are to be suspended from the ceiling of the lab.
As researchers use the CorteXlab cognitive radio testbed Nutaq’s PicoSDR will aid in achieving the goal of presenting the global SDR development community with an environment to research and develop for the “internet of things”.
The insights gained at INSA-Lyon will help interconnect tens of billions of devices over the coming decade, creating the opportunity for a vast number of improvements to our daily lives.