Open Source LTE Project | Interference Detection/Avoidance | High-Throughput DSA Video | Channel Modeling and Computing Task Demo for WDC | Position Location Using Distributed SDR Nodes | Wireless Communications Testbed (WCT)
The Spectrum Management Research Testbed – Self Sustaining Broadband Network (SMART-SSBN) will consist of a fleet of mobile or rapidly deployable Software-Defined Radio (SDR) nodes that offer unique economic value to Wireless Internet Service Providers (WISP), other private and public companies, research institutes as well as DoD and other government agencies who are quickly adapting with the spectrum sharing technology trend. SMART-SSBN creates this economic opportunity by offering a very affordable and flexible way for interested parties to perform research and testing on spectrum sharing in the real world environments of their choosing. These same radio nodes can also be used to set up broadband service infrastructure in remote and currently underserved locations where such service is unavailable, unreliable, or slow. The revenue from research and testing activities will be used to help offset the cost of providing service in areas that have relatively few potential subscribers. We have developed a proof of concept system that uses commercial off the shelf (COTS) SDR equipment to demonstrate some of the capabilities described above. These capabilities include broadband service provision, spectrum sensing, and basic cognitive radio experimentation. Additional SDR nodes will be used to test and demonstrate the proof of concept system. Outdoor CORNET (O-CORNET) provides an excellent opportunity to perform tests in a similar environment to that of the final system implementation. The final system will be similar in functionality, though much more capable than the proof of concept system which is limited by available time and resources.
The goal of this project is to implement a low cost cognitive radio network testbed to be used for educational, experimental, and demonstrative purposes in the areas of software-defined radio (SDR), cognitive radio (CR), and dynamic spectrum access (DSA). The final testbed consists of a number of SDR transceivers that can be spatially reconfigured anywhere with a connection to the local area network. Software was developed that enables Raspberry Pi’s to function as SDR transmitters, increasing the utility of this testbed while maintaining its low price point. Several applications and experiments were developed to demonstrate how this testbed can be used for research and education (Micronet-SDR14 ). It enables analyzing the performance of cognitive radio networks with varying degrees of feedback and control provided by a backbone network (the LAN), among others.
The deployment of cognitive radios requires the ability to evaluate their performance in a variety of scenarios. The goal of CRTS is enabling the testing of cognitive radios in diverse environments either by observing or inferring the decisions that the radios make as a function controlled modification to the radio environment. Appropriate metrics and test plans need to be developed to establish a common test procedure. Such procedure is not existing today and slows down the commercial interest in developing cognitive radios. Different cognitive radio implementations will likely exist. CRTS will therefore define a set of test benches that can be loaded to analyze the performance and compliance of a specific radio. CRTS will also assist in categorizing cognitive radios according to their abilities as well as provide the means to develop more efficient test plans by adjusting test parameters according to the results of radios whose abilities are already known without being biased to a specific product. CRTS today is able to measure a number of radio performance parameters and analyze the behavior of adaptive software radios. Our plans are expanding the system to enable testing cognitive radio networks in complex environments that incorporate interferers, dynamic spectrum access with spectrum sensing, and strict spectrum access regulations.
CORNET-3D is a web application that provides a 3D view of the CORNET testbed with information on which nodes and radios are operational and how to access them. The application can also display 2D and 3D plots of the spectrum usage, which is sensed by the radios in real time. CORNET 3-D provides a visual tool for teaching students about wireless resources, the existing tradeoffs and techniques for optimal use of spectral resources in an intuitive way. Spectrum challenges, where students collaborate on or compete for the use of radio resources, are currently being developed.
Students of the ECE 5674 Software Radio class that is taught each fall at Virginia Tech installed the open-source SDR framework ALOE and configured the open-source LTE waveform for real-time data transmission over the air. CORNET users have access to the waveform on CORNET Nodes 2-5 and 2-6. The video tutorial is an mp4 file that should open with Windows Media Player or any other media program.
Extending the concept of a basic DSA application, CORNET is also being used to explore more specific ways to identify the nature of a poor radio link between two nodes. Of particular interest is determining whether the loss of channel quality between two nodes is the result of interference or fading and attenuation. Channel quality is assessed by measuring the network’s aggregate percentage of dropped packet rate (PDR). The windowed PDR is monitored for short-term fluctuations and when it drops below a certain threshold, the nodes begin to sense the channel for signals, which may be causing interference. Man-made signals are separated from noise based on a measurement of the peak-to-average power ratio of a wide slice of spectrum. With this information, the radios can implement a variety of behaviors, such as changing frequency, increasing power, or reducing the order of modulation to improve performance.
The high-throughput DSA video link was built using GNU Radio and designed to run on the USRP2/Linux server configuration employed by CORNET. The high-throughput DSA radio establishes a 1.50 Mb/s, over-the-air video link between a master node and a client node, demonstrating the ability for the USRP2 platform to be reconfigured on the fly. The radio uses a random channel access and synchronization strategy, as well as energy detection to identify empty spectrum, whereby each radio briefly asserts itself at a random frequency until both arrive at a common channel. Once synchronized, the master and client employ a modified ARQ protocol as well as energy detection to actively search for and identify potential interference. The waveform exhibits how a “mission critical” data-link, such as a video feed, can be made more robust through the use of intelligent frequency hopping behavior.
WDC enables the radio nodes with reduced computing capabilities to cooperate with other nodes to accomplish complex computational tasks while minimizing the overall execution time. The uncertainty of the dynamic wireless environment, which is not an issue for the traditional distributed computing, poses an additional challenge for WDC. A theoretical framework, task mapping algorithm, and simulation results are introduced in , which considers a heterogeneous computing and radio environment. The algorithm has been implemented and tested on CORNET for gaining additional insights on the dominating factors for further improvements, currently being investigated by the researchers. Chen et al.  applies WDC for video compression. The demonstrations show that considering channel heterogeneity is key for building power efficient and robust WDC systems.
This project uses the 12 nodes available on a single floor of the ICTAS building hosting CORNET and applies a proximity-based algorithm for position location. Proximity algorithms provide symbolic relative location information, relying on a number of receiving nodes with well-known positions. When a mobile target is detected by a node, it is considered to be collocated with it. When more than one node detects the mobile target, it is considered to be collocated with the one that receives the strongest signal. This method has been implement on CORNET and is able to locate a moving user within 3 m accuracy. User tracking can be done by monitoring the sequence of the color changes of the nodes during the random walks of the mobile user. This is a first step towards the development of more complex positioning systems based on power finger printing or indoor path loss models, for instance. Click here for more information.