FP7 CREW

June 2014

Dear subscriber,

The FP7-CREW (Cognitive Research Experimentation World) project has now been running for almost four years! In this issue of the CREW newsletter, information is given about our presence at EuCNC next month as well as about the Open Access we are offering to our facilities. Furthermore, you will find testimonies about the completed Open Call 2 experiments and about the ongoing Open Call 3 experiments.

CREW booth @ EuCNC

The final results and demonstrators of the different OC2 experiments will be shown at the European Conference on Networks and Communications (EuCNC) in Bologna (Italy) on June 23-26.

Go back to the top of the page.

Open Access

CREW is now in a continuous open access phase, offering 2 modes for the use of the CREW facilities:

1. Best effort access & basic support: CREW offers best effort access to the facilities free of charge for non-commercial use, including basic support (i.e. information from portal, guidelines, tutorials, handbooks, and limted technical support). The CREW portal will guide you to find the most suitable test facility for your experiment along with further information on how to get started.
2. Guaranteed access & advanced support: If more guarantees are required on the availability of infrastructure and more advanced technical support is needed, it is possible to submit a proposal application for an open access experiment with guaranteed availability & support. If the request is granted. CREW commits to provide the necessary facility resources and manpower to the experimenter, free of charge. Both the template for 'request for experimentation' and announcement document are available (since June 2014).

Go back to the top of the page.

Completion of OC2 experiments

CNIT, Italy

As experimenters, one of the strongest initial motivations to use the CREW testbed was undoubtedly the possibility to run tests using advanced cognitive components and a large number of wireless nodes. Later, we discovered that CREW has much more to offer:it allows to easily add software components to facilitate MAC-layer experiments and to define network 'intelligence' (based on advanced sensing of different network parameters) by using the same OMF framework originally designed for experiment control. The CREW team offered support not only in teaching us how to use the testbed facilities, but also in the design and set-up of some of our experiments about inter-technology coordination mechanisms.

Go back to the top of the page.

IT-CMSF, Portugal

IT/CMSF has been working with geo-location databases to protect incumbent DVB-T signals from TVWS device interference. Wireless microphones (WM) are usually protected by using sensing techniques on the TVWS device itself. With our participation in CREW from the second open call on, we had the unique opportunity to experiment with a new approach for sensing WMs, using the Log-a-Tec testbed from JSI. In the past year, we conducted several sensing experiments on TVWS, relying on remote access to the testbed, The JSI team provided professional and intuitive technical support whenever we needed. As a result, we have successfully combined information from the sensing network with a geo-location database access that finally led to a TVWS trial in a real scenario.

Go back to the top of the page.

WINGS, Greece

Our main motivation for using the CREW facilities for experimentation was to be able to obtain results through actual devices operating in real environments, instead of limiting our assessment to simulation-based approaches. Thescenarios and concepts of our experiments originate from the OneFIT project which elaborated on the issue of opportunistic networks by capitalizing on device-to-device communications in order to achieve coverage and capacity extension of the infrastructure.

We had the opportunity to conduct various experiments in order to assess the performance of D2D networks under real conditions of operation of wireless networks. We have also benefited from the fact that we have learnt how to use the 802.11s protocol along with its functionality and capabilities, in order to create an opportunistic/D2D network, knowhow we would not have acquired without CREW’s experimentation facilities.

Go back to the top of the page.

UTH, Greece

For the purposes of the proposed experiment, we exploited the advanced spectrum sensing functionalities and the available sensing solutions that are provided by the CREW Testbed Federation. Among the 5 individual wireless testbeds that constitute CREW, we decided to use the w-iLab.t indoor testbed, as it offers all the required hardware and software components for the development of the proposed framework. More specifically, w-iLab.t features several sensing devices spanning from commercial sensor and Wi-Fi nodes, to SDR platforms and device prototypes. Another basic characteristic that motivated to use the w-ilab.t testbed, was the adoption of OMF as its testbed control and management frameworkenabling us to build the proposed framework as a plug-in compatible with the well-adopted OMF.

Our participation in CREW, as Open Call 2 experimenters, provided us with direct access to all the required hardware and software components, along with valuable support by the w-ilab.t testbed team. Through the development of the proposed framework, its deployment in a large-scale cognitive testbed and the successful execution of extensive remote experiments under realistic conditions, we verified the success of our efforts towards enabling cognitive experimenters to apply the developed tools and evaluated the performance of their solutions in terms of the Spectrum Sensing Delay and Energy Consumption metrics.

Go back to the top of the page.

Start of OC3 experiments

Tass, Belgium

Cognitive networking in heterogeneous networks is a concept that relies on interaction among different layers in the protocol stacks. A wireless mesh network of low-power sensor embedded devices can support a complete TCP/IP stack for remote socket communication. The CREW testbed, consisting of 150+ nodes, is an ideal platform to deploy and experiment with the mesh networking capabilities of a cognitive networking system. The sensor boards in the CREW wilab-t Office in Ghent are interconnected with IEEE 802.15.4e low-rate wireless interfaces. Using a modified device driver to control those interfaces, the TCP/IP stack is updated periodically with RSSI information on the physical link quality, which is taken into account by the OLSR protocol when determining the optimal routes.

Go back to the top of the page.

UKIM, Macedonia

Radio Environmental Awareness (REA) is an integral part of the newly emerging wireless communication systems, whereby such awareness promises to enhance the utilization of spectrum worldwide and to facilitate associated spectrum management techniques. In the generic context, REA is composed of a properly structured spectrum database, pervasive sensing and distributed computation in order to provide the requisite self-reconfiguration and improved access capabilities of the underlying radio system.

This Spectrum sensing supported and enabled Increased Radio environmental awareness (SIRI) experiments to evaluate key distinct features of the spectrum sensing process (e.g. primary user detection, sensing complexity, sensing agility and swiftness, cooperation/spatial diversity gain etc.) by exploiting the possibilities provided by the CREW federation facilities. The SIRI experiment targets the evaluation and extensive analysis of other REA enabling techniques, such as single/multiple source localization, Radio Interference fields estimation, indoor/outdoor propagation model estimation and evaluation etc.

Go back to the top of the page.

KULeuven, Belgium

The goal of the FACT (Furthering Airborne Cognitive Technology) project is to improve the performance of airborne ad-hoc networks. Currently, in this CREW Open Call 3 experiment we are investigating the interplay between heterogeneous networking technologies (802.15.4, 802.11), controlled mobility and cross-layer optimizations for 2D mobile ad-hoc networks. The goal is to optimize the trajectory and other link parameters to meet certain requirements (e.g. latency and throughput) under given constraints (e.g. energy). One cross-layer improvement under investigation is using multiple 802.15.4 channels simultaneously in a cognitive way to mitigate interference and congestion. We are using wireless nodes, including 2D mobile nodes, provided by the iMinds w-iLab.t testbed. The use of the testbed, offering a controlled test environment, allows us to obtain repeatable results. Furthermore, CREW greatly lowers project level equipment investment requirements by sharing a pool of hardware and software resources with the entire community.

Go back to the top of the page.

Paris Descartes University & Paris-Sud University, France

Cognitive radio networks permit to increase the network capacity by using opportunistically the unexploited bandwidth of the licensed spectrum. Nevertheless, the distributed spectrum access leads to several problems which are hard to solve with the current technology. Indeed, all spectrum access schemes require a common channel to share the information used by secondary users to coordinate their access. Such a problem is known in literature as blind rendezvous problem, since devices need to decide independently the control channel, maximizing the probability of choosing the same channel.

Our first motivation for using the CREW facilities has been the possibility to test the protocols that we devised for solving the distributed access spectrum problem on a large scale wireless testbed with the possibility to easily repeat the experiments. We were pleasantly surprised by the facility to set up and perform the experiments with respect to other remote open testbeds. Indeed, the system used to control the experiments is well integrated with the testbed and permit to quickly configure the network scenario, thus speeding up the experimental activity.

Go back to the top of the page.

Technical University of Cluj-Napoca, Romania

The experiment aims to validate a distributed interference-aware power control algorithm motivated by Game Theory (GT). In this respect we perform a thorough analysis of the occurence and stability of several equilibria, including Nash, in multi-player scenarios. Although a large number of GT-based models for CR networks have been proposed and theoretical and computational frameworks abound, there are still few experiments validating their potential.

As experimentation facilities play an important role in the transition from concept to implemented technology, we find CREW highly beneficial. The experiments are designed to run on the LOG-a-TEC CREW experimentation facility. This facility offers both outdoor and indoor configurations, it is simple yet flexible enough to accomodate our scenarios. It enables accurate control and mapping of the game as it is based on custom made hardware platforms. The access to the testbed is open and web based (remote access is provided). The tools are well documented and the overall support is very good.

Go back to the top of the page.

Televic, Belgium

Wireless roaming and handover for mobile devices is becoming more and more important for applications in healthcare and conferencing. In the CREW project, Televic is developing a test setup that enables fast evaluation of handover features of different mobile devices. The goal of these tests is to get a better insight in wifi roaming on smartphones. We want to test how existing smartphones roam across wifi networks with multiple access points. The outcome of the tests will be recommendations on what type of mobile device has the best roaming performance.

Go back to the top of the page.

AED, Germany

Within our R&D compartment we developed and implemented a coordinator for wireless sensor networks based on the IEEE 802.15.4 standard. In contrast to already existing coordinator solutions our coordinator is able to handle more than one personal area network (PAN) simultaneously while using several channels within each PAN. It is therefore a multi-channel transceiver based on SDR-technology which enables larger networks comprising more nodes and larger bandwidth. The coordinator is able to use different channels for data transmission simultaneously within one PAN based on energy detection and LQI (link quality indication). It is also able to split one PAN autonomously into several smaller ones due to enable prioritizing or to re-balance the quality of service (QoS).

To evaluate the cognitive behavior of the coordinator when interference occurs it is necessary to be able to restore a signal situation within a test bed for several successive tests without unwanted variations of the signal availability. Based on interference the coordinator adapts the use of channels for its nodes and it can clearly be estimated in which channel the interference occurs and how the coordinator reacts. In total three scenarios are planned to evaluate the AED coordinator.

Go back to the top of the page.

Awards

• TUB won the best Demo at EWSN-2014 in February
• TCD's Iris Based Spectrum Wars Demonstration won best Demo at DySPAN-2014 in April.

Go back to the top of the page.