uwicore. Ubiquitous Wireless Communications Research LaboratoryUwicore





The project activities will be conducted using several software and hardware platforms that have been developed over the previous years by the project team members. The use of such platforms within the project provides a competitive advantage, and represents a very valuable asset to the viability of the project proposal. In any case, it is important to note that continuous efforts (planned in WP4) will still be necessary to update and evolve the research testbeds according to the project needs.

Realt-time testbed for field testing of Multi-Hop Cellular Networks using Mobile Relays

The project members have developed, to their knowledge, the first MCN-MR testbed developed to investigate the performance improvements that can be achieved through MCN networks using mobile relays, and the operating conditions under which such improvements can be achieved. The platform, illustrated in the figure below, currently operates using the Orange HSDPA/HSUPA network with 802.11 ad-hoc links between mobile relays. In addition to two cellular links (one as part of the MCN-MR connection and another traditional cellular link to compare its performance against that achieved with MCN-MR) and a scalable number of 802.11 nodes, the platform also includes the necessary monitoring tools to analyse the end-to-end performance and GPS units to synchronise and geo-reference all measurements. In particular, the platform incorporates Nokia 6720c handsets with the powerful Nemo Handy monitoring tool, and 802.11 nodes (there is a plan to migrate them to handheld devices and/or 802.11n when the necessary linux drivers are available) with the linux Ath9k controller. The nodes incorporate several 802.11a/b/g/n wireless cards to be able to simultaneously transmit and monitor the performance of the mobile relay transmissions; such monitoring is conducting using the Wireshark and Kismet software tools.

An example of the measurement capability of the MCN-MR connectivity testbed is shown in the figure below. The field tests evaluates the capacity of MCN-MR to improve the cellular link quality, and thereby the end-user QoS, when operating under NLOS conditions. To this aim, the performance of a traditional cellular link with NLOS conditions with its serving BS (L2) is compared to that achieved with a MCN-MR link operating under LOS conditions through various hops (L1-mH1-mH2). The results obtained showed that MCN-MR could improve the average performance of a traditional cellular link by a factor between three and four.

The MCN-MR field testing platform will clearly represent a key asset for the project, and will be extended to include more reliable 802.11 technologies, such as 802.11n, in addition to integrating the opportunistic operation investigated within this project.



IEEE 802.11 Programmable Software Defined Radio Experimental Wireless Communications Research Testbed

A second hardware testbed available at the Uwicore laboratory and that will be extended and used within this project is a 802.11 SDR hardware testbed implementation using the USRP2. The platform includes the first 802.11 MAC implementation for the USRP2 platform (the implementation was conducted during the preceding mHOP project), including its DCF function based on CSMA/CA, a channel sensing function, a back-off process, and the RTS/CTS mechanism. The MAC implementation has been integrated with a 802.11 emitter implementation done at the FTW research institute in Austria (currently, only an emitter implementation is available, although several research teams are working on the receiver). The USRP2 802.11 platform will be extended with new functionalities to test advanced opportunistic and adaptive cross-layer transmission techniques to be investigated with WP2. The following figure depicts the USRP2 nodes, and a state machine diagram of the 802.11 MAC implementation conducted at the Uwicore laboratory.



Mobile Relaying in Multi-Hop Cellular System Simulator

To investigate the performance of multi-hop communications in MCN-MR, the project team members are developing a urban ns2 software simulation platform based on 802.11 technologies to emulate mobile relaying communications. The simulator incorporates accurate radio propagation models to adequately model the challenging peer-to-peer communication conditions. The simulator does not currently model cellular transmissions, although it can interact with the SPHERE cellular simulator described in the following section. In addition, the project members are working to integrate cellular connectivity and QoS traces obtained from field tests into the ns2 platform as an alternative to precisely model the transmission in cellular systems. This alternative is being investigated due to the fact that this platform focuses more on the study of the mobile relaying communications. In addition, the platform will be extended during the project to be able to investigate all possibilities and trade-offs offered by opportunistic MCN-MR systems. It is also important to emphasize that the level of modelling and large simulations needed to obtain accurate and valid simulation results requires large computational times. The team will use its small computing cluster, although further extensions will be necessary during the project.



SPHERE Heterogeneous System Simulator

The project also incorporates the SPHERE (Simulation Platform for HEterogeneous wiREIess systems) platform developed at the Uwicore laboratory, and capable to investigate heterogeneous management policies at the system level. The platform is capable to simultaneously emulate the transmission over different radio access technologies (currently HSDPA, EDGE and GPRS), a key characteristic to be able to adequately evaluate CRRM policies in heterogeneous beyond 3G systems. The platform has been evolved to include an API that allows the simulation of different RATs in distributed locations/equipment, which can reduce simulation times, and integrate various simulation platforms (currently working to integrate ns2 simulators). The computational efficient use of the SPHERE platform will also require updates in the simulation cluster of the projectís team.