|Presentation||People||Projects||Publications||News||Collaborations||Intranet||Spin-off||EIT ICT Labs||Contact us|
The WiMAX (IEEE 802.16e) standard includes the necessary features for supporting mobility while providing similar broadband multimedia services as fixed wireless. Due to the use of orthogonal frequency division multiplexing (OFDM) technology in these systems, the transmissions are scheduled using a time-frequency bidimensional matrix. This fact opens the possibility of exploiting frequency selectivity by mapping traffic flows with high transmission rate to subchannels with favorable channel conditions. In a scenario with full mobility, the use of a channel-quality indicator channel (CQICH) as defined by the WiMAX standard is not enough to allow a channel-aware radio resource allocation. Therefore, we proposed the use of a dynamic channel predictor to provide the required channel information to a cost-based scheduler in order to improve the overall throughput and spectral efficiency of the system.
Most of the ad hoc networks studies and implementations are done on the Wi-Fi radio interface. Wi-Fi is very popular and provides cheap devices for basic ad hoc networks installations. This has contributed to the high success of ad hoc networks such as the Freifunk networks in Berlin. However, ad hoc networks are still far from industrial applications because of the weakness of the Wi-Fi radio interface. To cope with this problem, we used the WiMAX standard, which is both more reliable than Wi-Fi and also offers more guarantees in terms of quality of service.
WiMAX uses a central way to communicate between devices (access point mode) that are not compliant with ad hoc networks. We are working on developing a peer-2-peer optimized communication mode. This optimization comes from the cross layer communication between routing and medium access. The topology information of the network graph is used to apply a multi-constraints optimization between reducing interference on communication and the control overhead information.