Anandaraman, Shanmuga Sundaram. Distributed spectrum coordination for multi-radio platform co-existence: an experimental study on the orbit testbed. Retrieved from https://doi.org/doi:10.7282/T3668DJ6
DescriptionThis thesis presents an experimental investigation of algorithms for protocol-assisted spectrum coordination of multi-radio platforms in a dense radio environment. With increasing proliferation of new wireless technologies and radio standards such as 802.11b/g, Bluetooth, Zigbee, UWB, WiMax etc, multi-radio devices such as laptop computers, cell phones and PDA's will need to co-exist in shared unlicensed frequency bands.
The common spectrum coordination channel (CSCC) protocol has previously been proposed as a method for nearby devices to exchange spectrum usage and traffic information necessary to execute decentralized co-existence algorithms. This work focuses on the application of CSCC to dense deployments of multi-radio platforms with both 802.11 WLAN and Bluetooth in a typical office/SOHO type environment. Distributed spectrum coordination algorithms listen to these CSCC announcements from radios within range, and back off their transmission parameters to avoid contributing excessive interference. We have developed a set of distributed coordination algorithms, with the objective of achieving efficient co-existence between interfering radios while maintaining acceptable QOS (Quality of Service) at every node.
Specific coordination algorithms considered include Bluetooth defer-transfer (Bo), Simple Source Rate adaptation (Rt), distance based SIR link budget rate adaptation (SIR-BT). Each of these algorithms is defined and evaluated using dual-radio nodes on the 400-node ORBIT radio grid. System performance parameters obtained from the experiments are throughput, file transmission delay (for TCP) and quality of data/audio/video streams (for UDP).
Experimental results are given for a number of device densities and topologies. Significant degradation in throughput and application performance is observed without spectrum coordination. The proposed CSCC-based coordination algorithms are shown to provide significant performance gains, both in terms of system throughput and application level parameters. Overall, for the scenarios considered, the proposed coordination algorithms provide ~50-100% improvement in system throughput when compared to the case with no coordination.