A Distributed Cluster-based Emulation Test Bed for Large Wireless Communication Networks

In this proposal, we propose an innovative technical approach for developing a real-time scalable emulation test bed for large wireless communication networks. Our approach has four main advantages: 1) it uses radio model with configurable fidelity levels (packet level and detailed signal level) according to user’s preference; 2) the emulation architecture is scalable to support large wireless networks. We adopt clustering technique to manage up to thousands of radios in the emulation network in real-time. All wireless nodes are clustered into multiple groups, and each group and the communication activities within the group is managed by a single cluster server. Inter-cluster activities are handled by cooperation of different cluster servers; 3) the emulation network also works with user interactions and environmental data, such as weather and terrain data; 4) it provides both perceptual and statistical evaluation results. Users can connect real communication terminal devices into the emulation network and run any type of network applications, including audio/video streams, and evaluate those applications by the actual usage experience.