Because of the nature of wireless networks, accurate evaluations can be quantified only through real experimentation. Experimental repeatability and controllability issues are largely overcome by the unique setting of QuRiNet, virtually free of external interference (neighboring wireless networks) and unpredictable fading environments (people/vehicles/doors). Current areas of research include:
- Layer 2 Routing
We would like to forward traffic at the MAC layer in order to avoid excess processing time and addressing required by current routing protocols. We are interested in finding ways to intelligently select paths for packets at the MAC layer.
- Monitoring and Maintenance Tools
With access to the Quail Ridge Test Bed, we would like to design and develop wireless mesh network monitoring and maintenance tools that would be useful to future mesh network administrators and researchers.
- Time Based Access Mechanisms
Current 802.11 mesh networks must utilize CSMA/CA in order to access the channel. With a large number of users, it may be more efficient to organize access based on time. We are interested in applying TDMA schemes over 802.11 in order to boost performance.
- Exploiting the Multiple Channel Advantage
The access points we have chosen for Quail Ridge are equipped with two radios. This will allow us to intelligently and dynamically select channels in order to avoid interference on high-traffic links.
- Mobility
We will study the impact of mobility using a set of fine-grain metrics, including signal strength, its variation, BER (Bit Error Rate), PER (Packet Error Rate), queue length, back-of window size, and packet drop rate. We will include the impact of different mobility types - low mobility (hand-held), medium mobility (Gators), and high mobility (vehicular) - in various geographical topology settings.
- Security
We will study unique wireless security features with a focus on physical layer security, namely, secret key generation and signal-print-based authentication. Generating a secret key between two parties by extracting the shared randomness in the wireless fading channel is an emerging area of research.
- Network Heterogeneity
With the fast proliferation of wireless devices, dense co-existence of heterogeneous wireless technologies is inevitable. For instance, B3G (Beyond 3G) cellular service is often envisioned to integrate different communication technologies such as cellular, WiFi, WiMAX, and cordless. Network integration allows users with multiple radio interfaces to intelligently decide the most appropriate network/technology on the fly based on a variety of factors, such as channel condition, traffic pattern, network congestion, and reliability.
- We plan to develop a framework to quantify and analyze the integration gain of heterogeneous networks. We have identified three contributing factors: spatial multiplexing (different networks have different coverage, data rate, and cost), multi-network diversity (a user with multiple wireless interfaces has multiple options), and multi-user diversity (a wireless network with multiple users can opportunistically schedule users with relatively good channel conditions while guaranteeing the performance of each user).
- We will develop distributed schemes to harness the benefits of network integration, analyze average integration gain, and develop upper and lower bounds for it.
- We will evaluate the modeling and analysis using experiments in the QuRiNet infrastructure.
- Another property of the heterogeneous network architecture is its resilience to network failures. We plan to evaluate the reliability of the integrated network through analysis and experiments at QuRiNet, and study the impact of an integrated macro-micro scale approach that consists of network deployment, multi-hop relay, and back-up route planning.