Wireless mesh networks are becoming popular for ubiquitous and low-cost wireless broadband connectivity. However, they, suffer from serious interference problems which limit their capacity. We use TDMA scheduling to address these performance problems. We use a SINR-based physical interference modeling for realism, and various forms of diversities - such as transmit power control, directional antennas, multiple channels, and rates - to maximize performance. We also introduce a measurement-based modeling technique to make the physical interference modeling practical. We support these innovations by experimental studies. Our goal is developing a complete set of algorithms, protocols, and system solutions that target "managed" mesh networks. The protocol solutions include the routing layer and below, and are compatible with existing Inter-networking protocols for the transport layer and up.
The project's intellectual merit has the following components: (i) New protocols for TDMA scheduling with diversity and physical interference modeling, (ii) integration of scheduling with routing, (iii) measurement-based modeling of physical interference, and (iv) simulation modeling and testbed experiments to support the protocol innovations.
The project contributes to the education and training of graduate students in the two universities -Georgia Tech and Stony Brook University. The project gains leverage from existing international collaborations of PIs with IIT-CNR in Italy and CDAC in India, and strengthens this collaboration. Finally, success in this project means low cost, ubiquitous broadband connectivity. It competes well with WLANs and wired connectivity, and is expected to level the technological playing field.
