Robust Geographic Protocols and Services

Geographic protocols are very promising for wireless ad hoc and sensor networks. These protocols take advantage of the information of node locations to provide higher efficiency and scalability. We study geographic protocols to assess their robustness to non-ideal conditions corresponding to the real-world environments. Our current studies cover protocols that provide basic functions at the network-layer such as geographic routing, geocasting, and geographic-based rendezvous mechanisms. Several robustness issues are considered, such as lossy wireless channels, location inaccuracy, obstacles, node failures, and mobility. Our results show the discrepancy in performance when the protocols are evaluated in non-ideal realistic environments compared to the common ideal environments typically used, revealing the inadequacy of the current protocols to be implemented as-is in real-world. By introducing simple fixes and strategies that take these conditions into account, we show that we are able to improve the performance significantly.

PUBLICATIONS

M&M: Multicast-based Architecture for Mobility Management

This project investigates problems of IP mobility management and efficient handoff for wireless (cellular and all-IP) networks. It addresses issues of address assignment, micro-mobility, handoff performance for mobile nodes.

PUBLICATIONS

AQM-Marking

PUBLICATIONS

VINT/NS Virtual InterNetwork Testbed

VINT is a DARPA-funded research project whose aim is to build a network simulator that will allow the study of scale and protocol interaction in the context of current and future network protocols. VINT is a collaborative project involving USC/ISI, Xerox PARC, LBNL, and UC Berkeley.

PUBLICATIONS

PIM

PIM-SM (Protocol Independent Mutlicast-Sparse Mode) maintains the traditional IP multicast service model of receiver-initiated membership. It uses explicit joins that propagate hop-by-hop from members' directly connected routers toward the distribution tree and builds a shared multicast distribution tree centered at a Rendezvous Point, and then builds source-specific trees for those sources whose data traffic warrants it. The protocol is not dependent on a specific unicast routing protocol and uses soft-state mechanisms to adapt to underlying network conditions and group dynamics.

PUBLICATIONS