Joseph Bibb Cain

A link scheduling and ad hoc networking approach using directional antennas

There is strong interest within DoD to utilize high-gain, directional antennas at both the transmitting and receiving end of the link in a dynamic, ad hoc network environment. However, the application of directional antennas (e.g., phased-array or sectorized antennas) in a dynamic network of mobile nodes requires coordination of antenna steering at both the receiver and transmitter ends of the link. Our solution is to apply adaptive, link-state routing (be performed by the OLSR ad hoc routing protocol [T. Clausen et. al., September 1, 2001]) supported by a distributed, adaptive time division multiple access (TDMA) scheduler, which determines schedules based on cooperative decisions between each pair of neighbor nodes. The architecture that has been developed contains a high rate mission data channel with an adaptive TDMA link scheduling protocol designed to take advantage of high-gain directional antennas. Time slots on this channel are adaptively scheduled to meet dynamic traffic demand requirements and to avoid interference from adjacent transmitting nodes. In addition, the link scheduling protocol must adapt to changes in node neighborhood topology caused by node mobility and link obstructions. This architecture was tested and evaluated in two ways: by OPNET simulations and by field demonstrations.

An adaptive link assignment algorithm for dynamically changing topologies

An adaptive link assignment algorithm for the distributed optimization of dynamically changing network topologies is presented. The algorithm is responsible for determining the network connectivity by controlling the selection of links to be established and disconnected. This algorithm is designed to recover from predictable link outages as well as massive unpredictable failures. To minimize computational time complexity as well as to improve transient response. Some known graph-theoretic algorithms are utilized. >

Interactions Between Routing and Flow Control Algorithms

In this paper we discuss the interactions between routing and flow control algorithms for packet-switched networks. We present a new variation on the Gallager-Golestaani flow control scheme [4] and also a new variable window scheme. The tradeoff between throughput and network congestion is examined. It is shown that these two flow control schemes have the unique feature that the parameters can be specified to place an upper bound on the expected amount of network congestion.

Improved probabilistic neural network and its performance relative to other models

This paper presents a new extension of the probabilistic neural network which utilizes one additional training pass to obtain significantly improved performance relative to the conventional probabilistic neural network. In addition it automatically sets certain algorithm parameters. The method substantially outperforms K-nearest neighbor techniques for the same number ofnodes. and it also offers performance competitive with LVQ2 which requires much longer training periods. 1.

A class of adaptive routing and link assignment algorithms for large-scale networks with dynamic topology

The authors present novel algorithms for adaptive routing and link assignment, which can be used in a large-scale network (1000-10000 nodes) with dynamic topology and stress. Special attention was given to performance (delay, throughput, and survivability) and implementation cost issues as the size of the network grows. These costs are controlled by a logical partitioning of the network into areas. The proposed class of algorithms addresses the following key issues: maintenance of up-to-date routing information in the presence of dynamics, robustness to failures that can partition areas, optimization of routing algorithm performance through multiple path routing, and optimization of network topology for performance metrics of delay, link performance, and connectivity.<<ETX>>