Agent-based communication in a mobile, wireless and ad-hoc host network

A Project of the D'Agents Laboratory at Dartmouth College

Qun Li, Daniela Rus

An ad-hoc network is formed by a group of mobile hosts upon a wireless local network interface. It is a temporary network without the aid of any established infrastructure or centralized administration. It is required when no infrastructure such as Internet is available to some mobile hosts, or the traditional network infrastructure is undesirable due to the high cost or inconvenience. For example, in some situations of disaster recovery personnel or military troops, the normal network infrastructure may be unavailable or destroyed. Other examples include business associates wishing to share files in an airport terminal, or a class of students needing to interact during a lecture.

We are planning to use mobile agents for the communication in a wireless, ad-hoc, and mobile-host network environment. A mobile agent is a program that can migrate from machine to machine in a heterogeneous network. It can be viewed as an intelligent entity that carries the message, as opposed to the plain message data transmitted in a network protocol implementation. An agent may stay in a machine waiting for the connection to another machine, or plan to transport itself to transmit the message in a changing, disconnected environment.

More specifically, our work here is to research how to use mobile agents to maintain the communication in a disconnected, ad-hoc network. We explore how the information about the motion of the destination host can be used to determine how the message can be sent by the cooperation of the intermediate hosts.

The field experiments will be done in the future. We also plan to apply the same strategy in the strategic robotic networks.

Results:

We have studied two algorithms, which are presented in the following paper.

1. In the first algorithm, we assume the information about the motions and locations of hosts is known to all hosts, or can be estimated with some error parameters.

2. The second algorithm doesn't assume the movement of the hosts is known. We use the Brownian movement to characterize the random movement of hosts. The algorithm gives the strategy on how to maintain network connectivity by location update.

Relevant Dartmouth papers:

Transportable information agents: Daniela Rus, Robert Gray, and David Kotz. Journal of Intelligent Information Systems, 9:215-238, 1997.

Agent Tcl: Targeting the needs of mobile computers: David Kotz, Robert Gray, Saurab Nog, Daniela Rus, Sumit Chawla, and George Cybenko. IEEE Internet Computing, 1(4):58-67, July/August 1997.

Modeling and analysis of active messages in volatile networks: C. Okino and G. Cybenko. In Proceedings of the 37th Allerton Conference on Communications, Control and Computing, Monticello, IL, 1999.

Sending Messages to Mobile Users in Disconnected Ad-hoc Wireless Networks: Qun Li and Daniela Rus. In Proceedings of the Sixth Annual International Conference on Mobile Computing and Networking, Boston, August, 2000. ACM Press.

Relevant projects, inside D'Agents:

MURI demo

Relevant projects, outside D'Agents:

H. T. Kung's Lab at Harvard

Network Lab at RPI