Agent-based Distance Vector Routing: A Resource Efficient and Scalable approach to Routing in Large Communication Networks Page: 19
The following text was automatically extracted from the image on this page using optical character recognition software:
The corresponding message activity is thus bounded by the number of con-
stituent agents. However, by limiting the number of agents in order to control
resource overhead, the degree of concurrency which the algorithm can employ
is restricted as well. We have conducted a number of experiments to analyze
the performance of an agent-based distance vector routing scheme. In partic-
ular, we have focused on the Instantaneous Routing Load (IRL), scalability,
path-cost convergence, and route discovery of ADVR and have compared the
results with that of DVR. We have also looked at the distributed manipulation
of the agent population in the network.
It was observed that as per our conjecture, the IRL in ADVR is considerably
low and scalable when compared to DVR. It was also verified that the IRL
for ADVR has a very small variation as opposed to DVR which results in
sharp spikes of routing loads for periods of activity. Further, it was verified
that although DVR is aggressively reactive in path-cost convergence and route
discovery, ADVR with a substantial number of agents can compete with the
performance of DVR. To validate our conjecture, a dynamic and distributed
mechanism was set up using pheromones to manipulate the number of agents
in the network in order to reduce the overall protocol overhead.
The results of this paper are expected to provide alternative ways to design
and implement resource efficient routing algorithms. Particularly in view of
the recent developments in ad-hoc and mobile networks, agent-based solutions
to routing may be alluring as the such system are inherently fault tolerant.
While the main objective of this paper is on routing, agent-based solutions are
deemed suitable for many other network centric applications. Network mon-
itoring, for instance, could take advantage of the mechanisms developed as
part of this approach. The dynamic population control mechanisms facilitate
the design of adaptive solutions for monitoring processes or sensors that un-
dergo complex dynamics and cannot rely on statically designed schedules and
itineraries. The distributed control mechanisms described above may help to
coordinate the actions of otherwise autonomous agents to find a global moni-
toring strategy. The management of large networks and distributed computing
environments can take advantage of the mobile agent paradigm and the tools
designed for this research. By exploiting mobility and intelligence, agents fa-
cilitate system fault tolerance through the expedient discovery of redundant
communication paths and/or alternative computing platforms. Resource man-
agement and distributed cluster scheduling in support of scientific applications
in Grid computing may take advantage of such properties. In general, we ex-
pect that this paper and its corresponding results will motivate the design of
agent-based solutions for large scale system-level applications.
Here’s what’s next.
This article can be searched. Note: Results may vary based on the legibility of text within the document.
Tools / Downloads
Get a copy of this page or view the extracted text.
Citing and Sharing
Basic information for referencing this web page. We also provide extended guidance on usage rights, references, copying or embedding.
Reference the current page of this Article.
Amin, Kaizar A. & Mikler, Armin R. Agent-based Distance Vector Routing: A Resource Efficient and Scalable approach to Routing in Large Communication Networks, article, March 25, 2002; [New York, New York]. (digital.library.unt.edu/ark:/67531/metadc111275/m1/19/: accessed November 21, 2018), University of North Texas Libraries, Digital Library, digital.library.unt.edu; crediting UNT College of Engineering.