Single Snapshot Exploratory Approach For Visualizing Very Large Network Technologies.

Single Snapshot Exploratory Approach For Visualizing Very Large Network Technologies

Introduction

The networks nowadays are growing rapidly. This explosive growth makes manual network analysis practically impossible. Hence keeping in view these problems we have developed a method of displaying the entire network in a single snapshot, So that the network administrator can have an overview of the network.

Motivation

The network visualization on which we are working provides a single snapshot view of the network infrastructure. The display is re-adjustable and updates itself in accordance with the changes in network. Hence the network managers know that what went wrong and where.

Problems Faced

Three problems are most commonly encountered for large topology networks

Display Clutter Node Positioning Perceptual Tension

Proposed Solution

The entire area of screen is divided into layers.

Nodes are positioned within these layers. We use color codes to define the type of

machine being displayed as a node. SVG is the preferred technology for the job.

SVG to Display Topology Discovery

SVG is used due to a number of reasons as its features fulfill all our display requirements.

Plain Text format Scalable Zoom able Searchable and Selectable Text Scripting and Animation Works With Java Technology Open Standard True XML

The Algorithm (at 800X600 Resolution)

Area of layer = n (16 X 12) No. of nodes in n (16 X 12 px) = n (4)

where n=1, 2, 3, 4,5… NL (n=layer number)

Maximum number of nodes that can be displayed = l = 1+∑lan = 1 + ∑n (4) n=1, 2, 3, 4, 5…NL. (n=layer number)

Inner Most Layer View (at 800X600 resolution)

Sample Visualization View (at 800X600 resolution)

Results

Hence according to the above calculations the following results can be derived:

The inner most layer contains 4 nodes The outer most layers that is the 50th layer

contains 200 nodes the maximum number of nodes that can be

placed in 50 layers is 1+∑lan

Conclusion

In this paper, we have discussed a technique for network visualization with minimum possibility of a cluttered display. The challenge was to achieve a clear and comprehensive view of the network, so that the network managers of vast telecommunication networks can have idea of their network. It is practically almost impossible to manually examine such networks node by node.

For this purpose the best choice was SVG. Graphics in SVG can be zoomed in for better viewing and there is negligible corrosion of the image. Further unlike raster images it does not produce stair case effect while printing images. One very major reason for using SVG was its compatibility with java. Being XML based SVG has vast compatibilities.

QUESTIONS?