Lukket

MSc Level Dissertation on ''GATHERING MESH NETWORK KNOWLEDGE''

Abstract : The project will carry out some variation of :

Survey the protocols and tools used to monitor network topology, size and traffic patterns in the infrastructure networks of today's Internet.

Study the differences between infrastructure networks and wireless mesh and sensor networks.

Propose new protocols to support efficient network knowledge capture and network management in mesh and sensor networks.

The entire Dissertation approximately has to be for around 10,000 - 12,000 words including the Initial research and the results etc. The java coding will be submitted separately than the report not to be included in the word limit.

## Deliverables

Abstract : The project will carry out some variation of :

Survey the protocols and tools used to monitor network topology, size and traffic patterns in the infrastructure networks of today's Internet.

Study the differences between infrastructure networks and wireless mesh and sensor networks.

Propose new protocols to support efficient network knowledge capture and network management in mesh and sensor networks.

Implement new protocols and show by experiment (e.g. simulation or emulation) the advantages (and possibly disadvantages) gained from using the new protocols.

The study will involve massive simulations to be run using a pre-defined software called JIST/SWANS details of which are all mentioned in this website <[url removed, login to view]> . The jar files and the user documentations can be downloaded from there too. The javadoc help file which explains all the packages involved and their use in the simulation is given here : <[url removed, login to view]>. This can be used fully for understanding all classes and the packages involved in the simulation.

Following are the steps to be followed from start:

1. Create a wireless world/environment (1000m * 1000m) with two nodes at a distance of 10m able to exchange packets with each other. (It can be in the form of phone calls too)

2. Study the call cost, duration, throughput and all other variants possible.

3. Gradually increase the number of nodes and the calls between them.

4. Start making random calls between nodes, add noise and congestion on the go.

5. Again note down the call cost, duration, throughput and all other variants possible.

6. Compare all these variants by making graphs and keep noting the packet delays, losses and the routing and congestion algorithms.

7. Collect all data and create lots of graphs for all the above comparisons and state in results which way is to be the best possible taking into consideration the cost too.

Some more additions recommended by my supervisor recently:

1. Find out how to generate random bursty internet style traffic to

imitate http, email, ftp traffic. Decide what parameters need

controlling in the generator.

2. Find out how to generate VOIP traffic. Decide what parameters need

controlling in the generator.

3. Decide what statistical data can sensibly be collected, especially

from the routing/network layer but also from transport and MAC.

The entire Dissertation approximately has to be for around 10,000 - 12,000 words including the Initial research and the results etc. The java coding will be submitted separately than the report not to be included in the word limit.

Evner: Ingeniørarbejde, Java, MySQL, PHP, Software Arkitektur, Software Testning, System Administrator, Teknisk Forfatning

Se mere: what are graphs used for, what are graphs for, what are algorithms, using graphs in java, understanding algorithms, study of algorithms, steps on making a website, steps on how to create a website, number algorithms, network world, my edu world, msc tools, msc software, making graphs in java, making algorithms, mac tools, level two, level s, level 3 support, java graphs

Om arbejdsgiveren:
( 0 bedømmelser ) India

Projekt ID: #3317784