Computer networking undergraduate project ideas

GothicGhost

GothicGhost

GothicGhost

Soldato
Joined
23 Aug 2007
Posts
2,794
The time has come for me to start my undergraduate project in computer networking but I am also struggling to come up with something to do.

It's only the first week back into university but I am trying to make a start on the project, but thinking about what I can do that has not already been done is harder than I thought it would lol.

Do any postgraduates have any words of advice?
Thank you very much.
 
I did my dissertation on physical vs. virtualised routing on commodity hardware.
 
Depends on what sort kind of thing you're doing.

I personally like load balanced, hypervisor based horizontal scaling web deployments - not sure if that's appropriate or not on your course though.

You could also see if you can do a modern version of this mad-lad's system. The trolling based security measures were a classic:
http://www.ex-parrot.com/pete/upside-down-ternet.html
 
Last edited:
DireEmblem said:
How about an 'authentication' network, that logs material from each participant and seeks out unusual activity?
Click to expand...
lol yeah maybe lol

Cooler running said:
Depends on what sort kind of thing you're doing.

I personally like load balanced, hypervisor based horizontal scaling web deployments - not sure if that's appropriate or not on your course though.

You could also see if you can do a modern version of this mad-lad's system. The trolling based security measures were a classic:
http://www.ex-parrot.com/pete/upside-down-ternet.html
Click to expand...

Thank you.

I guess the hard part finding something that is in the middle ground of not too simple and also not something I would find impossible to complete in time.
 
Protip. Choose a subject that allows easy comparisons and therefore clear path to generate data.

I did networking 15+ yrs ago at uni. My fyp was on Qos protocol efficacy with voip (which was hot at the time) as the payload.
 
GothicGhost said:
Do any postgraduates have any words of advice?
Thank you very much.
Click to expand...

Never went to uni, but here's a few ideas for you to get your teeth into;

- Application driven routing, where the user application decides it's own routing path through the network based on latency or performance, as opposed to traditional routing protocols (IGP/BGP). (look into segment routing, also ipv6 segment routing header)

- Building an open-source internet router on a standard server, capable of line-rate performance (10G/100G), using xdp (express data path) or dpdk (dataplane dev kit) (kernel bypass) complete with an open source routing stack. Then do pros vs cons against a traditional internet router like a Juniper MX or Cisco ASR9k or something.

- Come up with, and or / compare different types of IGP routing designs for hyper-scale cloud network fabrics, which solves the problem of IGPs (such as OSPF) breaking due to control-plane load (numbers of routers + LSAs, combined with the number of links, killing the network)
 
[FnG]magnolia said:
This is more amazing.
Click to expand...

Why? Cable comb dressers, for instance, are massively popular. They're essential for ensuring cable bundles don't get tangled. If the OP were to create an algorithm for designing and 3d-printing one with an arbitrary number of cables of arbitrary diameter, that would be incredibly useful. The Acom comb copes with 48 cables. That's small beer: what about 100? 200? What about dressing fibre cables? Actually, I expect it's already been done, but the OP is in a far better position to know that.

People are entranced by the sexy stuff but the un-sexy stuff is just as valid.
 
Thank you all for giving me some ideas to research and see what I can do with my current skill level.
Also going over some of the ideas you have mentioned with my adviser and see what direction to go.
Once I know my goal I can work my ### off to hopefully get a good grade.

Thank you very much.
 
You shouldn't necessarily need to come up with anything as you might well find that there are a bunch of academics at your uni with various interesting areas for prospective undergrads or master's students to look at. Some will have them listed on their academic web pages - undergrad and postgrad projects for this year etc.. Even for PhD students, the academic might have some specific funding for a doctoral student who is willing to investigate X etc..

You could just look at possible tutors within your department, see what they're offering, see if it is something you like and/or then suggest some possible modifications too if you have something in mind...

Also, can you do one in conjunction with an external supervisor too - that can be a good way into emplyoment, it is perhaps more applicable to applied projects than purely theoretical/academic ones but if you have some area you're interested in working in then...

Like say you're interested in finance, do any academics at your university have connections with financial institutions? Could they introduce you?

Ditto to anything else like security or whatever. There will generally be some who work on research in conjunction with industry.
 
GothicGhost said:
Also going over some of the ideas you have mentioned with my adviser and see what direction to go.
Click to expand...

Oh, wait you already have an advisor? Can you not choose one then?

I'd speak to them first tbh... don't take ideas that are too specific from randoms in this thread necessarily as they might not be in line with his or her research interests or areas of expertise etc..
 
dowie said:
Oh, wait you already have an advisor? Can you not choose one then?

I'd speak to them first tbh... don't take ideas that are too specific from randoms in this thread necessarily as they might not be in line with his or her research interests or areas of expertise etc..
Click to expand...
I think you over estimate what has become the final year project/dissertation in modern university degrees. Most people will be assigned a tutor and write 15000 words on something dull as dishwater of no real relevance to anything get a grade and move on. I got a first for wiring 15000 words on The use of Linus in SME’s that was 20 odd years ago though so maybe it was a little more ground breaking than if you did the same today!
 
We have to pick our supervisor next week so I am going to ask my networking lecturer if he would be mine.
Hopefully, he will then point me in the right direction.

I see university as a stepping stone to true learning you get from on the job experience, but I need the degree to open some doors for me I guess.
 
GothicGhost said:
We have to pick our supervisor next week so I am going to ask my networking lecturer if he would be mine.
Hopefully, he will then point me in the right direction.

I see university as a stepping stone to true learning you get from on the job experience, but I need the degree to open some doors for me I guess.
Click to expand...

Whatever you do, you'd be well advised to pick something that includes a decent amount of coding and automation, in some way or another. Getting a real job in the network engineering sector is very difficult without good working knowledge of automation and some degree of software engineering.
 
You must log in or register to reply here.
Back
Top Bottom