[distruction]

Hi, There


This is my First Post So bare with me.

I got this End of Year Software Project and i am interested in doing something different than others,

Basically what i had in mind is building a firewall with the use of some AI, so that the Firewall can Learn and make decisions on its own, this is a rough idea for now.

What i am interested in is, if it is possible to build, and if so pointers on how to go about the AI part, I am not asking for code i am just interested in your way of going about the problem.

Hope you can help

Regards.....

This post has been edited by distruction: 30 July 2011 - 11:10 AM

[Brewer]

That's a pretty broad question. First, if you don't mind, please answer some questions for me. How much studying have you done on either of these topics (artificial intelligence and computer security)? Have you taken a course in both? What other projects have you done in these fields?

[macosxnerd101]

Moved to Software Development.

It's obviously possible as many firewalls exist. However, to be honest, people who usually have no ideas on where to start with projects of this scope usually don't get very far with them.

I found this tutorial for developing a Linux firewall. Hopefully it will get you pointed in the right direction.

[distruction]

Brewer, on 30 July 2011 - 12:38 PM, said:

That's a pretty broad question. First, if you don't mind, please answer some questions for me. How much studying have you done on either of these topics (artificial intelligence and computer security)? Have you taken a course in both? What other projects have you done in these fields?

I am currently doing a National Diploma in Computer Science i don't have a lot of experience in AI, I Have been looking at
at a lot of forums but can't quite put together the two ideas and that is what i was looking for here someone to help me understand how to put AI concepts with Security.

I have Ethical Hacking Experience but none in AI, hope this Answers your questions.

The ideas for the Firewall is

To identify all outgoing and incoming communication from the computer and learn that some connections are OK and others have to be looked into by the user

Scan for open ports and closes them if the connection is suspicious
(This i don't know how to go about identify suspicious behavior and not)

Instead of having the user to block traffic on certain TCP/UDP ports, the program will be able to know what ports to close and which once to leave open. (This will for example leave open ports used in P2P connections and close others)

these are the basics for now and will evolve the program as i go along.

Thanks for the Replies