network security research

I am a masters student in computer science. I am interested in
Network Security. I am currently working (research) on network
security of wireless networks, to be more specific Denial of service
attacks on 802.11 networks.

I would greatly appreciate if anyone could please tell me if this
topic is of practical significance in the industry today or is there
some other related specific focus area that would be much better.

Also I would like to know what are the current hot topics in network
security which are both of research interest and have some
industrial significance.

Thanks,
Jitesh

802.11 security is a very hot topic but not really pertaining to DOS attacks, Think about it with wireless people dont want to do DOS they want to get in without being noticed. I would change it to 802.11 security authentication and encryption methods and how each method is vulnerable to what types of attacks. such as WEP,EAP-TLS,EAP-TTLS,EAP-MSCHAPv2,PEAP,LEAP,WPA-PSK,WPA,TKIP ETC....

Well i am actually working on Dos attack on shared medium which includes 802.11, sensor networks and others. I am looking(reading recently published research papers) at various prevalent dos attacks and intrusion detection mechanisms to understand the domain better. My ultimate aim is to create controlled attacks in shared networks and create analytical model for the same.

Do you think that it is worth the effort. Also comment on its practical significance in industry.

thanks

DOS on wireless networks is actually rather simple, just blast the appropriate wavelength at a high 'volumn'. There is also the well-known DOS in WEP and EAP where an attacker repeatedly fails to authenticate and causes the access point to deny all connections for a period of time. The architects have denied that this is a 'problem' precisely because it is so trivial to DOS the signal. since there are so many 'trivial' ways to DOS a wireless network, I agree that there is probably not a whole lot of point in investing a lot of time and energy in the research. If you wanted to try to develop ways to avoid active DOS attacks you could probably find a lot more interest, though the basic theory makes that unlikely (though the military has invested a great deal in that sort of research if you could get your hands on it).

even easier just plug a wireless phone in next to the access point LOL. Mitakeet is correct though the militarey has invested alot into wireless the reasons are obvious but commercialy it isnt as much interest as just securing their wireless from intruders. I have never been approached on protecting wireless from DOS but have had many requests to secure wireless access.

I know that not much research has been done in DOS attacks. But as per some FBI report for 2003 DOS attacks comprised about 42% of overall network intrusions. Also authentication and encryption are concerned with data security. But even these, place their trust on the underlying infrastructutre which is at risk from DOS attacks.

Isn't this a good enough reason to look at the field.

More importantly, I am quite new to the wireless domian, so could you please explain these flooding techniques in wireless without the technical jargon or Can you point me to some resource that i can refer for the same.

Thanks a lot for all the help

DOS on wired networks is a bit more difficult (though not terribly so) because you actually have to be connected to the network in order to execute the attack. DOS on wireless networks, on the other hand, does not require physical connection and can often be remarkably remote. It is just like a radio. If you are listening to your favorite channel and then there is some broadband burst of radio energy (sun flare, military exersize, lightning strike) then you get a burst of static and lose the signal (denial of service). If someone is doing something intentional, they can save a lot of energy and only broadcast on the particular wavelength you are recieving on. They can do even better with directional broadcast where they focus most of their energy on your reciever. With a little tuning you can probably get something that will run off a car battery for several days and can be disguised to look like just about anything. Put a nice random number generator on it to vary the intensity and duration of the signal and you can create nightmares for the people attempting to use the network. At this point we have a DOS and haven't even approached the protocol of the networking; we are just drowning out the signal. If you take into account the protocol (which, after all, is publicly available to anyone who is interested), then there are additional attacks you can use (either on encrypted or unencrypted networks). Since all these attacks are well known and essentially impossible to defend against, where are you going to contribute to the sum-total of knowlege as require for a graduate thesis? If, by some burst of inspiration, you come up with a way to defend against DOS in a wireless environment (if that happens, expect to get a huge job offer from the military, directly or via a contractor), then you are doing something interesting. I suggest that it will take an extensive background in energy physics, not something typically recieved by CS students. It is your time and energy though, and if you can convince your comittee that it is a worthwhile use of your time then you will get the degree you are there for, so what does it matter?