Hi..!

i m working on very simple agaent and manager of SNMP. i want to comapre DES and AES cryptographic algorithm that i will use in communication being carried out between agent and manager.

i want to know that which paramter i should select through which i can do this comaprison (that which technique is better for secure communication?)

for example the paramters like CPU USAGE, SYSTEM UP TIME, NO.OF PROCESS, MEMORY CAPACITY, STORAGE CAPACITY etc...
can i use these paramter for comparision? i mean by using both (AED & DES)techniques for secure communication if find out these parameters, will these help me to analyze that which technique is better?

please help me out i am so confused about these paramters!

thanx
Regards

I've not got access to my library at mo so...

http://searchsecurity.techtarget.com...797792,00.html
http://www.nexiondata.com/products/o...ypt/desaes.htm

There's loads more, see google search

I'd also suggest looking at block modes... http://en.wikipedia.org/wiki/Block_c...s_of_operation

A good reference book is Applied Cryptography: Protocols, Algorithms and Source Code in C

The short answer is "Use AES". DES is completely broken, and Triple-DES is deprecated.

@ brazilnut

thanx.. these are helpful links but actually these are to understand cryptography and i need to know their performance through some paramters.... that i still didnt get!!

tuxdev
yes! AES is much better than DES but for my project i have to compare the results of both.
for example i will note down the time taken to encrypt and decrypt the message by both techniques.

other than time what can i check or note????

Regards!

When you find the source code for your implementation (library) there is usually a benchmark app included so you can test (and check for errors).

To understand the performance you do sort of need to have some understanding of crypto. For instance if you get a data error (say one bit is wrong in one block) then using certain block modes it may be impossible to recover all blocks after, or you may just lose one block, etc (I provided a link above which explained that).

Certain implementations of the code are faster than others either because of the language or just the way it was written. Some versions have optimised code included for specific processors, etc...



Here's some more google results for ya...

http://www.cryptopp.com/benchmarks.html
http://forums.sun.com/thread.jspa?threadID=558004
http://gladman.plushost.co.uk/oldsit...esr1/index.php


What library are you using?

Thanx! for your help.!
and for these useful links.

well i dont know about library. as i have simple client and server application and
i just have to secure there communication and find out the results.

i m doing it in simple C language on Linux (Fedora) platform.

Regards.

It's completely and utterly *pointless* to even consider broken and deprecated algorithms. The other AES candidates like Twofish or Serpent are reasonable to compare with, not DES or Triple-DES. You're not really learning anything when you're stuck on such algorithms.

Tuxdev is absolutely correct:
Strike 2:
So you're not even evaluating the algorithms themselves - you're just timing two different implementations (and apparently not even interested in examining the code).

Good luck!

IMHO .. PSM

PS:
Suggestion:

ahaaan... i understand what u people are saying...and thankful to you as u r giving me your time but my problem is that my project topic is approved by my teachers. and the topic was secure simple network management system. in SNMP DES is already implemented even 3DES and AES.
when i told my supervisors they asked me to do analysis of both algorithms. because they can not change my topic.... so its just my department's head's requirement :-( i cant do any thing so just trying to do what they assigned. well i will try my level best to do something different regarding this.

hope you understands my problem.

and again thanks for nice suggestions.

Regards!

As for a specific implementation, OpenSSL is pretty much the top open-source C solution out there.

Some more research for you...

http://tools.ietf.org/html/rfc3826
http://tools.ietf.org/html/rfc3414