Re: Sun setting on stream ciphers?

From: Lassi Hippeläinen (lassi.hippelainen_at_welho.compromised.invalid)
Date: 02/28/04 

Date: Sat, 28 Feb 2004 16:15:23 +0200

Mok-Kong Shen wrote:
>
> In http://www.eweek.com/article2/0,4149,1538027,00.asp
> with the title 'RSA Panel: Cryptography Can't Foil Human
> Weakness' there is the following:
>
> In perhaps the only actual discussion of cryptography,
> the Weizmann Institute's Shamir said the sun was setting
> on stream ciphers used to encode real-time data streams.
> Instead, the power of today's microprocessors could be
> used to encode data in blocks via block ciphers, which
> are more powerful but require a large amount of
> information to be buffered and then encoded.

I disagree about two of those arguments.

First, whenever processors have gotten faster, the applications have
consumed that speed ("Moore giveth, Gates taketh away"). If there aren't
enough cycles to do block ciphering now, there won'tbe in near future
either. And there are still those low power low bandwidth applications
that have to play by the old rules anyway. So, IMHO there is a market
for stream ciphers also in future.

Secondly, the block size isn't an issue. From the Quality of Service
point of view, the most dificcult application is Voice over IP. VoIP
uses 20 octet packets, and needs low-jitter low-latency QoS. But 20
octets is still more than an AES block!

Stream ciphers are more efficient in terms of bandwidth, because they
don't need padding bits. Anybody interested in developing a 160 bit
block cipher for VoIP?

-- Lassi

> Are stream ciphers really to go away in practice? What are
> the opinions of the group? Aren't there ongoing projects
> of stream cipher designs? (I suppose the issue might
> depend on the 'definition' of stream cipher and block cipher.
> Note that combinations of both types of techniques are
> possible and in fact could be advantageous in my humble view.)
>
> M. K. Shen
> --------------------------------------
> http://home.t-online.de/home/mok-kong.shen

Relevant Pages

  • Re: Sun setting on stream ciphers?
    ... > with the title 'RSA Panel: Cryptography Can't Foil Human ... > on stream ciphers used to encode real-time data streams. ... > depend on the 'definition' of stream cipher and block cipher. ... It's true that most recent designs are actually PRP ...
    (sci.crypt)
  • Re: Stream Ciphers and Confusion/Diffusion
    ... Stream ciphers are not a subset of block ciphers. ... they are classified as such in AC2 and on the RSA FAQ (unless I ... I have also seen this classification in several other books. ... something less than what a block cipher might be used for). ...
    (sci.crypt)
  • Re: Sun setting on stream ciphers?
    ... > Stream ciphers are more efficient in terms of bandwidth, ... > don't need padding bits. ... > block cipher for VoIP? ...
    (sci.crypt)
  • Re: Class Distinction and the Winds of Change.
    ... Scalar Cryptography. ... The science of cryptography was autonomous, ... data types and should taken out of the symmetric stable to go on its ... Security Weaknesses in Maurer-Like Randomized Stream Ciphers. ...
    (sci.crypt)
  • Sun setting on stream ciphers?
    ... on stream ciphers used to encode real-time data streams. ... used to encode data in blocks via block ciphers, ...
    (sci.crypt)