>>>>>> My understanding is: 
>>>>>> H1 is already there so it should be considered. (notice that H1 
>>>>>> include both rsa-only node or ecc-only node according to the 
>>>>>> definition of definition);
>>>>>> H2 is the type that can be transfered to H4: need to
>>>>>>       
>>>>>>           
>>>>> generated a new CGA covering multiple keys. There is no divice 
>>>>> ability difference between H2 and H4. If requireing a new 
>multi-key 
>>>>> CGA is not too much, it is fair to rule out H2.
>>>>>     
>>>>>         
>>>>>>   
>>>>>>       
>>>>>>           
>>>>> mmm, i am not so sure about this
>>>>> I mean, one question that i have with respect to this is whehter 
>>>>> validating is cheaper than signing?
>>>>> I mean, wouldn't be possible that a node has not enough 
>processing 
>>>>> power to generate an RSA key and sign using the key, but it is 
>>>>> capable of verifying RSA signatures?
>>>>> (i really don't know the answer to this and maybe verifying is as 
>>>>> expensive as signing... i simply don't know)
>>>>>     
>>>>>         
>>>> For rsa, validating is cheaper than signing. So the situation you 
>>>> described is possible.
>>>>   
>>>>       
>>> ok, so do you think is reasonable to impose that all nodes must be 
>>> able to validate RSA signatures? (even light weigth nodes?)
>>>     
>> I don't think the it's proper to impose all nodes must be able to 
>> validate RSA signature. Especially when we are looking for general 
>> agility solutions, we should not consider one particular 
>algorithm different from others.
>>   
>but the fact is that rsa is different cause it is alreadys 
>pecified, and the cga have by default an rsa key and so on
>
>> Besides, to impose rsa validation is too strong, for the nodes which 
>> only want ECC for key size and speed reasons, this 
>requirement is not proper.
>>
>>   
>but you mentioned that RSA validation is cheaper than rsa 
>signing, so maybe these light weigth nodes can do rsa valdiation...

Yes, they can do it if power consumption is the only constraint.

>
>My point is unless we properly scope the types of nodes we 
>want to support and what are the requirements, we are arguing 
>in the abstract. I guess niether you nor me can have an 
>auhtoritative voice on this... 
>maybe we should ask some people more involved on these light 
>weight devices to figure out what are reasonable assumptions to make.
>
Light weight devices venders and CHES conference people will be interested
in this. Wider comments will be helpful, I will go ahead and explore. 


>>   
>>>> What I mean is, If an H2 can already generate an CGA 
>associated with 
>>>> one key and own another key, he can also generate an CGA
>>>>       
>>> associated with two keys.
>>>     
>>>>   
>>>>       
>>> sure, this is not the case i am concerned about
>>>
>>>     
>>>> Possiblly we can use this way to reduce combination of different 
>>>> hosts/routers.
>>>>   
>>>>       
>>> right
>>>
>>>     
>>>> The question is that: is it proper to require "have the ability to 
>>>> perform an algorithm A"<==>"have a public key for algorithm A".
>>>> For the "<==", I think it's easier to be accepted, a device should 
>>>> have the ability to play the algorithm A if it has the key of 
>>>> algorithm A;
>>>>   
>>>>       
>>> right
>>>
>>>     
>>>> For the "==>", I think this is fairly reasonable especially for 
>>>> terminal devices. If an device want to save resouce, it
>>>>       
>>> probably just
>>>     
>>>> want the ability of playing one algorithm.
>>>>   
>>>>       
>>> but i guess this depends on what resoruces are constrained in the 
>>> device.
>>> If what is constrained is processing power, then it may 
>make sense to 
>>> have a device that for instacne does not have a RSA key, but it is 
>>> capable of validating RSA signatures, since as you mentioned above, 
>>> validating RSA signatures is cheaper than signing, right?
>>>     
>> Even if  what is constrained is only processing power, 
>installing RSA 
>> only for validating RSA signature will not help saving power 
>compared 
>> with ECC. I mean, if a terminal can use ECDSA for signature, 
>it can just use ECDSA for
>> validation,     
>> Comparing ECDSA validation speed and RSA validation speed is a little
>> complex: overall, RSA algirithm complexity increase exponentially 
>> while ECDSA algorithm complexity increase sub-exponentialy; on the 
>> same level of cryptographic strength, ECDSA validation is a little 
>> slower but on the same level with RSA validation when key sizes are 
>> shorter than 300 bits, when key sizes are over 300 bits, ECDSA 
>> validation is more and more faster than RSA validation.
>>   
>right, i am not arguing with that
>I am just saying that there are 3 levels of power and CPU and 
>memory consuption The lowest level if ECC only The 
>intermediate level is ECC and RSA validation The highest level 
>is ECC and RSA validation plus RSA signing (wihc requires 
>having a RSA key pair)
>
>I guess the point is whehter the intermediate level is good 
>enough or we need to aim for the lowest level
Personally I think this depend on different devices, also depend on an
vendor's preference: an Certicom product probably don't want RSA at all, an
EMC product may not like ECDSA either, an third vendor might try to gain
best compatibility if device ability alowed... 
I think this is a important one, we will include this question in the draft
for more comments.

>
>> So overall, if a terminal can use ECDSA, it does not need 
>RSA only for 
>> validation in order to save power. If the purpose is compatibility 
>> with old RSA only routers, the RSA only nodes can not validate 
>> terminal's ECDSA signature, SeND can not be performed anyway.
>>
>>   
>well, the node could trust the router and verify the prefixes 
>and the default gateway, but i agree that we can't do full send
>

best,

Sean


_______________________________________________
CGA-EXT mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/cga-ext

Reply via email to