Hi Jing,

>-----Original Message-----
>From: Jing Chen [mailto:[EMAIL PROTECTED]
>Sent: Monday, August 04, 2008 10:34 AM
>To: songhaibin 64081
>Cc: [email protected]; Dan York
>Subject: Re: Re :Re: [P2PSIP] P2PSIP security
>
>Hi all,
>  I have a question about the nodeID assignment in P2PSIP. As it said
>that " In order to prevent so-called Sybil or join-leave  attacks, the
>attacker SHOULD NOT be able to easily register a  unlimited number of
>IDs of his choice in the P2SIP overlay.  The  P2PSIP system SHOULD be
>able to control ID assignment. " in "draft-matuszewski-p2psip-security
>requirements-03.txt", how to determine the node's real identification
>in nodeID assignment? By its IP addr? By the username used on it? Or
>by its MAC addr?  The IP addr may change, the username who uses the
>node may change,and so on. I mean that  the attacker may change his
>usernames or his node IP addresses to register  new nodeIDs or perhaps
>he can control many nodes to register many new nodeIDs although our
>P2PSIP system can control ID assignment indeed.

When you bind a node ID with its physical equipment or device, e.g. MAC
addr, a user who has many equipment or devices can obtain many node IDs if
he can afford the cost. When you bind a node ID to a username, then a user
who registered many usernames can also obtain many node IDs, however,
perhaps there are economic resorts to lessen the problem. I mean, you can
not absolutely prevent a "user" to obtain more than one node ID unless you
can bind the node ID to the user's unique characteristic, e.g. passport
number, etc. Most likely we can only prevent a user to "easily" obtain
unlimited number of node IDs.

Regards!
Song Haibin

>
>2008/7/30 songhaibin 64081 <[EMAIL PROTECTED]>:
>> Hi Jing,
>>
>>
>>>  I have three questions about the new security requirements draft (-03):
>>>(1)  The threats and security requirements on DHT network or overlay are
>discussed , but >what about the threats and security requirements on SIP?
>Especially that what are the >threats on decentralized SIP?
>>>
>>
>> Good question. I have discussed this question with Dan before, but we are
so
>busy. You will see the text about it in the next revision.
>>
>>>(2) Why the layers in "Figure 2    P2PSIP architecture" are different
from
>the >architecture layers in RELOAD? The KBR Layer isn't included in the
layers
>of RELOAD, and >this layer is not defined in draft-ietf-p2psip-concepts-02.
>>>
>>
>> I think the WG now have not achieved a consensus on the P2P layers. Bruce
didn't
>mention that issue in his presentation. But I think the architecture are
similar
>with Reload. We will surely make the architecture consistent with the WG's
>consensus.
>>
>> From many papers, you will find KBR layer is often used. Anyway, we will
keep
>it consistent with the WG's consensus item.
>>
>>>(3) As I know, PKI-based certificate is not flexible enough nor bandwidth
>>>efficient because of the size of keys and certificates used. Why we don't
>consider to use
>>>Identity-based security framework ?
>>>
>>
>> I think Identity-based security framework may be easy to use in the self
>organzed network. But the purpose of this draft mainly is not to provide
the
>concrete solutions. It just tell you what the security issues are in the
overlay,
>and you should consider them when you establish a p2p overlay.
>>
>> Best Regards!
>> -Song Haibin
>>
>>
>>
>>
>>

_______________________________________________
P2PSIP mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/p2psip

Reply via email to