[Ndn-interest] Regarding NDN packet security

[Lixia Zhang]

> On Dec 28, 2023, at 9:58 PM, Shashank G <shashank.girish007 at gmail.com> wrote:
> 
> Hi,
> 
> Thanks for the help, Lixia and Kundan. They helped clarify a lot of things for me. I however got a few more doubts going through the resources.
> 
> 1) Suppose a Mal-User floods the PIT with requests for packets that don't exist, how does the Router prevent itself from being overpopulated with bad Requests and continue to serve genuine Requests?

many papers have been published over the years on mitigation solutions against NDN interest flooding.
I just did a quick google search "ndn ddos mitigation", google returned a long list.

> 2) Regarding the Certificate Revocation framework, is there an existing methodology to know if the producer has been compromised?

i) It seems to me that compromise detection is a general problem, not an NDN specific one.
ii) monitoring and auditing seem a general approach here (one of the security's 3 A's: authentication, authorization, audit) 

> 3) I was also hoping for a few resources on How the FIB is populated.

different from IP networks where FIBs can only be populated by routing protocols, FIBs in NDN networks may be populated by multiple means, a few examples:
i) routing protocols, see 
"A Secure Link State Routing Protocol for NDN"
https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8246713

ii) self-learning, see 
"On Broadcast-based Self-Learning in Named Data Networking"
https://named-data.net/wp-content/uploads/2017/04/on_broadcast-based_self-learning_ndn.pdf

iii) through special use of NDN interest packets, see 
"KITE: Producer Mobility Support in Named Data Networking"
https://conferences.sigcomm.org/acm-icn/2018/proceedings/icn18-final23.pdf

iv) more recent work, 
"Reflexive Forwarding for CCNx and NDN Protocols"
https://datatracker.ietf.org/doc/draft-oran-icnrg-reflexive-forwarding/

Happy New Year to all, Lixia


> On Mon, Dec 25, 2023 at 10:56 PM Lixia Zhang <lixia at cs.ucla.edu> wrote:
> "CertRevoke: A Certificate Revocation Framework for Named Data NetworkingShashank, thanks for exploring NDN! From your questions, I guess you may be looking into NDN's name-based access control (NAC) solutions?
> 
> Your first question: the concern seems that, given the data decryption key is encrypted in an object X using a legit's user Y's public key, an attacker Z could easily fetch the decryption key. As Kundan explained, Z can get a copy of X, but wont be able to read the decryption key in it, because Z doesn't have Y's private key.
> 
> Your 2nd question: a good question, yes it is possible to compromise a producer and its key needs to be revoked. NDN based systems try to minimize the danger and damage of key/cert compromises by shortening their life times, though compromises detection is still needed.
> for revocation, there's initial work on NDN cert revocation: "CertRevoke: A Certificate Revocation Framework for Named Data Networking", https://dl.acm.org/doi/pdf/10.1145/3517212.3558079, which utilizes distributed ledgers. The paper referenced DLedger from several years ago, a more recent work is
> "CLedger: A Secure Distributed Certificate Ledger via Named Data"
> https://ieeexplore.ieee.org/document/10279244
> again this is very initial work, we are extending it in another ongoing project to cover more general cases.
> 
> Lixia
> 
> > On Dec 24, 2023, at 11:02 PM, Shashank G via Ndn-interest <ndn-interest at lists.cs.ucla.edu> wrote:
> > 
> > Hi all,
> > 
> > I am Shashank, a sophomore at National Institute of Technology, Karnataka from India. I recently began exploring NDN and have been fascinated by its data security aspect. However, since I am new to the field, I have quite a few doubts regarding the same, and I was hoping for your patience and guidance to clarify them.
> > 
> > 1) I was trying to understand how cryptographically signing packets works, and have got a certain grasp of it's advantages, however, I had a doubt - If the public keys themselves are named, then with the right naming convention, couldn't an attacker get access to data that he is not supposed to view. How is this prevented?
> > 
> > 2) Is there any mechanism to detect if the producer of data has been compromised, i.e, his private key has been obtained by a third party? If so, since the certificates are cached, how do we detect if the producer is safe or not?
> > 
> > I look forward to learning a lot here and eagerly await your response. Thank you.
> > 
> > Yours sincerely,
> > _______________________________________________
> > Ndn-interest mailing list
> > Ndn-interest at lists.cs.ucla.edu
> > https://www.lists.cs.ucla.edu/mailman/listinfo/ndn-interest
>