Re: [TLS] [T2TRG] ITDA - IoT Device Authentication

Sankalp, The schema below is the wrong way to use PUF technology for IoT device authentication. PUF is already being used in many fielded IoT systems. You simply use the device unique PUF secret to create a public/private key pair. The public key is then extracted and used to authentic the de

Re: [TLS] ITDA - IoT Device Authentication

Hi, Please see the comments inline. On Mon 18 Feb, 2019, 8:59 AM Salz, Rich, wrote: > If a client will sign any challenge presented by any server, it open up a > big hole. An attacker can present a challenge that a (future) valid server > will present, and the attacker, having the answer, can

Re: [TLS] ITDA - IoT Device Authentication

If a client will sign any challenge presented by any server, it open up a big hole. An attacker can present a challenge that a (future) valid server will present, and the attacker, having the answer, can then act as the device to the server. For example, the attacker can intercept the server/cl

Re: [TLS] ITDA - IoT Device Authentication

Hi, I looked in net for Oracle and found the following limitation. PUF depends on physical variations of chip at time of manufacturing and after processing, should function as a better Oracle than any other algorithm. Limitations (as per wikipedia)Edit

Re: [TLS] ITDA - IoT Device Authentication

Do you know what I mean by an oracle? * PUF provides a cheaper alternative to complex and expensive cryptographic circuitry. As keys need not be stored at the IoT device. When PUF receives a challenge from server, it calculates response and sends it to server. __

Re: [TLS] ITDA - IoT Device Authentication

Hi, Servers are usually more secure and can store challenge/ response pairs for all clients it connects with in Oracle. Remote IoT devices can be attacked physically and keys retrieved from them. To prevent this, costly and complex tamper proof cryptographic circuitry is used. PUF provides a che

Re: [TLS] ITDA - IoT Device Authentication

I would also be concerned about adding a "new" scheme that easily functions as an oracle. On 2/16/19, 8:01 PM, "Peter Gutmann" wrote: Sankalp Bagaria writes: >We propose that the server is authenticated using X509 certificate in a TLS >1.3 like protocol. The Server sends 32-b