Device Pairing Using Short Authentication Strings

To engage in secure and privacy preserving communication, hosts need to differentiate between authorized peers, which must both know about the host's presence and be able to decrypt messages sent by the host, and other peers, which must not be able to decrypt the host's messages and ideally should not be aware of the host's presence. The necessary relationship between host and peer can be established by a centralized service, e.g. a certificate authority, by a web of trust, e.g. PGP, or -- without using global identities -- by device pairing. This document proposes a device pairing mechanism that provides human operated devices with pairwise authenticated secrets, allowing mutual automatic re-discovery at any later point in time along with mutual private authentication. We especially care about privacy and user-friendliness. The proposed pairing mechanism consists of three steps needed to establish a relationship between a host and a peer: Discovery of the peer device. The host needs a means to discover network parameters necessary to establish a connection to the peer. During this discovery process, neither the host nor the peer must disclose its presence. Agreeing on pairing data. The devices have to agree on pairing data, which can be used by both parties at any later point in time to generate identifiers for re-discovery and to prove the authenticity of the pairing. The pairing data can e.g. be a shared secret agreed upon via a Diffie-Hellman key exchange. Authenticate pairing data. Since in most cases the messages necessary to agree upon pairing data are send over an insecure channel, means that guarantee the authenticity of these messages are necessary; otherwise the pairing data is in turn not suited as a means for a later proof of authenticity. For the proposed pairing mechanism we use manual interaction involving an SAS (short authentication string) to proof the authenticity of the pairing data.

The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in .

The general pairing requirement is easy to state: establish a trust relation between two entities in a secure manner. But details matter, and in this section we explore the detailed requirements that guide our design.

Many pairing protocols have been already developed, in particular for the pairing of devices over specific wireless networks. For example, the current BlueTooth specifications include a pairing protocol that has evolved over several revisions towards better security and usability . The Wi-Fi Alliance defined the Wi-Fi Protected Setup process to ease the setup of security-enabled Wi-Fi networks in home and small office environments . Other wireless standards have defined or are defining similar protocols, tailored to specific technologies. This specification defines a pairing protocol that is independent of the underlying technology. We simply make the hypothesis that the two parties engaged in the pairing can discover each other and can exchange establish connections over IP. TODO: discuss the actual goal. Is it a shared secret or a public key?

The parties in the pairing must be able to identify each other. To put it simply, if Alice believes that she is establishing a pairing with Bob, she must somehow ensure that the pairing is happening with Bob, and not with some interloper like Eve or Nessie. Providing this assurance requires care at the user interface (UI), and also in the design of the protocol. Consider for example an attack in which Eve tricks Alice into establishing a pairing, and then pretends to be Bob. Alice must be able to discover that something is wrong, and refuse to establish a pairing. At a minimum, the parties engaged in the pairing must be able to verify the identity under which the pairing is established.

Because the pairing protocol is executed without prior knowledge, it is typically vulnerable to "Man-in-the-middle" attacks. While Alice is trying to establish a pairing with Bob, Eve positions herself in the middle. Instead of getting a pairing between Alice and Bob, both Alice and Bob get paired with Eve. This requires specific features in the protocol to detect man-in-the-middle attacks, and if possible resist them. The reference analyzes the various proposals to solve this problem, and in this document, we present a layman description of these issues in . The various protocols proposed in the literature impose diverse constraints at the UI interface, which we will review here.

The initial Bluetooth pairing protocol relied on a four digit PIN, displayed by one of the devices to be paired. The user would read that PIN and provide it to the other device. The PIN would then be used in a Password Authenticated Key Exchange. Wi-Fi Protected Setup offered a similar option. There were various attacks against the actual protocol; some of the problems were caused by issues in the protocol, but most were tied to the usage of short PINs. In the reference implementation, the PIN is picked at random by the paired device before the beginning of the exchange. But this requires that the paired device is capable of generating and displaying a four digit number. It turns out that many devices cannot do that. For example, an audio headset does not have any display capability. These limited devices ended up using static PINs, with fixed values like "0000" or "0001". Even when the paired device could display a random PIN, that PIN will have to be copied by the user on the pairing device. It turns out that users do not like copying long series of numbers, and the usability thus dictated that the PINs be short -- four digits in practice. But there is only so much assurance as can be derived from a four digit key. It is interesting to note that the latest revisions of the Bluetooth Pairing protocol do not include the short PIN option anymore. The PIN entry methods have been superseded by the simple "just works" method for devices without displays, and by a procedure based on an SAS (short authentication string) when displays are available. A further problem with these PIN based approaches is that -- in contrast to SASes -- the PIN is a secret instrumental in the security algorithm. To guarantee security, this PIN had to be transmitted via a secure out of band channel.

Some devices are unable to input or display any code. The industry more or less converged on a "push button" solution. When the button is pushed, devices enter a "pairing" mode, during which they will accept a pairing request from whatever other device connects to them. The Bluetooth Pairing protocol denotes that as the "just works" method. It does indeed work, and if the pairing succeeds the devices will later be able to use the pairing keys to authenticate connections. However, the procedure does not provide any protection against MITM attacks during the pairing process. The only protection is that pushing the button will only allow pairing for a limited time, thus limiting the opportunities of attacks. As we set up to define a pairing protocol with a broad set of applications, we cannot limit ourselves to an insecure "push button" method. But we probably need to allow for a mode of operation that works for input-limited and display limited devices.

There have been several attempts to define pairing protocols that use "secure channels." Most of them are based on short range communication systems, where the short range limits the feasibility for attackers to access the channels. Example of such limited systems include for example: QR codes, displayed on the screen of one device, and read by the camera of the other device. Near Field Communication (NFC) systems, which provides wireless communication with a very short range. Sound systems, in which one systems emits a sequence of sounds or ultrasounds that is picked by the microphone of the other system. A common problem with these solutions is that they require special capabilities that may not be present in every device. Another problem is that they are often one-way channels. Yet another problem is that the side channel is not necessarily secret. QR codes could be read by third parties. Powerful radios antennas might be able to interfere with NFC. Sensitive microphones might pick the sounds. We will discuss the specific case of QR codes in .

The evolving pairing protocols seem to converge towards a "display and compare" method. This is in line with academic studies, such as or . This points to a very simple scenario: Alice initiates pairing Bob selects Alice's device from a list. Alice and Bob compare displayed strings that represent a fingerprint of the key. If the strings match, Alice and Bob accept the pairing. Most existing pairing protocols display the fingerprint of the key as a 6 or 7 digit numbers. Usability studies show that gives good results, with little risk that users mistakenly accept two different numbers as matching. However, the authors of found that people had more success comparing computer generated sentences than comparing numbers. This is in line with the argument in to use sequences of randomly chosen common words as passwords. On the other hand, standardizing strings is more complicated than standardizing numbers. We would need to specify a list of common words, and the process to go from a binary fingerprint to a set of words. We would need to be concerned with internationalization issues, such as using different lists of words in German and in English. This could require negotiation of word lists or languages inside the pairing protocols. In contrast, numbers are easy to specify, as in "take a 20 bit number and display it as an integer using decimal notation."

It is tempting to believe that once two peers are connected, they could create a secret with a few simple steps, such as for example exchange two nonces, hash the concatenation of this nonces with the shared secret, display a short authentication string composed of a short version of that hash on each device, and verify that the two values match. The sequence of messages would be something like:

Alice Bob g^xA --> <-- g^xB nA --> <-- nB Computes Computes s = g^xAxB s = g^xAxB h = hash(s|nA|nB) h = hash(s|nA|nB) Displays short Displays short version of h version of h If the two short hashes match, Alice and Bob are supposedly assured that they have computed the same secret, but there is a problem. The exchange may not deter a smart attacker in the middle. Let's redraw the same message flow, this time involving Eve:

Alice Eve Bob g^xA --> g^xA'--> <-- g^xB <--g^xB' nA --> nA --> <-- nB Picks nB' smartly <--nB' Computes Computes s' = g^xAxB' s" = g^xA'xB h' = hash(s|nA|nB') h" = hash(s"|nA|nB) Displays short Displays short version of h' version of h" Let's now assumes that to pick the nonce nB' smartly, Eve runs the following algorithm:

s' = g^xAxB' s" = g^xA'xB repeat pick a new version of nB' h' = hash(s|nA|nB') h" = hash(s"|nA|nB) until the short version of h' matches the short version of h" Of course, running this algorithm will require in theory as many iterations as the possible values of the short hash. But hash algorithms are fast, and it is possible to try millions of values in less than a second. If the short string is made up of fewer than 6 digits, Eve will find a matching nonce quickly, and Alice and Bob will hardly notice the delay. Even if the matching string is as long as 8 letters, Eve will probably find a value where the short versions of h' and h" are close enough, e.g. start and end with the same two or three letters. Alice and Bob may well be fooled. The classic solution to such problems is to "commit" a possible attacker to a nonce before sending it. This commitment can be realized by a hash. In the modified exchange, Alice and Bob exchange a secure hash of their nonces before sending the actual value:

Alice Bob g^xA --> <-- g^xB Computes Computes s = g^xAxB s = g^xAxB h_a = hash(s|nA) --> <-- nB nA --> verifies h_a == hash(s|nA) Computes Computes h = hash(s|nA|nB) h = hash(s|nA|nB) Displays short Displays short version of h version of h Alice will only disclose nA after having confirmation from Bob that hash(nA) has been received. At that point, Eve has a problem. She can still forge the values of the nonces but she needs to pick the nonce nA' before the actual value of nA has been disclosed. On first sight, it seems that once she has committed by sending hash(nA'), it will be impossible to send anything but the nA'. Eve would still have a random chance of fooling Alice and Bob, but it will be a very small chance: one in million if the short authentication string is made of 6 digits, even fewer if that string is longer. Nguyen et al. survey these protocols and compare them with respect to the amount of necessary user interaction and the computation time needed on the devices. The authors state that such a protocol is optimal with respect user interaction if it suffices for users to verify a single b-bit SAS while having a one-shot attack success probability of 2^-b. Further, n consecutive attacks on the protocol must not have a better success probability then n one-shot attacks. There is still a theoretical problem, if Eve has somehow managed to "crack" the hash function. We build some "defense in depth" by some simple measures. In the design presented above, the hash "h_a" depends of the shared secret "s", which acts as a "salt" and reduces the effectiveness of potential attacks based on pre-computed catalogs. For simplicity, the design used a simple concatenation mechanism, but we could instead use a keyed-hash message authentication code (HMAC, , ), using the shared secret as a key, since the HMAC construct has proven very robust over time. Then, we can constrain the size of the random numbers to be exactly the same as the output of the hash function. Hash attacks often require padding the input string with arbitrary data; restraining the size limits the likelyhood of such padding.

Pairing exposes a relation between several devices and their owners. Adversaries may attempt to collect this information, for example in an attempt to track devices, their owners, or their "social graph." It is often argued that pairing could be performed in a safe place, from which adversaries are assumed absent, but experience shows that such assumptions are often misguided. It is much safer to acknowledge the privacy issues and design the pairing process accordingly. In order to start the pairing process, devices must first discover each other. We do not have the option of using the private discovery protocol since the privacy of that protocol depends on the pre-existing pairing. In the simplest design, one of the devices will announce a "friendly name" using DNS-SD. Adversaries could monitor the discovery protocol, and record that name. An alternative would be for one device to announce a random name, and communicate it to the other device via some private channel. There is an obvious tradeoff here: friendly names are easier to use but less private than random names. We anticipate that different users will choose different tradeoffs, for example using friendly names if they assume that the environment is "safe," and using random names in public places. During the pairing process, the two devices establish a connection and validate a pairing secret. As discussed in , we have to assume that adversaries can mount MITM attacks. The pairing protocol can detect such attacks and resist them, but the attackers will have access to all messages exchanged before validation is performed. It is important to not exchange any privacy sensitive information before that validation. This includes, for example, the identities of the parties or their public keys.

The pairing algorithms typically combine the establishment of a shared secret through an [EC]DH exchange, with the verification of that secret through displaying and comparison of a "short authentication string" (SAS). As explained in , the secure comparison requires a "commit before disclose" mechanism. We have three possible designs: create a pairing algorithm from scratch, specifying our own crypto exchanges; use an [EC]DH version of TLS to negotiate a shared secret, export the key to the application as specified in , and implement the "commit before disclose" and SAS verification as part of the pairing application; or, use TLS, integrate the "commit before disclose" and SAS verification as TLS extensions, and export the verified key to the application as specified in . Creating an algorithm from scratch is probably a bad idea. We would need to reinvent a lot of the negotiation capabilities that are part of TLS, not to mention algorithm agility, post quantum, and all that sort of things. It is thus pretty clear that we should use TLS. It turns out that there was already an attempt to define SAS extensions for TLS (). It is a very close match to our third design option, full integration of SAS in TLS, but the draft has expired, and there does not seem to be any support for the SAS options in the common TLS packages. In our design, we will choose the middle ground option -- use TLS for [EC]DH, and implement the SAS verification as part of the pairing application. This minimizes dependencies on TLS packages to the availability of a key export API following . We will need to specify the hash algorithm used for the SAS computation and validation, which carries some of the issues associated with "designing our own crypto." One solution would be to use the same hash algorithm negotiated by the TLS connection, but common TLS packages do not not always make this algorithm identifier available through standard APIs. A fallback solution is to specify a state of the art keyed MAC algorithm.

In , we reviewed a number of short range communication systems that can be used to facilitate pairing. Out of these, QR codes stand aside because most devices that can display a short string can also display the image of a QR code, and because many pairing scenarios involve cell phones equipped with cameras capable of reading a QR code. QR codes could be particularly useful when starting discovery. QR codes can encode an alphanumeric string, which could for example encode the selected name of the pairing service. This would enable automatic discovery, and would be easier to use than reading the random name of the day and matching it against the results of DNS-SD. In addition to the instance name, a QR code could also be leveraged for authentication. It could encode an SAS or even a longer authentication string. Transmitting the output of a cryptographic hash function or HMAC via the OOB channel would make an offline combinatorial search attack infeasible and thus allow to not sent the commitment discussed in saving a message. Further, if a single device created both QR codes for discovery and verification, respectively, and the other device scans these, the users could just wait while both QRs are scanned subsequently as no user interaction is necessary between these two scans (but it needs a QR scanner (app) that support this). This could make the process feel like a single user interaction. But still, from a users point of view, scanning QR codes may not be more efficient than visual verification of a short string. The user has to take a picture of the QR code, which is arguably not simpler than just "look at the number on the screen and tell me whether it is the same as yours". In the case of a man-in-the-middle attack, the evaluation of the QR code will fail. The "client" that took the picture will know that, but the "server" will not. The user will still need to click some "Cancel" button on the server, which means that the process will not be completely automated.

In this section we discuss the design of pairing protocols that use manually verified short authentication strings (SAS), focusing on the user experience. We will make provision for the optional usage of QR codes. We divide pairing in three parts: discovery, agreement, and authentication, detailed in the following subsections.

During the discovery phase, the client will find the server. There are two possible experiences depending of whether QR codes are supported or not. If QR codes are supported, the discovery proceeds as follow: The server displays a QR code containing the its chosen instance name. The client scans the QR code, and discovery is performed automatically, without further client interaction. If QR codes are not supported, the discovery will require some user interaction: The server displays its chosen instance name on its screen. The client performs a discovery of all the "pairing" servers available on the local network. This may result in the discovery of several servers. The client user selects from the list of available servers the instance name that matches the name displayed by the server.

Once the server has been selected, the client connects to it without further user intervention. Client and server exchange the required messages, and both display an SAS, normally a number encoded over up to 7 decimal digits. The users are asked to validate the pairing by comparing the SASes. If they match, each user enters an agreement, for example by pressing a button labeled "OK". If they do not match, each user should cancel the pairing, for example by pressing a button labeled "CANCEL".

The length of the SAS is a compromise between the capacity to resit MITM attacks and a the burden placed on the users performing the comparison. We chose the number 7 based on a "well known" psychology result that "the longest sequence a normal person can recall on the fly contains about seven items" . We chose decimal numbers because they are a least common denominator, and to not require translations in differnt languages. A simple way to obtain a 7 digit is to print a binary number 20 to 23 bits long. In the design, we will use 20 bits because 7 is in fact an upper limit. 20 bits print on 7 digits about 5% of the time, on 6 digits about 86% of the time, and on 5 digits or fewer less than 10% of the time.

Once the agreement has been performed, each device displays the "real" name of the other device, as exchanged through the pairing protocol. Each user verifies that the name matches what they expect. If the names are validated, each user approves the pairing, which will then be remembered.

Users can pair their own devices in secure (home) networks without any interaction using a special DNS-SD pairing service. Verification methods where a single user holds both devices, e.g. synchronously pressing buttons on both devices a few times, are also suitable. Further, a secure OOB could be established by connecting two devices with an USB channel. Pairing via an USB connection is also used by some Bluetooth devices, e.g. when pairing a controller with a gaming console. [[TODO: elaborate]]

To make it sufficient for users to pair only one of their devices to one of their friends devices while still being able to engage in later communication with all of this friend's devices using any of the own devices, we offer the possibility to synchronize pairing data among devices of the same user. Pairing data synchronization is performed via a special DNS-SD service (_pdsync._tls). [[TODO: elaborate]]

[[TODO: Should we discuss public authentication keys whose fingerprints are verified during pairing?]]

[[TODO: elaborate on all subsections]] In the proposed pairing protocol, one of the devices acts as a "server", and the other acts as a "client". The server will publish a "pairing service". The client will discover the service instance during the discovery phase, as explained in . The pairing service itself is specified in .

The discovery uses DNS-SD over mDNS . The pairing service is identified in DNS SD as "_pairing._tcp". When the pairing service starts, the server starts publishing the chosen instance name. The client will discover that name and the corresponding connection parameters.

The pairing protocol is built using TLS. The following description uses the presentation language defined in section 4 of . The protocol uses five message types, defined in the following enum:

enum { ClientHash(1), ServerRandom(2), ClientRandom(3), ServerSuccess(4), ClientSuccess(5) } PairingMessageType; Devices implementing the service MUST support TLS 1.2 , and SHOULD support TLS 1.3 as soon as it becomes available. When using TLS, the client and server MUST negotiate a ciphersuite providing forward secrecy (PFS), and strong encryption (256 bits symmetric key). All implementations using TLS 1.2 SHOULD be able to negotiate the cipher suite TLS_DH_anon_WITH_AES_256_CBC_SHA256. Once the TLS connection has been established, each party extracts the pairing secret S_p from the connection context per , using the following parameters: "PAIRING SECRET" empty. 32 bytes (256 bits). Once S_p has been obtained, the client picks a random number R_c, exactly 32 bytes long. The client then selects a hash algorithm, which SHOULD be the same algorithm as negotiated for building the PRF in the TLS connection. If there is no suitable API to retrieve that algorithm, the client MAY use SHA256 instead. The client then computes the hash value H_c as: H_c = HMAC_hash(S_p, R_c) Where "HMAC_hash" is the HMAC function constructed with the selected algorithm. The client transmits the selected hash function and the computed value of H_c in the Client Hash message, over the TLS connection:

struct { PairingMessageType messageType; hashAlgorithm hash; uint8 hashLength; opaque H_c[hashLength]; } ClientHashMessage; Set to "ClientHash". The code of the selected hash algorithm, per definition of HashAlgorithm in section 7.4.1.1.1 of . The length of the hash H_c, which MUST be consistent with the selected algorithm "hash". The value of the client hash. Upon reception of this message, the server stores its value. The server picks a random number R_s, exactly 32 bytes long, and transmits it to the client in the server hash message, over the TLS connection:

struct { PairingMessageType messageType; opaque R_s[32]; } ServerRandomMessage; Set to "ServerRandom". The value of the random number chosen by the server. Upon reception of this message, the client discloses its own random number by transmitting client random message:

struct { PairingMessageType messageType; opaque R_c[32]; } ClientRandomMessage; Set to "ClientRandom". The value of the random number chosen by the client. Upon reception of this message, the server verifies that the number R_c hashes to the previously received value H_c. If the number does not match, the server MUST abandon the pairing attempt and abort the TLS connection. At this stage, both client and server can compute the short hash SAS as: SAS = first 20 bits of HMAC_hash(S_p, R_c + R_s) Where "HMAC_hash" is the HMAC function constructed with the hash algorithm selected by the client in the ClientHashMessage. Both client and server display the 20 bit value as a decimal integer, and ask the user to compare the values. If the values do not match, the user cancels the pairing. Otherwise, the protocol continues with the exchange of names, both server and client announcing their own preferred name in a Success message

struct { PairingMessageType messageType; uint8 nameLength; opaque name[nameLength]; } ClientSuccessMessage; Set to "ClientSuccess" if transmitted by the client, "ServerSuccess" if by the server. The length of the string encoding the selected name. The selected name of the client or the server, encoded as a string of UTF8 characters. After receiving these messages, client and servers can orderly close the TLS connection, terminating the pairing exchange.

[[TODO: elaborate]] There are a variety of attacks against pairing systems. They may result in compromised pairing keys. To mitigate such attacks, nodes MUST be able to quickly revoke a compromised pairing. This is however not sufficient, as the compromise of the pairing key could remain undetected for a long time. For further safety, nodes SHOULD assign a time limit to the validity of pairings, discard the corresponding keys when the time has passed, and establish new pairings. The requirement of limiting the Time-To-Live of a pairing can raise doubts about the usability of the protocol. The usability issues would be mitigated if the initial pairing provided both a shared secret and the means to renew that secret over time, e.g. using authenticated public keys.

This draft does not require any IANA action.

TODO