ESTABLISHING SECURE MUTUAL TRUST USINGAN INSECURE

PASSWORD

BACKGROUND OF THE INVENTION

[0001] Computer networks are subject to ever increasing security risks.

To protect against attacks some network security protocols utilize public-key

encryption techniques for secure communication. Public-key techniques use two

separate keys - a public key which is made public for others to use and a private

key that is only known to its owner. Each user generates a pair of keys to be used

for encryption and decryption of message. A device's private key is kept secure

and the public key is available to all users. If a first user wishes to send a private

message to a second user, the first user encrypts the message using the second

user's public key. When the second user receives the message, the second user

decrypts it using its private key.

[0002] Alternatively, the public-key technique may be utilized to

authenticate the sender of the message, instead of securely exchanging the message.

In particular, the first user encrypts the message using its own private key. When

the second user receives the message, the second user decrypts it using the first

user's public key. It is appreciated that no other user but the first user could have

generated the encrypted message. Thus, the entire encrypted message serves as a

digital signature. In addition it is not possible to alter the message without the first

user's private key, so the message is authenticated both in terms of source and in

terms of data integrity.

[0003] In yet another implementation, a portion of the message is

encrypted to generate an authenticator. The authenticator is sent with the

unencrypted message. If the authenticator is encrypted with the private key of the

first user, the authenticator serves as a signature that may be utilized to verify the

origin and content of the message.

[0004] In all of the implementations, a method of authenticating the public

key of a given device is needed when using public key techniques. Typically, a

trusted third party (e.g., certificate authority) verifies the identity of entities, such as

individuals and devices. A unique digital certificate is issued to each authenticated

entity which confirm their identity. The certificate typically contains the device

identifier of the certificate holder, a serial number, expiration dates, a copy of the

certificate holder's public key and the digital signature of the certificate-issuing

authority so that a recipient can verify that the certificate is real.

[0005] A recipient of an encrypted message uses the certificate authority's

public key to decode the digital certificate attached to the encrypted message. The

receiving device verifies that the digital certificate was issued by the certificate

authority and then obtains the sender's public key and identification information

encoded in the certificate. With this information, the recipient can securely

communicate on the network with the other device.

[0006] The trusted third party, however, adds a significant amount of

overhead in small networks and/or networks implementing a modest level of

security. A certificate authority can be eliminated and mutual trust can be

established directly between the devices if the certificates are transferred outside of

the network (e.g., output-of-band transfer of certificates). However, to achieve a

reasonable level of security the certificates need to be a large sequence of bits,

typically 1024-bits or more. As a result of their length the digital certificates are

not readily remembered by users, making manual transfer difficult. Furthermore,

the manual transfer of certificates is also difficult if the devices are separated by

large distances (e.g., in another building, across town, in another country).

Accordingly, it is not practical for a user to manually transfer certificates between

devices. Nor can the certificates always be transferred by a.portable computer-

readable medium (e.g., floppy disk, USB key, SD Flash, portable memory card or

the like) because the devices may not have a common portable computer-readable

medium interface.

SUMMARY QF THE INVENTION

[0007] The techniques described herein are directed toward establishing

secure mutual trust between devices over a network using an insecure password. In

one embodiment, the password is acquired by both devices in an exchange off of

the network (e.g., out-of-band transfer). The devices then exchange bit-commit

cryptographic encodings of the password and authentication certificates. Each

device step-wise reveals, to the other device, its digital certificate and the keys used

in the bit-commit cryptographic primitive. Each device step-wise verifies the other

device's certificate by recalculating the bit-commit cryptographic encodings based

upon the revealed certificate and keys. Accordingly, each device verifies that the

other device knew the password and the respective authentication certificate at the

time that the bit-commit cryptographic encodings were exchanged.

BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Embodiments of the present invention are illustrated by way of example and not by way of limitation, in the figures of the accompanying drawings and in which like reference numerals refer to similar elements and in which:

Figures 1, 2 and 3 show a block diagram of a system for implementing

secure mutual trust between devices using an insecure password.

Figures 4, 5 and 6 show a flow diagram of a method of establishing secure

mutual trust by a first device and a second device using an insecure password.

Figures 7, 8 and 9 show a flow diagram of another process of establishing

secure mutual trust between a first and second device using an insecure password.

Figures 1O ₃ 11, 12 and 13 show a flow diagram of another process of

establishing secure mutual trust between a first and second device using an insecure

password.

Figure 14 shows a block diagram of an exemplary operating architecture for implementing secure mutual trust between two devices using an insecure password.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT

[0009] Systems and methods are described herein for establishing secure

mutual trust between two devices using an insecure password. The password is

transferred outside of the network (e.g., out-of-band) from one device to the other.

The password is leveraged by each device to securely exchange authentication

certificates over the network. The authentication certificates are then step-wise

revealed to verify that they were successfully exchanged between the intended

devices. The password is insecure in the sense that it is short. However, a

relatively high level of security is achieved by utilizing the password to exchange

the authentication certificates.

[0010] Figures 1, 2 and 3 show a system for implementing secure mutual

trust between devices using an insecure password. The system includes a plurality

of computing and/or electronic devices 110, 140, 160 communicatively coupled to

each other. In one implementation, the devices 110, 140, 160 may be

communicatively coupled by one or more communication channels 182, 184, 186

directly or through one or more networks 180. The networks 180 may include local

area networks, wide area networks, intranets, extranets, the Internet and/or the like.

[0011] A first device 110 may be a control point device (e.g., host), such

as a personal computer, server computer, client computer, hand-held or laptop

device, set top box, programmable consumer electronic, or similar device. A

second device 140 may be an endpoint device (e.g., responder), such as a personal

computer, server computer, client computer, hand-held or laptop device, set top

box, programmable consumer electronic, or similar device. In an exemplary

implementation, the first device may be a personal computer and the second device

may be an electronic appliance, such as a microwave oven, multimedia digital

recorder, security system or the like. The personal computer is communicatively

coupled to the appliance by a home network. In one implementation, the devices

may be universal plug-and-play (UPnP) devices.

[0012] Mutual trust can be established between the first and second

devices 110, 140 utilizing an insecure one-time-password 120 generated each time

trust establishment is attempted. The password 120 is generated by the second

device 140 and transferred to the first device 110 out-of-band 122. The out-of-band

transfer 122 may be any transfer that does not occur over the network 180. In one

implementation, a user manually transfers 122 a short or medium sized password

120 from the endpoint device to the control point device.

[0013] The one-time-password 120 is then leveraged to securely exchange

the authentication certificates 112, 142 of the devices 110, 140 over the network

180 and to verify that the certificates were successfully exchanged between the intended devices HO ₅ 140. Each device's digital certificate 112, 142 contains

information that establishes the credentials of the device 110, 140. The information

includes the device's public key and a variety of other identification information. In one implementation, the digital certificates 112, 142 are X509 V3 certificates.

[0014] A bit-commit cryptographic primitive is utilized to securely

exchange the password 120 and digital certificate 112, 142 of the given device 110,

140. More specifically, the one-time-password 120 is decomposed by each device

110, 140 into a plurality of password sub-strings 120'. Each device also generates

its own set of random numbers 114, 144. Each random number, in the respective

set 114, 144, is logically paired with a password sub-string 120'. The respective

certificate 112, 142 is then hashed 116, 146 with each one of the corresponding pairs of random numbers 114, 144 and the password sub-strings 120' to generate a

corresponding set of authenticators 118, 148. The respective sets of authenticators

118', 148' are exchanged by the devices 110, 140.

[0015] In a second implementation (not shown), each device generates its

own confirmation random number. The respective certificate 112, 142 is then

hashed 116, 146 with the corresponding random number and one-time-password

120 to generate a corresponding confirmation authenticator. The devices 110, 140 then exchange the respective confirmation authenticators along with the device's certificate 112, 142.

[0016] Referring now to Figure 2, the devices 110, 140 then iteratively

reveal each random number generated by the given device along with the device's

certificate 112, 142. More specifically, the first device 110 may reveal its first random number 114-1' that it generated along with its certificate 112'. The second

device 140 recalculates the second device's first authenticator 218 using the random

number 114-1" and certificate 112" revealed by the first device 110. The second

device 140 verifies that the recalculated authenticator 218 matches the

corresponding authenticator 118' previously received from the first device. The

second device 140 then reveals its first random number along with its certificate, if

the other device's 110 certificate was successfully verified in the previous step.

The first device 110 recalculates the first authenticator using the random number and certificate revealed by the second device 140. The first device 110 verifies that

the recalculated authenticator matches the corresponding authenticator previously

received from the second device 140. The step-wise process of verifying

authenticators is performed for each set of authenticators generated by each device.

If all of the authenticators are verified, secure mutual trust is established between

the devices 110, 140.

[0017] In the second implementation (not shown), each device generates

its own set of validation nonces. Each validation nonce, in the respective set, is

logically paired with a password sub-string 120'. The respective certificate 112,

142 is then hashed 116, 146 with each one of the corresponding pairs of validation nonces and the password sub-strings 120' to generate a corresponding set of validation authenticators. The devices 110, 140 exchange each one of the set of

validation authenticators and then iteratively reveal the respective one of the set of

validation nonces. More specifically, the first device 110 transfers its set of

validation authenticators to the second device 140. The second device 140 then

transfers its set of validation authenticators to the first device 110. The first device

110 then reveals the first one of its set of validation nonces to the second device.

The second device 140 recalculates the first device's corresponding validation authenticator using the validation nonces and certificate 112 revealed by the first

device 110. The second device 140 verifies that the recalculated validation

authenticator matches the corresponding validation authenticator received from the

first device 110. The second device 140 then reveals the first one of its set of

validation nonces, if the first device's certificate was successfully verified in the

previous step. The first device 110 recalculates the first validation authenticator

using the validation nonce and certificate revealed by the second device 140. The

first device 110 verifies that the recalculated validation authenticator matches the

corresponding validation authenticator previously received from the second device

140. The step-wise process of verifying the validation authenticators is performed

for each validation authenticator generated by each device 110, 140.

[0018] In the second implementation (not shown), if all of the validation

authenticators generated by the second device 140 are verified, the first device 110

reveals the confirmation nonces that it previously generated. The second device

140 recalculates the confirmation authenticator using the confirmation nonce and

certificate revealed by the first device 110. The second device 140 verifies that the

recalculated confirmation authenticator matches the corresponding confirmation

authenticator previously received from the first device 110. If all of the validation

authenticators generated by the first device 110 are verified, the second device 140

then reveals the confirmation nonces that it previously generated. The first device

110 recalculates the confirmation authenticator using the confirmation nonce and

certificate revealed by the second device 140. The first device 110 verifies that the

recalculated confirmation authenticator matches the corresponding confirmation

authenticator previously received from the second device 140. If both of the

confirmation authenticators are verified, secure mutual trust is established between

the devices 110, 140.

[0019] In both implementations, the digital certificates 112, 142 are not

authenticated by a trusted third part and therefore may have any format.

Furthermore, it is appreciated that the one-time-password 120 may be any length

and may be decomposed into any number of sub-strings 120'. The length of the

one-time-password 120 and the number of sub-strings 120' may be selected based

upon the level of security required by a particular application. Generally, a one¬

time-password 120 decomposed into a large number of sub-strings 120' provides

more security than the same password 120 decomposed into a smaller number of

substrings 120'. While it may be easier for an attacker to guess a shorter sub-string

120', the attacker only acquires a smaller portion of the one-time-password 120 if

divided into a larger number of strings. For example, if the password 120 is eight

digits long and is divided into four sub-strings 120', there is a one-in-one hundred

chance of guessing a subs-string 120'. However, if the attacker guesses correctly,

only 25% of the one-time-password 120 is acquired by the attacker. If the password

is divided into two sub-strings 120', there is a one-in-ten thousand chance of

guessing a sub-string. However, if the attacker guesses correctly, they acquire 50%

of the one-time-password 120. Similarly, while an attacker can steal a first sub-

string 120' by initiating the establishment of trust, the attacker acquires a smaller

portion of the one-time-password 120 if it is divided into a larger number of sub¬

strings 120'. Thus, if an attacker steals the first sub-string 120' and guess correctly

once, security is compromised if the password 120 was divided into two sub-strings

120'. However, only 50% of the one-time-password 120 is acquired by the attacker

if the password 120 was divided into four sub-strings 120'.

[0020] Once secure mutual trust is established, the public keys contained

in the certificates 112, 142 may be utilized by the devices 110, 140 to securely

communication across the network 180. The secure communication may utilize any

conventional public-key based communication protocol. For example, the first

device 110 may encrypt 310 a message 320 using the second device's public key, as

illustrated in Figure 3. The encrypted message 330 is transmitted across the

network 180 and received by the second device 140. The second device 140

decrypts 350 the message 330" using its private key. Accordingly, the message 330'

transmitted across the network 180 is encrypted and no other device 160 can

decrypt the message 330'. In another implementation (not shown), the first device

110 may utilize its own private key to encrypt the message, which would allow any

device in possession of the first device's 110 public key to decrypt the message and

authenticate that it was sent by the first device 110. In yet another implementation

(not shown), the first device 110 may generate an authenticator utilizing the

device's private key. The authenticator is appended to the message and transmitted

across the network 180. Accordingly, the second device 140 can verify that the first

device 110 sent the message and it has not been modified during transmission.

[0021] Figures 4 ₅ 5 and 6 show a method of establishing secure mutual

trust by a first device and a second device using an insecure password. The method

includes a password generation and out-of-band transfer stage, an information

exchange stage and a validation stage. The method may also include a confirmation

stage. Referring to Figure 4, the method begins with generation of a one-time-

password (OTP) by the second device (e.g., endpoint device), at 405. The one¬

time-password is randomly generated each time trust establishment is attempted.

The one-time-password may be a string of a plurality of characters, numbers or the

like. In one implementation, the one-time-password is a four digit decimal number.

At 410, the one-time-password is received by the first device (e.g., control point

device), via an out-of-band transfer. The out-of-band transfer may be any transfer

that does not occur over the network. In an implementation for establishing secure

mutual trust between an electronic appliance and a computer, the one-time-

password is output on the display of an electronic appliance and a user manually

enters it into the computer.

[0022] At 420, the first device decomposes the one-time-password into a

first plurality of password sub-strings (OTPJ-OTPN). At 425, the second device

decomposes the one-time-password into a second plurality of password sub-strings.

Each substring is a portion of the password. The first and second set of password

sub-strings are equivalent but possessed by the respective devices. In one

implementation, the four-digit decimal number password is decompressed by each

device into four sub-strings of one decimal digit each, In other implementations,

the one-time-password may be eight decimal digits and may be decomposed into

four sub-strings of two decimal digits each, the one-time-password may be six

characters and may be decomposed into three sub-strings of two characters each, or

any similar variation. The length of the password and the number of sub-strings

that it is decomposed into should be selected based upon a desired level of security

and acceptable increase in computation costs.

[0023] At 430, the first device also generates a first set of nonces. At 435,

the second device also generates a second set of nonces. The nonces are each one¬

time only random numbers. In one implementation, the nonces are each 160-bit

random numbers. It is appreciated that the first and second sets of nonces are

independently generated by the respective device and will probabilisticly be

composed of different random numbers. Use of one-time only random number

nonces and a one-time-password protects against an attack that has acquired one or

more password sub-strings and/or nonces from a previous session.

[0024] At 440, the first device generates a first set of authenticators (HA _j -

HA]vf) as ^a function of the first plurality of password sub-strings and the first set of

nonces. At 445, the second device generates a second set of authenticators (D A^-

DAN) ^{as a} function of the second plurality of password sub-strings and the second

set of nonces. More specifically, each device generates its respective set of

authenticators by hashing the respective set of nonces with the respective set of

password sub-strings and the device identifier (e.g., Device_IDs) and certificates

(e.g., Device_Certifϊcates) of the corresponding device. The certificate typically

contains the device identifier of the certificate holder, a serial number, expiration

dates and a copy of the certificate holder's public key. The device identifier is a

string of bits that identifier the type of device.

[0025] The hash algorithm should produce the same output for the same

input. It should be impractical to find a different input that will produce the same

output or to deduce the input given the output. The hash algorithm should also be

verifiable, such that when the input is revealed the authenticity of the revealed

information can be verified. In one implementation, the password sub-strings are

hashed utilizing a bit-commit cryptographic primitive. More particularly, a

message authentication code (MAC) function, such as HMAC-SHAl or the like

may be utilized. The HMAC-SHAl function takes a 160-bit random number key

(e.g., nonce), a body of text (e.g., device identifier, device certificate and sub-string

one-time-password) and produces a 160-bit message authentication code (MAC).

10026] In the exchange of information stage, the devices exchange their

respective set of authenticators. In particular, the first device outputs the first set of

authenticators and its device identifier and certificate, at 450. The first set of

authenticators and the device identifier and certificate of the first device are

received by the second device, at 455. At 460, the second device outputs the

second set of authenticators and its device identifier and certificate. Referring now

to Figure 5, the second set of authenticators and the device identifier and certificate

of the second device are received by the first device, at 510.

[0027] In the validation stage, the devices verify that the other device's certificate was received intact and unaltered. Verification includes the step-wise

revealing of information by the devices. In particular, the first device outputs (e.g., reveals) the first one of the nonces that it generated, its device identifier and

certificate, at 515. The first nonce generated by the first device and the first

device's identifier and certificate are received by the second device, at 520. At 525,

the second device calculates a first validation parameter (e.g., recalculates the first

authenticator received from the first device) based upon the first one of the

password sub-strings that the second device possesses and the nonce, device

identifier and certificate received at operation 520. At 530, the second device

compares the first validation parameter to the first one of the first set of

authenticators received at operation 455.

[0028] At 535, the second device outputs (e.g., reveals) the first one of the

nonces that it generated, its device identifier and certificate. Referring now to

Figure 6, the first nonce generated by the second device and the second device's

identifier and certificate are received by the first device, at 610. At 615, the first

device calculates a first validation parameter (e.g., recalculates the first

authenticator received from the second device) based upon the first one of the

password sub-strings that the first device possesses and the nonce, device identifier

and certificate received at operation 610. At 620, the first device compares the first

validation parameter to the first one of the second set of authenticators received at

operation 510.

[0029] The operations of 515 to 620 are repeated, at 625 and 63O ₅ by the

first and second devices for each of the authenticators. If each validation parameter

matches the corresponding authenticator, mutual trust is established between the

first and second devices at 635 and 640. If either the first or second device

determines that any one of the validation parameters do not match the

corresponding authenticator, the method may be aborted at the applicable operation

or mutual trust may be rejected after completion of operations 405-630. Although

not shown, the method may further include confirmation of the validation. The

confirmation provides an affirmative communication of acceptance or rejection of

trust by each device.

[0030] Figures 7, 8 and 9 show another process of establishing secure

mutual trust between a first and second device using an insecure password. The

process begins with generation of a one-time-password (OTP) by the second device

(e.g., endpoint device), at 705. The one-time-password may be randomly generated

each time the process is initiated. The one-time-password may be a string of

characters, numbers and/or the like. In one implementation, the one-time-password

is a four digit decimal number. At 710, the one-time-password is transferred out-

of-band from the second device to the first device. In one implementation, the out-

of-band transfer entails an electronic appliance generating and displaying the one¬

time-password. A user then types the one-time-password into a personal computer

implementing a home control point.

[0031] The first device may check the received one-time-password against

a list of previously used OTPs, at 715. If the one-time-password has already been

used, generation of a new one-time-password by the second device and subsequent

out-of-band transfer of the one-time-password to the first device may be performed

prior to continuation of the process.

[0032] The first device may generate a first plurality (e.g., N+2) of nonces

(HNi-HN^f ₄ -2), at 720. The second device may generate a second plurality (e.g.,

N+2) of nonces (DN1-DNN+2X at 725. The nonces may be cryptographically

generated random numbers. It is appreciated that the nonces generated by the first

device have a very high probability of being different from the nonces generated by

the second device. In one implementation, each device generates a set of six nonces.

[0033] At 730, the first device decomposes the one-time-password into a

plurality (e.g., N) of password sub-strings (OTPI-OTPN). In one implementation,

the four digit decimal number of the one-time-password is decomposed by the first

device into four sub-strings of one decimal digit each. At 735, the first device

generates a first set of authenticators as a function of the password sub-strings, the

first set of nonces, and the device identifier and certificate of the first device. In

one implementation the certificate is an X509 V3 certificate or the like. The

authenticators may be generated by hashing the respective password sub-strings, the

nonces, the device identifier and certificates. The hash algorithm may be a bit-

commit cryptographic primitive, such as an HMAC or the like. In particular, the

first device generates a plurality of authenticators, denoted as:

MACJHAj = hmac(HN _ls OTP ₁ , HostID, HostCertifϊcate)

MAC-HA ₂ = hmac(HN ₂ , OTP ₂ , HostID, HostCertificate)

MACJHA _N = hmac(HN _N , OTP _N , HostID, HostCertificate)

MAC_HA _A = nmac(HN _N+1 , OTP, HostID, HostCertificate)

MAC_HA _R = hmac(HN _N+2 , OTP, HostID, HostCertificate)

The hmac function takes a 160-bit key (HN _N ) and information (OTP _N ,

HostID, HostCertificate) and produces a 160-bit authentication code (MAC_HA _N ).

The N authenticator values (MACJSAC] through MACJHA _N ) are used in a

corresponding round of the subsequent validation stage. The confirmation and

rejection authenticators (MACJHAA, MAC_HAR) may be calculated based upon

the whole one-time-password and are utilized in the subsequent confirmation stage.

In one implementation, the first device generates six authenticators for a validation

stage implemented in four rounds.

[0034] At 740, the second device also decomposes the one-time-password

into a plurality (e.g., N) of password sub-strings (OTPI-OTPN). In one

implementation, the four digit decimal number of the one-time-password is

decomposed by the second device into four password sub-strings of one decimal

digit each. At 745, the second device generates a second set of authenticators as a

function of the password sub-strings, the second set of nonces, and the device

identifier and certificate of the second device. In particular, the second device

generates a plurality of authenticators, denoted as :

MAC-DA ₁ = hmac(DN _ls OTP ₁ , DevID, DevCertificate)

MAC_DA ₂ = hmac(DN ₂ , OTP ₂ , DevID, DevCertificate)

MAC_DA _N = hmac(DN _N , OTP _N , DevID, DevCertificate)

MAC_D A _A = hmac(DN _N+ i , OTP, DevID, DevCertificate)

MAC_DA _R = hmac(DN _N+2 , OTP, DevID, DevCertificate)

Each authenticator value (MAC_DA] through MAC-DA _N ) is used in a

corresponding round of the subsequent validation stage. The MACJDAA or

MACJDAR may be calculated based upon the whole one-time-password and are

utilized in the subsequent confirmation stage.

[0035] In the exchange of information stage, the devices exchange their

respective device identifiers, device certificates and set of authenticators. In

particular, the first device transfers its device identifier, certificate and the first set

of authenticators to the second device, at 750. At 755, the second device may check the format of the information received from the first device. If the format is valid,

the process may continue. If the format is invalid, the process may be terminated or

the second device may request that the first device resend the information.

[0036] At 760, the second device transfers its device identifier, certificate

and the second set of authenticators to the first device. The first device may check

the format of the information received from the second device, at 765. If the format is valid, the process may continue. If the format is invalid, the process may be

terminated or the first device may request that the second device resend the

information.

[0037] Accordingly, the one-time-password (e.g., a secret shared by both

devices) is broken-up into a plurality of pieces. Hashing each of the pieces with a

nonce effectively commits to the password sub-strings. The commitments are then

exchanged so that the subsequent multistage reveal of the password can validate the

authentication certificates.

[0038] In the validation stage, the devices step-wise reveal the one-time-

password. More specifically, each device verifies that the other device knew the

one-time-password when they generated and exchanged the authenticators.

Referring now to Figure 8, the first device sends the first nonce from the first set of

nonces along with the first device's identifier and certificate, at 810. At 815, the

second device calculates a first one of a first set of validation parameters by hashing

(e.g., MAC) the first nonce received from the first device, the device identifier and

certificate of the first device and the first password sub-string that the second

device possesses. The second device validates that the first device knew the first

password sub-string if the first validation parameter matches the first authenticator

from the first set of authenticators received from the first device, at 820. If the

second device validates that the first device knew the first sub-string one-time-

password, the second device sends the first nonce from the second set of nonces

along with the second device's identifier and certificate, at 825. At 830, the first

device calculates a first one of a second set of validation parameters by hashing the

first nonce received from the second device, the device identifier and certificate of

the second device and the first password sub-string that the first device possesses.

The first device validates that the second device knew the first password sub-string

if the first validation parameter of the second set of validation parameters matches

the first authenticator from the second set of authenticators received from the

second device, at 835.

[0039] Procedures 810-835 are iteratively repeated to step-wise reveal

each of the N password sub-strings. For each iteration, the given device to which

the information has been revealed is able to verify that the other device knew each

of the sub-strings and the authentication certificate when the devices generated and

exchanged the authenticators. Accordingly, after all of the sub-strings of the

password have been revealed by each device, the respective given device is able to

verify that the other device received the device identifier and certificate of the given

device intact and unaltered.

[0040] In the confirmation stage, each device provides an affirmative

communication accepting or rejecting establishment of the trusted relationship. In

particular, the first device indicates to the second device whether verification of the

exchanged information has succeeded or failed by revealing either the acceptance

nonce (HN^+i) or rejection nonce (HNN+2) along with the device identifier and

certificate of the first device, at 840. Referring now to Figure 9, the second device

calculates a confirmation parameter based upon the revealed information, at 910.

At 915, the confirmation parameter is compared to the acceptance authenticator

(MAC_HCA) and the rejection authenticator (MACJHCR). If the confirmation

parameter matches the previously received acceptance authenticator, the second

device knows that the first device has accepted establishment of the trust

relationship. If the confirmation parameter matches the previously received

rejection authenticator,, the second device knows that the first device has rejected

establishment of a trust relationship.

[0041] Similarly, the second device indicates to the first device whether

verification of the exchanged information has succeeded or failed by revealing

either the acceptance nonce (DNjsj+i) or rejection nonce (DN] _\ f-f2) along with the

device identifier and certificate of the first device, at 920. At 925, the first device

calculates a confirmation parameter based upon the revealed information. At 930,

the confirmation parameter is compared to the acceptance authenticator

(MAC JDC pj and the rejection authenticator (MACJDCR). If the confirmation

parameter matches the previously received acceptance authenticator, the first device

knows that the second device has accepted establishment of the trust relationship.

If the confirmation parameter matches the previously received rejection

authenticator, the first device knows that the second device has rejected

establishment of a trust relationship.

[0042] Figures 10, 11, 12 and 13 show another process of establishing

secure mutual trust between a first and second device using an insecure password.

The process begins with generation of a one-time-password (OTP) by the second

device, at 1005. The one-time-password may be randomly generated each time the

process is initiated. The one-time-password may be a string of characters, numbers

and/or the like. In one implementation, the one-time-password is a four digit

decimal number.

[0043] At 1010, the one-time-password is transferred out-of-band from the

second device to the first device. In one implementation, an end point device, such

as a home appliance, generates and displays the one-time-password. A user then

types the one-time-password into a personal computer implementing a home control

point. The first device may check the received one-time-password against a list of

previously used OTPs. If the one-time password has already been used, the out-of-

band transfer may be retried a limited number of times and/or the process may be

aborted. If the one-time-password has not already been used, it may be added to a

list of previously used one-time-passwords and the process may proceed forward.

[0044] At 1015, the first device selects a number of iterations to

performing the subsequent commit and validation process. At 1020, the first device

also generates a first confirmation nonce. The first confirmation nonce may be a

cryptographically generated random number. The first device then generates a first

confirmation authenticator as a function of the one-time-password and the first

confirmation nonce, at 1025. The first confirmation authenticator may also be

generated as a function of the specified iteration count, the first device's identifier,

the first device's certificate and/or the like. In one implementation, the first

confirmation authenticator is generated by hashing the one-time-password, the

confirmation nonce, the specified iteration count, the device identifier and the

device certificate. The hash algorithm may be a bit-commit cryptographic

primitive, such as an HMAC or the like. In particular, the first confirmation

authenticator (MACJHCA) may be specified as:

MACJHCA = hmac(HCN, OTP, HostID, HostCertificate, N)

Wherein HCN is the first confirmation nonce, OTP is the one-time-password,

HostID is the device identifier of the first device, HostCertificate is the certificate

of the first device and N is the specified iteration count.

[0045] At 1030, the first device transfers its device identifier, certificate

and the first confirmation authenticator to the second device. The second device

may check the format of the confirmation authenticator received from the first

device. If the format is valid, the process may continue. If the format is invalid, the

transfer may be retried a limited number of times and/or aborted.

[0046] At 1035, the second device generates a second confirmation nonce.

The second confirmation nonce may be a cryptographically generated random

number. The second device then generates a second confirmation authenticator as a

function of the one-time-password and the second confirmation nonce, at 1040.

The second confirmation authenticate* may also be generated as a function of the

specified iteration count, the second device's identifier, the second device's

certificate and/or the like. In one implementation, the second confirmation

authenticator is generated by hashing the one-time-password, the confirmation

nonce, the specified iteration count, the device identifier and the device certificate.

The hash algorithm may be a bit-commit cryptographic primitive, such as an

HMAC or the like. In particular, the second confirmation authenticator

(MACJL)CA) may be specified as:

MACJ)CA = hmac(DCN, OTP, DevID, DevCertificate, N)

Wherein DCN is the second confirmation nonce, OTP is the one-time-password, HostID is the device identifier of the second device, HostCertificate is the

certificate of the second device and N is the specified iteration count.

[0047] At 1045, the second device transfers its device identifier, certificate

and the second confirmation authenticator to the first device. The first device may

check the format of the confirmation authenticator received from the second device.

If the format is valid, the process may continue. If the format is invalid, the transfer

may be retried a limited number of times and/or aborted.

[0048] At 1105 and 1110, the first and second devices respectively

decompose the one-time-password into a plurality of password sub-strings. The

number of password sub-strings is equal to the selected number of iterations. At 1115, the first device generates a validation nonce for the corresponding iteration of

a commit and validation phase. The nonce may be a cryptographically generated

random number. At 1120, the first device generates a validation authenticator for

the given iteration as a function of a corresponding one of the password sub-strings

and the nonce generated by the first device. The validation authenticator may also

be generated as a function of the corresponding iteration count, the first device's identifier, the first device's certificate and/or the like. In one implementation, the

validation authenticator is generated by hashing the respective password sub¬

strings, the respective validation nonce, the device identifier, the device certificate

and the respective iteration value. The hash algorithm may be a bit-commit

cryptographic primitive, such as an HMAC or the like. In particular, the validation

authenticator (MACJHVAi) generated by the second device for the given iteration

may be specified as: ₍

MACJWAI = hmac(HNi, OTPi, HosHD, HostCertificate, I)

Wherein N is the total number of iterations, I is the value corresponding to the

given iteration, HNi * ^{s tne} respective validation nonce and OTPi * ^{s tne} respective

password sub-string.

[0049] At 1125, the first device transfers its device identifier, certificate

and the validation authenticator corresponding to the particular iteration to the

second device. The second device may check the format of the validation

authenticators. If the format is valid, the process continues. If the format is invalid,

the transfer may be attempted a limited number of times and/or aborted.

[0050] At 1130, the second device generates a respective validation nonce

for the corresponding iteration of the commit and validation phase. At 1135, the

second device generates a validation authenticator for the given iteration as a

function of the corresponding password sub-string and the nonce generated by the

second device. The validation authenticator may also be generated as a function of

the corresponding iteration count, the second device's identifier, the second

device's certificate and/or the like. In one implementation, the validation

authenticator is generated by hashing the respective password sub-string, the

respective validation nonce, the device identifier, the device certificate and the

respective iteration value. In particular, the validation authenticator (MAC_DVAi)

generated by the second device for a given iteration may be specified as:

MAC_DVAi = hmac(DNi, OTPi, DevID, DevCertificate, I)

Wherein N is the total number of iterations, I is the value corresponding to a given

iteration, DNj is the respective validation nonce and OTPi * ^{s tne} respective

password sub-string.

[0051] At 1140, the second device transfers its device identifier, certificate

and the validation authenticator corresponding to the particular iteration to the first

device. The first device may check the format of the validation authenticators. If

the format is valid, the process continues. If the format is invalid, the transfer may

be retried a limited number of times and/or aborted

[0052] At 1145, the first device reveals the respective validation nonce

that it generated along with the first device's identifier and the corresponding

iteration value. Referring now to Figure 12, the second device calculates a

validation parameter by hashing the validation nonce, device identifier and iteration

value received from the first device and the password sub-string corresponding to

the given iteration that the second device possesses, at 1205. At 1210, the second

device validates that the first device knew the first password sub-string if the

validation parameter matches the validation authenticator received from the first

device for the given iteration.

[0053] If the second device validates that the first device knew the first

sub-string, the second device reveals the respective validation nonce that it

generated along with the second device's identifier and the corresponding iteration value, at 1215. At 1220, the first device calculates a validation parameter by

hashing the validation nonce, the device identifier and the iteration value received

from the second device and the password sub-string corresponding to the given

iteration that the first device possesses. The first device validates that the second device knew the first password sub-string if the validation parameter matches the

validation authenticator received from the second device for the given iteration, at

1225.

[0054] Procedures 1115-1225 are iteratively repeated to step-wise commit

and subsequently reveal each of the N password sub-strings. For each iteration, the

given device to which the information has been revealed is able to verify that the

other device knew the corresponding password sub-string when the device

generated and exchanged the validation authenticators. Accordingly, after all of the

sub-strings of the password have been revealed by each device, the respective given

device is able to verify that the other device received the device identifier and

certificate of the given device intact and unaltered.

[0055] If all of the validation authenticators are verified, the first device

reveals the first confirmation nonce, its device identifier and the iteration count, at

1230. The second device calculates a first confirmation parameter by hashing the

first confirmation nonce, device identifier and iteration count received from the first

40

32

device and the one-time-password that the second device possesses, at 1235. At

1240, the second device verifies that the confirmation parameter matches the first

confirmation authenticator previously received from the first device. If the first

confirmation authenticator is verified, the second device reveals the second

confirmation nonce, its device identifier and the iteration count, at 1305. At 1310,

the first device calculates a second confirmation parameter by hashing the second

confirmation nonce, device identifier and iteration count received from the second

device and the one-time-password that the first device possesses. At 1315, the first

device verifies that the confirmation parameter matches the second confirmation

authenticator previously received from the second device. If both of the

confirmation authenticators are verified, secure mutual trust is established between

the first and second devices.

[0056] Figure 14 shows an exemplary operating architecture 1400 for

implementing secure mutual trust between two devices using an insecure password.

The exemplary operating environment 1400 includes a control point device 1410,

and one or more endpoint devices 1420 communicatively coupled to the control

point device 1410. The control point device 1410 and the endpoint devices 1420

may include one or more personal computers, server computers, client computers,

hand-held or laptop devices, set top boxes, programmable consumer electronics

and/or the like.

[0057] An exemplary control point device 1410 or endpoint device 1420

may include one or more processors 1450, one or more computer-readable media

1460, 1470 and one or more input/output devices 1480, 1485. The computer-

readable media 1460, 1470 and input/output devices 1480, 1485 maybe

communicatively coupled to the one or more processors 1450 by one or more buses

1490. The one or more buses 1490 may be implemented using any kind of bus

architectures or combination of bus architectures, including a system bus, a memory

bus or memory controller, a peripheral bus, an accelerated graphics port and/or the

like. The one or more buses 1490 provide for the transmission of computer-

readable instructions, data structures, program modules, code segments and other

data encoded in one or more modulated carrier waves. Accordingly, the one or

more buses 1490 may also be characterized as computer-readable media.

[0058] The input/output devices 1480, 1485 may include one or more

communication ports 1485 for communicatively coupling the exemplary device

1410 to the other devices 1420. One or more of the other devices 1420 may be

directly coupled to one or more of the communication ports 1485 of the exemplary

device 1410. In addition, one or more of the other devices 1420 may be indirectly

coupled through a network 1430 to one or more of the communication ports 1485 of

the exemplary device 1410. The networks 1430 may include an intranet, an

extranet, the Internet, a wide-area network (WAN), a local area network (LAN),

and/or the like.

[0059] The communication ports 1485 of the exemplary device 1410 may

include any type of interface, such as a network adapter, modem, radio transceiver,

or the like. The communication ports may implement any connectivity strategies,

such as broadband connectivity, modem connectivity, digital subscriber link (DSL)

connectivity, wireless connectivity or the like. The communication ports 1485 and

the communication channels 1432, 1434 that couple the devices 1410, 1420 provide

for the transmission of computer-readable instructions, data structures, program

modules, code segments, and other data encoded in one or more modulated carrier

waves (e.g., communication signals) over one or more communication channels

1432, 1434. Accordingly, the one or more communication ports 1485 and/or

communication channels 1432, 1434 may also be characterized as computer-

readable media.

[0060] The exemplary device 1410 may also include additional

input/output devices 1480, such as one or more display devices, keyboards, and

pointing devices (e.g., a "mouse"). The input/output devices may further include

one or more speakers, microphones, printers, joysticks, game pads, satellite dishes,

scanners, card reading devices, digital and video cameras or the like. The

input/output devices 1480 may be coupled to the bus 1490 through any kind of

input/output interface and bus structures, such as a parallel port, serial port, game

port, universal serial bus (USB) port, video adapter or the like.

[0061] The computer-readable media 1460, 1470 may include system

memory 1470 and one or more mass storage devices 1460. The mass storage

devices 1460 may include a variety of types of volatile and non- volatile media, each

of which can be removable or non-removable. For example, the mass storage

devices 1460 may include a hard disk drive for reading from and writing to non¬

removable, non-volatile magnetic media. The one or more mass storage devices

1460 may also include a magnetic disk drive for reading from and writing to a

removable, non-volatile magnetic disk (e.g., a "floppy disk"), and/or an optical disk

drive for reading from and/or writing to a removable, non- volatile optical disk such

as a compact disk (CD), digital versatile disk (DVD), or other optical media. The

mass storage devices 1460 may further include other types of computer-readable

media, such as magnetic cassettes or other magnetic storage devices, flash memory

cards, electrically erasable programmable read-only memory (EEPROM), or the

like. Generally, the mass storage devices 1460 provide for non- volatile storage of

computer-readable instructions, data structures, program modules, code segments,

and other data for use by the exemplary device. For instance, the mass storage

device 1460 may store the operating system 1462, code segments 1464 for

establishing secure mutual trust using an insecure password, and other code and

data 1466.

[0062] The system memory 1470 may include both volatile and non¬

volatile media, such as random access memory (RAM) 1472, and read only memory

(ROM) 1474. The ROM 1474 typically includes a basic input/output system

(BIOS) 1476 that contains routines that help to transfer information between

elements within the exemplary device 1410, such as during startup. The BIOS 1476

instructions executed by the processor 1450, for instance, causes the operating

system 1462 to be loaded from a mass storage device 1460 into the RAM 1472.

The BIOS 1476 then causes the processor 1450 to begin executing the operating

system 1462' from the RAM 1472. The code segment 1464 for establishing mutual

trust may then be loaded into the RAM 1472 under control of the operating system

1462'.

[0063] The processor 1450 of the exemplary device 1410 executes the

various instructions of the code segment 1464' to establish secure mutual trust

between the exemplary device 1410 and another device 1420 using an insecure

password, In particular, the code segment 1464' may either generate a one-time-

password which is output for an out-of-band transfer to the other device 1420, or

the code segment may receive a one-time-password via an out-of-band transfer.

The code segment 1464' generates a first set of authenticators as a function of the

one-time-password and the device's 1410 authentication certificate. The code

segment 1464' iteratively transfers the first set of authenticators to the other device

1420 and receives a second set of authenticators from the other device 1420. The

received authenticators and the one-time-password 1466' may be stored in RAM

1472. The code segment 1464' then iteratively reveals information such that each

of a plurality of sub-strings of the one-time-password may be verified by the other

device 1420. The code segment 1464' verifies that the other device knew the sub¬

string and the authentication certificate as a function of information revealed by the

other device 1420. In addition, the code segment 1464' may also reveal information

such that the other device 1420 may determine if validation has been confirmed by

the code segment 1464'. The code segment 1464' may also verify that the other

device 1420 has confirmed validation.

[0064] Generally, any of the functions, processes of establishing secure

mutual trust using an insecure password described above can be implemented using

software, firmware, hardware, or any combination of these implementations. The

term "logic, "module" or "functionality" as used herein generally represents

software, firmware, hardware, or any combination thereof. For instance, in the case

of a software implementation, the term "logic," "module," or "functionality"

represents computer-executable program code that performs specified tasks when

executed on a computing device or devices. The program code can be stored in one

or more computer-readable media (e.g., computer memory). It is also appreciated

that the illustrated separation of logic, modules and functionality into distinct units

may reflect an actual physical grouping and allocation of such software, firmware

and/or hardware, or can correspond to a conceptual allocation of different tasks

performed by a single software program, firmware routine or hardware unit. The

illustrated logic, modules and functionality can be located at a single site, or can be

distributed over a plurality of locations.

[0065] It is further appreciated that the described control point device

1410 is only one example of a suitable implementation and is not intended to

suggest any limitations as to the scope of use or functionality of the invention. The

computing systems, electronic devices, environments and/or configurations suitable

for use with the invention may be implemented in hardware, software, firmware or

any combination thereof. For example, one or more devices may implement the

process of establishing secure mutual trust utilizing a logic circuit 1422 (e.g.,

hardware and/or firmware) communicatively coupled between an input/output

interface 1424 and a communication channel 1434 of the network 1430. Neither

should the operating architecture be interpreted as having any dependency or

requirement relating to any one component or combination of components

illustrated in the exemplary operating architecture 1400.

[0066] It is appreciated from the above description that embodiments may

advantageously be utilized to establish two-way trusted communications. The

multi-step process of establishing mutual trust uses an insecure password to

exchange full authentication certificates. Embodiments use the relatively short

password to achieve a relatively high level of security utilizing a multistage

iterative technique for exchanging authentication certificate. The password is

insecure in the sense that it is short, not that it is easily stolen. Accordingly, the

multi-step process of establishing mutual trust advantageously mitigates man-in-

the-middle attacks.

[0067] The foregoing descriptions of specific embodiments have been

presented for purposes of illustration and description. Numerous specific details

were set forth in the detail description in order to provide a thorough understanding.

However, it is understood that the present invention may be practiced without these

specific details. In other instances, well-known methods, procedures, components,

and circuits were not described in detail so as not to unnecessarily obscure aspects

of the invention. The described embodiments are not intended to be exhaustive or

to limit the invention to the precise forms disclosed, and obviously many

modifications and variations are possible in light of the above teaching. The

embodiments were chosen and described in order to best explain the principles of

the invention and its practical application, to thereby enable others skilled in the art

to best utilize the invention and various embodiments with various modifications as

are suited to the particular use contemplated. It is intended that the scope of the

invention be defined by the Claims appended hereto and their equivalents.