Split tokenization

Some conventional secure computer systems protect a secret by exchanging a token for that secret. For example, suppose that a merchant accepts credit cards for payment. For such a merchant, locally storing customers' credit card numbers carries a risk of exposing those credit card numbers to an adversary. The merchant can replace each credit card number with a corresponding token, or a number meaningless to the adversary, generated by a secure tokenization server. The merchant would then recover a credit card number by sending the corresponding token back to the tokenization server.

The tokenization server generates tokens from secrets in such a way that an adversary would have very little chance in deducing the secret from the token. For example, the tokenization server can generate a token from a credit card number by applying a cryptographic function to the credit card number. Such a cryptographic function relies on a key which is used to recover the credit card number from the token. The tokenization server can also use a lookup table in a database to recover the credit card number from the token.

Unfortunately, there are deficiencies with the above-described conventional secure computer systems. For example, an adversary may compromise the tokenization server in order to illicitly gain access to secrets stored there. Along the lines of the above example, if such an adversary were to gain access to a key used in a cryptographic function for generating tokens, that adversary could compromise credit card numbers from any tokens generated from that key. Further, that adversary could also compromise credit card numbers taken from a database on the tokenization server.

In contrast to the conventional secure computer systems having a tokenization server that stores secrets that can be compromised by a single illicit access of the tokenization server, an improved technique involves providing protection of secrets by splitting the secret into secret shares and providing tokens for each secret share. Along these lines, a terminal splits a secret such as a credit card number into shares. The terminal then transmits each share to a separate and distinct token server. Each token server, upon receiving a secret share, generates a corresponding token and sends that token to an application server. In some cases, when a user at the application server requires access to the secret, the application server sends each token to the token server form which the token was generated. The token servers each send, in return, a secret share to the application server. The application server combines the secret shares to recover the secret.

Advantageously, the improved technique creates conditions in which an adversary has a much smaller chance of retrieving a secret because that secret is split among several token servers. For example, if the adversary can access a token server with probability ρ, then that adversary may access two token servers with probability ρ 2 <ρ. Further, the chance of such an adversary successfully compromising the secret may be further reduced by proactively updating the shares in such a way as to preserve the secret.

One embodiment of the improved technique is directed to a method of protecting a secret. The method includes performing a secret splitting operation that produces a first secret share and a second secret share. The method also includes sending the first secret share to a first token server, the first token server being constructed and arranged to i) generate a first token from the first secret share and ii) send the first token to an application server. The method further includes sending the second secret share to a second token server that is different from the first token server, the second token server being constructed and arranged to i) generate a second token from the second secret share and ii) send the second token to the application server, the application server being constructed and arranged to recover the secret from the first secret share and the second secret share by exchanging the first token for the first secret share with the first token server and exchanging the second token for the second secret share with the second token server.

Another embodiment of the improved technique is directed to a method of recovering a secret. The method includes sending a first token to a first token server, the first token server being constructed and arranged to generate a first secret share in response to receiving the first token, the first secret share being based on the first token. The method also includes sending a second token to a second token server, the second token server being constructed and arranged to generate a second secret share in response to receiving the second token, the second secret share being based on the second token. The method further includes receiving the first secret share from the first token server and receiving the second secret share from the second token server. The method further includes performing a combination operation on the first secret share and the second secret share to produce the secret.

Additionally, some embodiments of the improved technique are directed to a system constructed and arranged to protect a secret. The system includes a network interface, memory, and a controller including controlling circuitry constructed and arranged to carry out the method of protecting a secret.

Furthermore, some embodiments of the improved technique are directed to a computer program product having a non-transitory computer readable storage medium which stores code including a set of instructions to carry the method of protecting a secret.

The foregoing and other objects, features and advantages will be apparent from the following description of particular embodiments of the invention, as illustrated in the accompanying figures in which like reference characters refer to the same parts throughout the different views.

FIG. 1 is a block diagram illustrating an example electronic environment for carrying out an aspect of the improved technique.

FIG. 2 is a block diagram illustrating an example electronic environment for carrying out another aspect of the improved technique.

FIG. 3 a is a block diagram illustrating an example terminal within the electronic environment shown in FIG. 1 .

FIG. 3 b is a block diagram illustrating an example application server within the electronic environments shown in FIG. 2 .

FIG. 3 c is a block diagram illustrating an example token server within the electronic environments shown in FIGS. 1 and 2 .

FIG. 4 is a block diagram illustrating an example proactivization engine within the token server shown in FIG. 3 .

FIG. 5 is a flow chart illustrating an example variable schedule based on epochs of random length within the electronic system shown in FIG. 1 .

FIG. 6 is a flow chart illustrating a method of carrying out the improved technique within the electronic environment shown in FIG. 2 .

An improved technique involves providing protection of secrets by splitting the secret into secret shares and providing tokens for each secret share. Along these lines, a terminal splits a secret such as a credit card number into shares. The terminal then transmits each share to a separate and distinct token server. Each token server, upon receiving a secret share, generates a corresponding token and sends that token to an application server. In some cases, when a user at the application server requires access to the secret, the application server sends each token to the token server form which the token was generated. The token servers each send, in return, a secret share to the application server. The application server combines the secret shares to recover the secret.

Advantageously, the improved technique creates conditions in which an adversary has a much smaller chance of retrieving a secret because that secret is split among several token servers. For example, if the adversary can access a token server with probability ρ, then that adversary may access two token servers with probability ρ 2 <ρ. Further, the chance of such an adversary successfully compromising the secret may be further reduced by proactively updating the shares in such a way as to preserve the secret.

FIG. 1 illustrates an example electronic environment 10 for carrying out the improved technique. Electronic environment 10 includes terminal 12 , application server 14 , token server 16 a , 16 bm 16 c , 16 d (token servers 16 ), and communications medium 32 .

Terminal 12 is configured to accept a secret 18 ′ such as a credit card number from a user (not pictured). Terminal 12 is further configured to split secret 18 into a set of secret shares 18 a , 18 b , 18 c , 18 d (secret shares 18 ) and send secret shares 18 to corresponding token servers 16 . In some arrangements, such as in an on-line shopping scenario, terminal 12 takes the form of a desktop computer. In other arrangements, such as in a brick-and-mortar shopping scenario, terminal 12 takes the form of a credit card reader.

Application server 14 is configured to receive tokens 20 a , 20 b , 20 c , and 20 d (tokens 20 ) from token servers 16 . Application server 14 typically takes the form of a web server, although, in some arrangements, application server may be a desktop computer.

Token servers 16 a , 16 b , 16 c , and 16 d (token server 16 ) are configured to receive a secret share 18 a , 18 b , 18 c , or 18 d , respectively, and generate a token 20 a , 20 b , 20 c , or 20 d , respectively, from corresponding secret share 18 . Token servers 16 are further configured to send tokens 20 to application server 14 .

Communication medium 32 provides network connections between terminal 12 , application server 14 , and token servers 16 . Communications medium 32 may implement a variety of protocols such as TCP/IP, UDP, ATM, Ethernet, Fibre Channel, combinations thereof, and the like. Furthermore, communications media 32 may include various components (e.g., cables, switches/routers, gateways/bridges, NAS/SAN appliances/nodes, interfaces, etc.). Moreover, the communications medium 32 are capable of having a variety of topologies (e.g., queue manager-and-spoke, ring, backbone, multi drop, point to-point, irregular, combinations thereof, and so on).

In some arrangements, communications medium 32 includes a content delivery network (CDN) including a set of Edge servers (not pictured) through which content is routed between terminal 12 and token servers 16 .

During operation, terminal 12 receives a secret 18 ′ from a user. For example, terminal 12 receives a credit card number from the user as the user attempts to make a purchase. In some arrangements, the user inputs the credit card number into terminal 12 via an internet browser. In other arrangements, the user inputs the credit card number into terminal 12 via a swiping mechanism in a dedicated device or a smartphone.

Upon receiving secret 18 ′, terminal 12 performs a splitting operation on secret 18 ′ to produce secret shares 18 a , 18 b , 18 c , and 18 d such that a combination operation (i.e., an inverse of the splitting operation) recovers secret 18 ′. For example, suppose that the splitting operation involves generating three random, 16-digit numbers, and then performing a bitwise XOR operation on those three random numbers and a credit card number 18 ′ input into terminal 12 to produce a fourth number. Each of the generated random numbers and the fourth number then is a secret share 18 . The credit card number 18 ′ may be recovered by performing the bitwise XOR operation on the secret shares 18 . In some arrangements, the Edge servers in the CDN may perform the splitting operation.

Terminal 12 also generates a random nonce 22 along with secret shares 18 . Terminal 12 provides random nonce 22 as proof that terminal 12 generated secret shares 18 together.

Upon producing secret shares 18 a , 18 b , 18 c , and 18 d , terminal 12 sends each secret share 18 to a corresponding token server 16 a , 16 b , 16 c , or 16 d via communications medium 32 . Terminal 12 also sends random nonce 22 to each token server 16 with corresponding secret share 18 . Terminal 12 also sends a destination address (i.e., IP address) of application server to each token server with random nonce 22 and corresponding secret share 18 . In some arrangements, terminal 12 encrypts secret share 18 , random nonce 22 , and the destination address (i.e., using a public key of the corresponding token server 16 ) prior to sending. In some arrangements, terminal 12 also send identification information corresponding to the owner of secret 18 ′.

Token servers 16 a , 16 b , 16 c , and 16 d each receive corresponding secret share 18 , random nonce 22 , and the destination address from terminal 12 . In some arrangements, each token server 16 decrypts what was received using a private key corresponding to the public key used for encryption to reveal secret share 18 , random nonce 22 , and the destination address,

Upon receiving secret share 18 a , 18 b , 18 c , or 18 d , each token server 16 a , 16 b , 16 c , or 16 d generates a corresponding token 20 a , 20 b , 20 c , or 20 d . Token server 16 applies a transformation to secret share 18 such as a keyed cryptographic function having a key in order to produce token 20 . In some arrangements, upon generating token 20 , token server 16 stores the key from the keyed cryptographic function with the stored token in a database (not pictured).

After generating token 20 , token server 16 sends token 20 and random nonce 22 to application server 14 at the destination address received from terminal 12 . Upon receiving tokens 20 , application server stores tokens 20 on an accessible storage device. In some arrangements, application server 14 combines tokens 20 into a single, combined token.

It should be understood that, in some arrangements, each token server 16 sends its corresponding token 20 back to terminal 12 . In such a case, terminal 12 sends tokens 20 to application server 14 after receiving tokens 20 . Further, terminal 12 need not send a destination address to token servers 16 when sending corresponding secret shares 18 .

FIG. 2 illustrates another example electronic environment 10 ′ for carrying out the improved technique. Electronic environment 10 ′ includes application server 14 , token servers 16 a , 16 bm 16 c , 16 d (token servers 16 ), secret-handling terminal 28 , and communications medium 32 .

Application server 14 is configured to send tokens 24 a , 24 b , 24 c , 24 d (tokens 24 ) to corresponding token server 16 a , 16 b , 16 c , and 16 d from which tokens 24 were generated. Application server 14 is further configured to receive secret shares 30 a , 30 b , 30 c , and 30 d (secrets shares 30 ) from corresponding token servers 16 . Application server 14 is further configured to combine secret shares 30 to recover secret 18 ′. In some arrangements, application server is further configured to send secret 18 ′ to a secret-handling terminal 26 . As an example, a merchant that recovers a credit card number 18 ′ from tokens 20 would send credit card number 18 ′ to a terminal 26 at a credit card company.

During operation, application server 14 sends tokens 24 and random nonce 22 (see FIG. 1 ) to corresponding token servers 16 in response to a request to recover secret 18 ′. Tokens 24 are equivalent to corresponding tokens 20 , but in some arrangements, tokens 24 each represent identical, combined tokens. In some arrangements, application server generates a hash of random nonce 22 before transmission to token servers 16 .

Upon receiving token 24 , corresponding token server 16 recovers secret share 30 from token 24 . For example, token server 16 applies a cryptographic function having a key related to the key that used to generate token 20 from secret share 18 to token 24 to recover secret share 30 . In some arrangements, token 24 is identical to token 20 , and secret share 30 is identical to secret share 18 . In other arrangements, however, token 24 results from a combination of tokens 20 a , 20 b , 20 c , and 20 d and is identical for all token servers. In still other arrangements, each secret share 30 differs from the corresponding secret share 18 in such a way that a combination operation performed on secret shares 30 still yields secret 18 ′. Details of such an arrangement are discussed below with respect to FIG. 4 .

It should be understood that, as an additional security measure, token servers 16 check each request for secret shares 30 from application server 14 for random nonce 22 . Token servers 16 will generate secret shares 30 when random nonce 22 matches that token servers 16 sent to application server 14 along with tokens 20 . As illustrated in FIG. 2 , token servers 16 generate a hash of random nonce 22 previously stored, and compare this hash to hash 28 sent from application server 14 .

It should also be understood that application server 14 may also encrypt tokens 30 , random nonce 22 (or its hash), and any identification information pertaining to the owner of secret 18 ′. Token servers 16 are then configured to decrypt such a transmission using a private key corresponding to a public key used to encrypt the transmission.

When random nonce 22 provides a match, token servers 16 send corresponding token shares 30 to application server 14 . Upon receipt of token shares 30 , application server 14 performs a combination operation on token shares 30 to produce secret 18 ′. In some arrangements, application server 14 sends secret 18 ′ to secret-handling terminal 26 and does not store secret 18 ′ on an accessible storage device. Further details of an example combination operation are discussed below with respect to FIG. 3 c.

Further details of terminal 12 are discussed below with respect to FIG. 3 a below.

FIG. 3 a illustrates an example terminal 12 . Terminal 12 includes controller 40 which in turn includes processor 44 and memory 46 , and network interface 42 .

Network interface 42 takes the form of an Ethernet card; in some arrangements, network interface 42 takes other forms including a wireless receiver and a token ring card.

Memory 46 is configured to store code which includes share generation code 54 , random number generation code 56 , and encryption code 58 . Memory 46 generally takes the form of, e.g., random access memory, flash memory or a non-volatile memory.

Processor 44 takes the form of, but is not limited to, Intel or AMD-based MPUs, and can include a single or multi-cores each running single or multiple threads. Processor 44 is coupled to memory 46 and is configured to execute instructions from share generation code 54 , random number generation code 56 , and encryption code 58 . Processor 44 includes share generation engine 48 , random number generation engine 50 , and encryption engine 52 which are configured to execute instructions derived from share generation code 54 , random number generation code 56 , and encryption code 58 , respectively.

During operation, share generation engine 48 generates random shares 18 from secret 18 ′ by performing a splitting operation on secret 18 ′. In some arrangements, the splitting operation involves generating three random, 16-digit numbers, and then performing a bitwise XOR operation on those three random numbers and a credit card number 18 ′ input into terminal 12 to produce a fourth number. Each of the generated random numbers and the fourth number then is a secret share 18 . The credit card number 18 ′ may be recovered by performing the bitwise XOR operation on the secret shares 18 .

In other arrangements, however, secret 18 ′ is split into multiplicative factors according to a cyclic group having a particular generator. Further details of such a splitting will be discussed below with respect to FIG. 3 b.

Random number generation engine 50 generates random nonce 22 . For example, random number generation engine outputs a 256-bit nonce according to a uniform distribution. In some arrangements, random number generation engine 50 generates nonce 22 upon generation of random shares 18 ; in other arrangements, random number generation engine 50 generates nonce 22 periodically.

Encryption engine 52 then performs an encryption operation on each transmission to token servers 16 that includes a shared secret 18 . For example, in some arrangements, processor 44 forms a string by concatenating a shared secret 18 , random nonce 22 , and a destination address of application server 14 . Encryption engine 52 then applies, to the string, a public key corresponding to the private key of the token server 16 to which that secret share 18 is to be transmitted. Processor 44 then sends the encrypted string to a token server 16 via network interface 42 .

Further details of each token server 16 are discussed below with respect to FIG. 3 b.

FIG. 3 b illustrates an example token server 16 . Token server 16 includes controller 60 which in turn includes processor 64 and memory 66 , and network interface 62 .

Network interface 62 takes the form of an Ethernet card; in some arrangements, network interface 62 takes other forms including a wireless receiver and a token ring card.

Memory 66 is configured to store code which includes token generation code 74 , hashing code 76 , and proactivization code 78 . Memory 66 generally takes the form of, e.g., random access memory, flash memory or a non-volatile memory.

Processor 64 takes the form of, but is not limited to, Intel or AMD-based MPUs, and can include a single or multi-cores each running single or multiple threads. Processor 64 is coupled to memory 66 and is configured to execute instructions from token generation code 74 , hashing code 76 , and proactivization code 78 . Processor 64 includes token generation engine 68 , hashing engine 70 , and proactivization engine 72 which are configured to execute instructions derived from token generation code 74 , hashing code 76 , and proactivization code 78 , respectively.

During operation, processor 64 receives a secret share 18 via network interface 82 . In some arrangements, processor 44 receives an encrypted string that, when decrypted by processor 64 using a private key, reveals secret share 18 , random nonce 22 , and the destination address of application server 14 .

Token generation engine 88 generates a token 20 from secret share 18 . In some arrangements, token generation engine 88 corresponding to token server 16 i , where iε{a,b,c,d}, utilizes a cryptographic function T I (K i ,S i ) to generate a token T i , S i being secret share 18 i , K i representing a secret key, and Iε{A,B,C,D} corresponding to iε{a,b,c,d}. That is, T i =T I (K i ,S i ). Token generation engine 88 keeps track of pairs (K i ,T i ) for eventual recovery of a secret share from a token. Hashing engine 70 also generates a hash of nonce 28 .

Processor 64 then sends token 20 (T i ) and nonce 22 to application server 14 . Further details of application server 14 are discussed below with respect to FIG. 3 c.

FIG. 3 c illustrates an example application server 14 . Application server 14 includes controller 80 which in turn includes processor 84 and memory 86 , and network interface 82 .

Network interface 82 takes the form of an Ethernet card; in some arrangements, network interface 82 takes other forms including a wireless receiver and a token ring card.

Memory 86 is configured to store code which includes share combination code 92 and hashing code 94 . Memory 86 generally takes the form of, e.g., random access memory, flash memory or a non-volatile memory.

Processor 84 takes the form of, but is not limited to, Intel or AMD-based MPUs, and can include a single or multi-cores each running single or multiple threads. Processor 84 is coupled to memory 86 and is configured to execute instructions from share combination code 92 and hashing code 94 . Processor 84 includes share combination engine 88 and hashing engine 90 which are configured to execute instructions derived from share combination code 92 and hashing code 94 , respectively.

During operation, processor 84 receives, via network interface, token 20 and nonce 22 from each token server 16 . Processor 84 checks that each nonce 22 received is identical to the other nonces received. Processor 84 then stores each token 20 in an accessible storage device. In some arrangements, processor 84 performs a combination operation on tokens 20 to produce a combined token T C , preferable by concatenating tokens T i for iε[a,b,c,d] in order.

Hashing engine 70 applies a hash function h to random nonce 22 , denoted by N to produce hashed nonce 28 , i.e., h(N). In some arrangements, processor 84 concatenates h(N) with combined token T C .

Sometime later, when a user (not pictured) wishes to recover secret 18 ′, denoted by S, processor 84 sends tokens 24 and hashed nonce 28 to each token server 16 . It should be understood that tokens 24 may be identical to corresponding tokens 18 , i.e., T i . In some arrangements, however, each token 24 is the combined token T C .

Referring back to FIG. 3 b , processor 64 of token server 16 receives, via network interface 62 , corresponding token 24 and hashed nonce 28 . In response, processor 64 compares the hashing result from hashing engine 70 to hashed nonce 28 . If there is a mismatch, processor 70 sends an error message to application engine 14 .

If there is a match, processor 64 performs a lookup operation on token T i in order to find the corresponding key K i . In some arrangements, when processor 64 receives combined token T C , processor 64 performs a token separation operation to recover the token T i . Such a token separation operation is common to all token servers 16 .

Token generation engine 68 then utilizes another cryptographic function, D I (K i ,T i ) to produce a new secret share 30 , denoted as D i =D I (K i ,T i ) such that, when the secret splitting operation split secret 18 ′ (S) into products under a cyclic group G, the product of the new secret shares 30 produce secret S. That is, S=S a S b S c S d =D a D b D c D d , where multiplication is understood to be under G.

In some arrangements, token servers 16 , via proactivization engine 72 , proactively update keys at regular intervals in order to provide deeper security for the split tokenization process outlined here. That is, token servers synchronously update the keys while preserving the value of S=D a D b D c D d . The following represents a scheme for proactive updates in the manner described, and is described with respect to FIG. 4 below.

FIG. 4 represents the above scheme by which cryptographic keys used to generate tokens and secret shares are updated proactively. Consider the following tokenization function: T I (K i ,S i )=S i g K i , where g is the generator of G. Consider further the following “share-ization” function D I (K i ,T i )=T i g −K i that returns a secret share D i from a token. At some point in time, token servers 16 update their respective keys by adding to each key a quantity x i . Then each new share D′ i =D I (K i +x i ,T I (K i ,S i ))=T i g −K i −x i =S i g K i g −K i −x i =S i g −x i , where addition is understood to be under G. The product of the new shares over the token servers is then S a S b S c S d g −(x a +x b +x c +x d ) . To recover the secret, the updates should satisfy the following condition: x a +x b +x c +x d =0.

FIG. 4 illustrates a scenario in which there are two token servers, 16 a and 16 b , involved in the updates. In this scenario, x a =−x b . Note that proactivization engine 72 in each token server has a synchronized random number generator (not pictured) for generating such updates. In the case of four token servers above, such synchronized random number generators would follow a more complex scheme. In general, the split tokenization scheme outlined here may work with any number of token servers.

It should be understood that, in the above examples, it was assumed that there was no interaction between token servers 16 . In some arrangements, however, more general cryptographic functions may be used when there is such interaction. In such a scenario, application server produces a combined token T C =T a T b T c T d and sends T C to token servers 30 . A token server 16 then utilizes a “share-ization” function D I (K a ,K b ,K c ,K d ,T C )=T C g −[(K] a +K b +K c +K d ) . An updating scheme as outlined above will also produce the secret 18 ′ here as well.

It should also be understood that there are alternatives to using cryptographic functions in generating tokens from secret shares and vice-versa. For example, token server 16 may utilize tables of tokens and corresponding secret shares. Such tables are implemented using a function T i =T I (S i ). Such tables may be proactively updated as follows. Suppose that there are two token servers. Further suppose that each token server maintains a first skip list that stores entries [[T] I (S i ),h(N),S i ] ordered according to T I (S i ) primarily and h(N) secondarily, and a second skip list that stores entries [[S i ,T] I (S i )] according to S i . The token servers then agree on a joint random bit string Σ and chunk size c. The first c entries in [[S i ,T] I (S i )] in the second skip list are updated to [[S I i,T] I I(S I i)]⊕ “Hash” (Σ∥S I i), where ⊕ represents a bitwise XOR operation. During this update, the corresponding entries [[T] I (S i ),h(N),S i ] in the first skip list are updated to [[T] I I(S I i)⊕“Hash” (Σ∥S I i),h(N),S I i] and copied to a new ordered list. Each token server 16 then replaces Σ with a one-way function f(Σ) and proceeds with a second chunk of c entries. Continuing along these lines for each token server 16 will result in a secure update that may be used to recover secret shares from tokens [T a ∥T b ∥h(N)].

FIG. 5 illustrates a method 100 of protecting a secret, which includes steps 102 , 104 , and 106 . In step 102 , a secret splitting operation that produces a first secret share and a second secret share is performed. In step 104 , the first secret share is sent to a first token server, the first token server being constructed and arranged to i) generate a first token from the first secret share and ii) send the first token to an application server. In step 106 , the second secret share is sent to a second token server that is different from the first token server, the second token server being constructed and arranged to i) generate a second token from the second secret share and ii) send the second token to the application server, the application server being constructed and arranged to recover the secret from the first secret share and the second secret share by exchanging the first token for the first secret share with the first token server and exchanging the second token for the second secret share with the second token server.

FIG. 6 illustrates a method 110 of recovering a secret, which includes steps 112 , 114 , 116 , 118 , and 120 . In step 112 , a first token is sent to a first token server, the first token server being constructed and arranged to generate a first secret share in response to receiving the first token, the first secret share being based on the first token. In step 114 , a second token is sent to a second token server, the second token server being constructed and arranged to generate a second secret share in response to receiving the second token, the second secret share being based on the second token. In step 116 , the first secret share is received from the first token server. In step 118 , the second secret share is received from the second token server. In step 120 , a combination operation is performed on the first secret share and the second secret share to produce the secret.

While various embodiments of the invention have been particularly shown and described, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.

For example, the examples above assumed that the secret took the form of personal credit information (PCI) such as credit card numbers. Nevertheless, the improved techniques described above are also useful for health records, insurance information, social security numbers, and other examples of PII and PHI information.

Furthermore, it should be understood that some embodiments are directed to terminal 12 , which is constructed and arranged to protect a secret. Some embodiments are directed to a process of protecting a secret. Also, some embodiments are directed to a computer program product which enables computer logic to protect a secret.

Moreover, it should be understood that some embodiments are directed to application server, which is constructed and arranged to recover a secret. Some embodiments are directed to a process of recovering a secret. Also, some embodiments are directed to a computer program product which enables computer logic to recover a secret.

In some arrangements, terminal 12 is implemented by a set of processors or other types of control/processing circuitry running software. In such arrangements, the software instructions can be delivered, within terminal 12 , in the form of a computer program product 130 (see FIG. 3 a ), each computer program product having a computer readable storage medium which stores the instructions in a non-volatile manner. Alternative examples of suitable computer readable storage media include tangible articles of manufacture and apparatus such as CD-ROM, flash memory, disk memory, tape memory, and the like.

In some arrangements, application server 14 is implemented by a set of processors or other types of control/processing circuitry running software. In such arrangements, the software instructions can be delivered, within application server 14 , in the form of a computer program product 150 (see FIG. 3 c ), each computer program product having a computer readable storage medium which stores the instructions in a non-volatile manner. Alternative examples of suitable computer readable storage media include tangible articles of manufacture and apparatus such as CD-ROM, flash memory, disk memory, tape memory, and the like.

In some arrangements, token server 16 is implemented by a set of processors or other types of control/processing circuitry running software. In such arrangements, the software instructions can be delivered, within token server 16 , in the form of a computer program product 140 (see FIG. 3 b ), each computer program product having a computer readable storage medium which stores the instructions in a non-volatile manner. Alternative examples of suitable computer readable storage media include tangible articles of manufacture and apparatus such as CD-ROM, flash memory, disk memory, tape memory, and the like.