Mainly in the management system, management protocol and key generation, distribution, replacement and injection. For the military computer network system, the user mobility is strong, and the subordinate relationship and cooperative operation command are complex, which puts forward higher requirements for key management.
Key management includes all aspects from key generation to key destruction. Mainly in the management system, management protocol and key production management, key distribution, replacement and injection. For the military computer network system, the user mobility is strong, and the subordinate relationship and cooperative operation command are complex, which puts forward higher requirements for key management.
procedure
(1) key generation
The key length should be long enough. Generally speaking, the longer the key length, the larger the corresponding key space, and the more difficult it is for an attacker to guess the password by exhaustive method.
Choose a good key and avoid a weak key. The random bit string generated by automatic processing equipment is a good key. When choosing a key, we should avoid choosing a weak key.
For public-key cryptosystems, it is more difficult to generate keys, because keys must meet certain mathematical characteristics.
Key generation can be achieved through online or offline interactive negotiation, such as encryption protocol.
(2) Key distribution
Using symmetric encryption algorithm for secure communication requires * * * sharing the same key. Usually, one member of the system first selects a secret key and then transmits it to another member or other members. X9. 17 standard describes two kinds of keys: key encryption key and data key. Key encryption key encrypts other keys that need to be distributed; The data key only encrypts the information flow. Key encryption keys are usually distributed manually. In order to enhance confidentiality, the key can be divided into many different parts and then sent out through different channels.
(3) Authentication key
Some error detection and correction bits are appended to the key for transmission. When an error occurs in key transmission, it can be easily detected and the key can be retransmitted if necessary.
The receiver can also verify whether the received key is correct. The sender encrypts a constant with a key, and then sends the first 2-4 bytes of the ciphertext together with the key. At the receiving end, do the same work. If the decryption constant of the receiver can match the original constant, the transmission is error-free.
(4) update the key
When the key needs to be changed frequently, it is really difficult to distribute new keys frequently. A simpler solution is to generate a new key from an old key, sometimes called a key update. You can use the one-way function to update the key. If both parties share the same key and operate with the same one-way function, the result is the same.
(5) key storage
Keys can be stored in brains, magnetic stripe cards and smart cards. The key can also be divided into two parts, half of which is stored in the terminal and the other half is stored in the ROM key. You can also use a method similar to key encryption key to encrypt and save difficult-to-remember keys.
(6) Backup key
The backup of key can adopt the methods of key escrow, secret division and secret sharing.
The easiest way is to use a key escrow center. Key escrow requires all users to give their keys to the key escrow center, which will back up their keys (such as locking them in a safe somewhere or encrypting them with the master key). Once the user's key is lost (for example, the user forgets to bring the key or the user dies unexpectedly), according to certain rules and regulations, the user's key can be obtained from the key escrow center. Another backup scheme is to use smart card as temporary key hosting. For example, Alice stores the key in a smart card and gives it to Bob in Alice's absence. Bob can use this card to do Alice's work. When Alice came back, Bob returned the card. Bob doesn't know what the key is because it is stored in the card.
Secret segmentation divides the secret into many blocks, each block does not mean anything in itself, but when these blocks are put together, the secret will reappear.
A better method is to adopt secret sharing protocol. The key k is divided into n blocks, and each part is called its "shadow". If you know any m or more blocks, you can calculate the key k, but if you know any m-/kloc-0 or less blocks, you can't calculate the key k, which is the so-called (m, n) threshold scheme. At present, many secret sharing schemes have been proposed based on Lagrange interpolation polynomial method, projective geometry, linear algebra, Sun Tzu theorem and so on.
Lagrange interpolation polynomial scheme is an easy-to-understand secret * * * shared (m, n) threshold scheme.
Secret sharing solves two problems: first, if the key is accidentally or intentionally exposed, the whole system will be vulnerable to attacks; Second, if the key is lost or damaged, all the information in the system can't be used.
(7) the validity period of the key
Encryption keys cannot be used indefinitely for several reasons: the longer the key is used, the greater the chance of leakage; If the key has been leaked, the longer the key is used, the greater the loss; The longer the key is used, the more people are tempted to spend energy to decipher it, and even use exhaustive attacks; It is usually easy to analyze multiple ciphertexts encrypted with the same key.
Different keys should have different validity periods.
The validity period of data key mainly depends on the value of data and the number of encrypted data in a given time. The greater the value and data transmission rate, the more frequent the key changes.
Key encryption keys do not need to be changed frequently because they are only used occasionally for key exchange. In some applications, the key encryption key is changed only once a month or a year.
The encryption key used to encrypt the saved data file cannot be changed frequently. Usually each file is encrypted with a unique key, and then all the keys are encrypted with a key encryption key. The key encryption key is memorized or stored in a safe place. Of course, losing this key means losing all the file encryption keys.
In the application of public key cryptography, the validity period of private key varies according to different applications. The private key used for digital signature and identification must be kept for several years (even for life), and the private key used for coin toss agreement should be destroyed immediately after the agreement is completed. Even if the security of the key is expected to be lifelong, the key needs to be changed every two years. The old key still needs to be kept secret in case the user needs to verify the previous signature. However, the new key will be used as the signature of the new file to reduce the number of signature files that can be attacked by password analysts.
(8) Destroy the key
If the key must be changed, the old key must be destroyed and the key must be physically destroyed.
(9) Key management of public key
Public key encryption makes keys easier to manage. No matter how many people are on the network, everyone has only one public key.
It is not enough to use public/private key pairs. Any good implementation of public key encryption needs to separate the encryption key from the digital signature key. But a pair of encryption and signing keys is not enough. Like ID cards, private keys prove a relationship, and people have more than one relationship. For example, Alice can sign documents in her own name or the name of the company's vice president.