English French
This principle is important because it remains the basic assumption of all cryptographers. Any system that relies on the secrecy of its algorithm to be considered secure is doomed to fail and be broken one day.
With the Kerckhoff principle, we can now discuss a simple but powerful encryption scheme that relies on the `XOR` logic operation. This operation is easily implemented in hardware and is supported by all microprocessors. Given a secret, :math:`K`, it is possible to encode a message `M` by computing :math:`C_M = K \oplus M`. The receiver of this messages can recover the original message as since :math:`M = K \oplus (K \oplus M)`. This `XOR` operation is the key operation of the perfect cipher that is also called the Vernam cipher or the one-time pad. This cipher relies on a key that contains purely random bits. The encrypted message is then produced by XORing all the bits of the message with all the bits of the key. Since the key is random, it is impossible for an attacker to recover the original text (or plain text) from the encrypted one. From a security viewpoint, the one-time-pad is the best solution provided that the key is as long as the message.
Unfortunately, it is difficult to use this cipher in practice since the key must be as long as the message that needs to be transmitted. If the key is smaller than the message and the message is divided into blocks that have the same length as the key, then the scheme becomes less secure since the same key is used to decrypt different parts of the message. In practice, `XOR` is often one of the basic operations used by encryption schemes. To be usable, the deployed encryption schemes use keys that are composed of a small number of bits, typically 56, 64, 128, 256, ...
A secret key encryption scheme is a perfectly reversible functions, i.e. given an encryption function `E`, there is an associated decryption function `D` such that :math:`\forall k \forall M : D(K, E(M,K))=M`.
Various secret key cryptographic functions have been proposed, implemented and deployed. The most popular ones are :
DES, the Data Encryption Standard that became a standard in 1977 and has been widely used by industry. It uses 56 bits keys that are not considered sufficiently secure nowadays since attackers can launch brute-force attacks by testing all possible keys. Triple DES combines three 56 bits keys, making the brute force attacks more difficult.
RC4 is an encryption scheme defined in the late 1980s by Ron Rivest for RSA Security. Given the speed of its software implementation, it has been included in various protocols and implementations. However, cryptographers have identified several weaknesses in this algorithm. It is now deprecated and should not be used anymore :rfc:`7465`.
AES or the Advanced Encryption Standard is an encryption scheme that was designed by the Belgian cryptographers Joan Daemen and Vincent Rijmen in 2001 [DR2002]_. This algorithm has been standardized by the U.S. National Institute of Standards and Technology (NIST). It is now used by a wide range of applications and various hardware and software implementations exist. Many microprocessors include special instructions that ease the implementation of AES. AES divides the message to be encrypted in blocks of 128 bits and uses keys of length 128, 192 or 256 bits. The block size and the key length are important parameters of an encryption scheme. The block size indicates the smallest message that can be encrypted and forces the sender to divide each message in blocks of the supported size. If the message is larger than an integer number of blocks, then the message must be padded before being encrypted and this padding must be removed after decryption. The key size indicates the resistance of the encryption scheme against brute force attacks, i.e. attacks where the attacker tries all possible keys to find the correct one.
AES is widely used as of this writing, but other secret key encryption schemes continue to appear. ChaCha20, proposed by D. Bernstein is now used by several internet protocols :rfc:`7539`. A detailed discussion of encryption schemes is outside the scope of this book. We will consider encryption schemes as black boxes whose operation depends on a single key. A detailed overview of several of these schemes may be found in [MVV2011]_.
In the 1970s, Diffie and Hellman proposed in their seminal paper [DH1976]_, a different type of encryption : `public key cryptography`. In public key cryptography, each user has two different keys :
a public key (:math:`K_{pub}`) that he can distribute to everyone
a private key (:math:`K_{priv}`) that he needs to store in a secure manner and never reveal to anyone
These two keys are generated together and they are linked by a complex mathematical relationship that is such that it is computationally difficult to compute :math:`K_{priv}` from :math:`K_{pub}`.
A public key cryptographic scheme is a combination of two functions :
The public key is used to encrypt a message so that it can only be read by the intended recipient. For example, let us consider two users : Alice and Bob. Alice (resp. Bob) uses the keys :math:`A_{priv}` and :math:`A_{pub}` (resp. :math:`B_{priv}` and :math:`B_{pub}`). To send a secure message `M` to Alice, Bob computes :math:`CM=E_p(A_{pub},M)` and Alice can decrypt it by using :math:`D_p(A_{priv},CM)=D_p(A_{priv},E_p(A_{pub},M))=M`.
Several public key encryption schemes have been proposed. Two of them have reached wide deployment :
The Rivest Shamir Adleman (RSA) algorithm [#frsa]_ proposed in [RSA1978]_ that relies on modular exponentiation with large integers.
The Elliptic Curve Cryptography techniques [#fecc]_ that rely on special properties of elliptic curves.
Another interesting property of public key cryptography is its ability to compute `signatures` that can be used to authenticate a message. This capability comes from the utilization of two different keys that are linked together. If Alice wants to sign a message `M`, she can compute :math:`SM=E_p(A_{priv},M)`. Anyone who receives this signed messaged can extract its content as :math:`D_p(A_{pub},SM)=D_p(A_{pub},E_p(A_{priv},M))=M`. Everyone can use :math:`A_{pub}` to check that the message was signed by using Alice's private key (:math:`A_{priv}`). Since this key is only known by Alice, the ability to decrypt `SM` is a proof that the message was signed by Alice herself.
In practice, encrypting a message to sign it can be computationally costly, in particular if the message is a large file. A faster solution would be to summarize the document and only sign the summary of the document. A naive approach could be based on a checksum or CRC computed over the message. Alice would then compute :math:`C=Checksum(M)` and :math:`SC=E_p(A_{priv},C)`. She would then send both `M` and `SC` to the recipient of the message who can easily compute `C` from `SC` and verify the authenticity of the message. Unfortunately, this solution does not protect Alice and the message's recipient against a man-in-the-middle attack. If Mallory can intercept the message sent by Alice, he can easily modify Alice's message and tweak it so that it has the same checksum as the original one. The CRCs, although more complex to compute, suffer from the same problem.