Definitions:
A communication system, as illustrated in the following
figure, is modeled as a data source providing either
continuous or discrete output, a source encoder transforming
source data into binary digits (bits), a channel encoder,
and a channel.

In many communication systems the channel is analog,
that is, continuous in amplitude and time, in which case a
modulator/demodulator (modem) is required to transform between analog channel data and discrete encoder/decoder data.
The source encoder, the aim of which is to minimize the number of bits
required to represent source data while still allowing subsequent
reconstruction, typically includes data compression to remove
unnecessary redundancy.
The objective of the channel encoder is to maximize the
rate at which information can be reliably conveyed by the channel, in the
presence of disruptive channel noise.
Coding theory is the study of the
translation between source data representations and the corresponding
representative symbols (coded data) used to transmit source data over a
communication channel.
Error-correction coding, located in the
channel encoder, adds systematic redundancy to messages to allow transmission
errors not only to be detected but also to be corrected.
Encryption, located after the source
encoder but not always before the channel encoder, is designed to render data
unintelligible to all but the intended recipient, and thereby preserve the secrecy
of messages in the presence of unfriendly monitoring of the channel.
Cryptography is the science of maintaining
secrecy of data, that is, protecting data from malicious or unauthorized
actions, including passive intrusion (eavesdropping) and active
intrusion (injection or modification of messages).
Authentication is corroboration that a party,
or the origin of a message, is as claimed.
Data integrity is the property that data have
not been modified in an unauthorized manner.
Non-repudiation is the preclusion of parties
from making undetectable false denials.