This paper reviews "A Mathematical Theory of Communication" by Shannon, introducing information theory. It discusses the concept of information entropy, ergodic sources, and channel capacity. The paper shows that error probability can be minimized without reducing data rate, and that there is an upper bound on data rate for reliable transmission. The paper is divided into five parts, but the authors reorganize it into four. The first part covers preliminary concepts like entropy, ergodicity, and channel capacity. Entropy measures uncertainty, ergodic sources allow the application of the AEP theorem, and channel capacity measures information transfer capability. The second part discusses discrete sources and channels, proving that if a source's entropy is less than or equal to the channel capacity, error-free transmission is possible. The third part addresses continuous channels, where capacity is defined using differential entropy. The fourth part deals with continuous sources and channels, introducing fidelity as a measure of transmission accuracy. The paper also discusses practical encoding schemes, such as Turbo and LDPC codes, which approach Shannon's limit but require large block sizes. The authors also explore the assumption of ergodic processes and note that non-ergodic sources can still satisfy the AEP property. The paper concludes that entropy and ergodicity are fundamental to information theory, and that Shannon's work remains foundational despite advancements in the field.This paper reviews "A Mathematical Theory of Communication" by Shannon, introducing information theory. It discusses the concept of information entropy, ergodic sources, and channel capacity. The paper shows that error probability can be minimized without reducing data rate, and that there is an upper bound on data rate for reliable transmission. The paper is divided into five parts, but the authors reorganize it into four. The first part covers preliminary concepts like entropy, ergodicity, and channel capacity. Entropy measures uncertainty, ergodic sources allow the application of the AEP theorem, and channel capacity measures information transfer capability. The second part discusses discrete sources and channels, proving that if a source's entropy is less than or equal to the channel capacity, error-free transmission is possible. The third part addresses continuous channels, where capacity is defined using differential entropy. The fourth part deals with continuous sources and channels, introducing fidelity as a measure of transmission accuracy. The paper also discusses practical encoding schemes, such as Turbo and LDPC codes, which approach Shannon's limit but require large block sizes. The authors also explore the assumption of ergodic processes and note that non-ergodic sources can still satisfy the AEP property. The paper concludes that entropy and ergodicity are fundamental to information theory, and that Shannon's work remains foundational despite advancements in the field.