A wireless communication system employs a variety of data compression, encoding, and modulation techniques to efficiently transmit messages via communication channels, aiming to reproduce the information at the receiver with the fewest possible errors. To counteract the impact of noise and interference encountered by the signal during its journey through the communication channel, the channel encoder introduces redundancy to the binary information sequence. The channel decoder at the receiver utilizes this redundancy to combat errors. To enhance data redundancy, the channel encoder utilizes errorcorrecting codes, including block codes, convolutional codes, Low-Density Parity Check (LDPC) codes, and turbo codes. These coding methods play a crucial role in error detection and correction. However, the configuration of a wireless communication system can now be simplified by leveraging Deep Neural Networks (DNNs). This streamlined communication system can be conceptualized as a specific type of autoencoder in the realm of Deep Learning (DL). The primary goal of this research is to develop an autoencoder model for Additive White Gaussian Noise (AWGN) and fading channels with a low error probability, ensuring reliable communication.