Sparse Code Multiple Access (SCMA) has proved to be a fascinating research in order to curtail the complications faced by the wireless communication networks. SCMA being a Non-Orthogonal Multiple Access technique evinces to be an outstanding candidate, to cater the complications faced by 5G communication networks to improve the bit error rate and reduce the complexity of decoding the transmitted signal from received signal. This paper explains the concept of SCMA by explaining the basic structure of encoder, decoder and codebook design with the help of neural networks. It explains the concept of reducing the complexity of the traditional decoder of the SCMA by implementing Neural Networks. Further sections explain the use of Convolutional Neural Networks for blind decoding, that outperforms the complexity of decoding carried by conventional SCMA using Message Passing Algorithm. This further explains the use of Deep Neural Networks for designing the codebook and decoding it, by adopting an autoencoder structure.

为了减少无线通信网络面临的复杂性，稀疏代码多址访问（SCMA）已被证明是一项引人入胜的研究。SCMA是一种非正交多址技术，因此有望成为出色的候选者，以应对5G通信网络面临的复杂情况，从而提高误码率并降低从接收信号中解码发送信号的复杂性。本文通过借助神经网络解释编码器，解码器和码本设计的基本结构来解释SCMA的概念。它解释了通过实现神经网络来降低SCMA传统解码器复杂性的概念。进一步的章节将说明使用卷积神经网络进行盲解码，胜过了传统SCMA使用消息传递算法进行解码的复杂性。通过采用自动编码器结构，这进一步说明了深度神经网络在设计码本和对其进行解码方面的用途。