Some of the recent literature related to the presented works are listed below,
Qinwei He et al [21] suggested a closed-form-based Symbol Error Rate (SER) derivative to estimate the reliability performances of the NOMA system. In this model, the reliability performance of the NOMA system with the Quadrature Amplitude Model (QAM) and Pulse Amplitude Modulation (PAM) are analyzed. Furthermore, the power allocation technique is introduced to decelerate the SER of the user. However, the suboptimal approach used in the model does not provide an effective solution.
In recent times, OFDM multicarrier modulation approach is used in wireless communication systems especially in 5th -generation systems for providing better spectral density. The major challenge in OFDM modulation is high PAPR which occurs when the nonlinear power amplifiers are operated near the saturation region. Thus, Khaled Tahkoubit et al [22] presented an efficient PAPR reduction model named Iterative Dichotomy. This method used analytical derivatives for decelerating the PAPR range. However, the implementation complexity is more.
In communication networks, NOMA is widely used to enhance network connectivity, and stability. However, the high error rate in NOMA demands an efficient channel estimation technique. Thus, Hamad Yahya et al [23] proposed a power-tolerant NOMA approach, which decelerates the network sensitivity by adaptively varying the signal power of the users. Although this scheme minimizes the error rate by 10dB, it is applicable for multicarrier modulation.
Saud Althunibat et al [24] presented an uplink multiple access approaches named Index modulation multiple access (IMMA). Here, each user's transmitted blocks are used to modulate a complex symbol, which is then transferred to a particular time slot. This presented model reduces the system complexity and error rate in the NOMA system. Furthermore, the designed model is validated with a comparative analysis. But, the probability of collision is high in this model.
Compared to OMA networks, NOMA provides greater spectral and network efficiency to multi-user communication systems. The existing techniques like Rayleigh fading channels cause low performances in the NOMA system. Therefore, Kuang-Hao Liu et al [25] developed a Quasi-degradation probability of 2-user NOMA over the Rician fading channel. This technique improves the system's performance and enhances network stability. However, the complexity and cost are high in this approach.
Generally, to minimize the MIMO dimension Single Beam selection approach is used in networks. But the user cannot properly select the multiple Radio Frequency (RF) chain, groups. Hence, Satyanarayana Murthy Nimmagadda [26] developed a hybrid based on the two optimization approaches named Alternative Grey Wolf with Beetle Swarm Optimization. This hybrid system acts as a multi-objective function to solve challenges regarding energy consumption, and power efficiency. However, it does not support future wireless communication systems.
Jawad Mirzaś et al [27] presented a two-stage channel estimation technique to overcome the issue with the traditional Reconfigurable Intelligent Surface (RIS) channels. The conventional RIS channels face ill-conditioned dictionary learning issues. The presented model utilizes a bilinear adaptive vector approximate message passing method to determine the RIS channels. This algorithm helps in identifying the ill-conditioned dictionary problems accurately. Moreover, the estimate of the phase shift matrix an approximate closed-form function is deployed in the system. However, the algorithm complexity and time consumption are more in this model.
In cellular networks, Intelligent Reflecting Surface (IRS) is widely used to improve the wireless propagation environment. However, the tuning of phase shifters in IRS is very complex. Thus, Tao Jiang et al [28] designed a machine-learning technique that tunes the beam formers optimally. This method deploys a deep neural network algorithm to tune the phase shifters optimally. In addition, a Graph neural network is utilized to capture the interrelations between the users. However, the error rate is more in this approach.
In 5th generation and beyond networks, IRS aided system is widely used to accelerate spectral power and energy efficiency. But the noise randomness is one of the major challenges in the IRS system. Therefore, Anastasios Papazafeiropoulos et al [29] developed a channel estimation method for reducing the computational complexity. The introduced technique includes closed-form derivative and low training overhead. Although the presented technique achieved better spectral efficiency, it does not allow for performing optimization which makes the IRS ineffective.
The major task in the IRS-aided communication system is finding the optimal channel estimation approach. Typically, the IRS system deploys a parallel channel with a sophisticated statistical distribution. The problem with the IRS communication system is a denoising issue. Hence, Chang Liu et al [30] designed a deep residual learning algorithm to retrieve the channel coefficients from the noisy parameters. Furthermore, the simulation outcomes of the developed model show better performances in minimizing the noise effects. Still, it does not provide the optimal solution for the denoising problem.