Spectral Efficiency of Millimeter Wave Multiple-Input-Multiple-Output System for Different Precoding Techniques

The search for new frequency bands and cutting-edge communication technologies has been driven by the rapid growth of wireless data traffic and the increasing demand for faster transmission speeds. Frequency bands in the RF region are utilized for various purposes, including satellite communication, radio communications, television broadcasting, and more. Being high-band fifth-generation (5G), this range works best across shorter distances in highly dense environments. 5G mmWave offers the fastest speeds and the maximum capacity, enabling advanced technologies such as automated cars.

The two main subjects of the feasibility of mmWave spectrum are its ability to support users on the go and its reach. On both points, new developments and antenna technologies are pushing the envelope. For instance, autonomous cars can communicate with other vehicles using the 5G mmWave frequency to steer clear of traffic and obstacles. Because it can support very high data rates and large bandwidths, mmWave communication, which operates in the 30–300 GHz range, has emerged as a promising contender for future wireless systems [1-7]. Rappaport et al. [8] discussed the feasibility of using mmWave communication for mobile applications, highlighting the significant path loss and the requirement for directional beamforming to achieve system reliability. El Ayach et al. [9] examined the basis pursuit, which can be used to construct precoders that are limited by hardware constraints and can be applied by realistic mmWave transceivers. According to Payami et al. [10], a fully digital precoding provides the most flexibility by regulating the amplitude and phase of the transmitted signals at each antenna.

However, because it requires numerous costly and power-hungry RF chains, its application at mmWave wavelengths is challenging. Analog and hybrid precoding systems were developed to overcome these constraints. Alkhateeb and Heath [11] suggested that such precoders offer less control over spatial multiplexing, while analog precoding uses a network of phase shifters to reduce hardware complexity. Hybrid precoding architectures strike a balance between performance and complexity by integrating analog beamforming with a limited number of RF links. Du et al. [12] proposed a convolutional neural network-based precoder to enhance SE and energy efficiency with minimal delay. Ma et al. [13] proposed a deep learning-based CE and feedback scheme for reducing overhead in mmWave MIMO systems. Kebede eta al. [14] have discussed the impact of fully connected and partially connected hybrid architecture on the performance of the mmWave MIMO system. Albreem et al. [15] have proposed a machine learning-based non-linear precoding scheme to estimate features such as the peak-to-average power ratio and constant envelope. Reddy et al. [16] have reviewed various algorithms for CE and hybrid precoding to analyze the performance of the latest communication technologies. Khalid et al. [17] have suggested that selecting a transmit antenna using an optimized algorithm based on probability distribution learning can yield a solution that provides selective antennas in real-time, thereby enhancing SE. Samir et al. [18] have estimated and proposed a low-complexity performance for mmWave MIMO systems using orthogonal matching pursuit and MMSE precoding techniques. Liu et al. [19] have optimized the sum SE using discrete Fourier transform in digital and analog precoders. Alkhateeb et al. [20] have compared precoding techniques at low and mmWave frequencies. Based on the comparison, hybrid analog-to-digital precoding techniques have been proposed. Yu et al. [21] have proposed a hybrid precoder design optimization that can be applied to both fully connected and partially connected architectures, utilizing an effective alternating minimization process. This approach demonstrates a performance gain over current hybrid precoding schemes. Businesses can now achieve unparalleled performance and network management with an independent 5G system. 5G mmWave is essential for applications that can benefit from connectivity via public mobile networks, as it offers focused, high-performance service for moving cars, customers, and devices throughout urban areas. Although computer aficionados tend to experience more problems, perhaps due to their higher expectations, there is a slight variation in how each views the utility of connectivity in crowded areas. It is crucial to acknowledge that each group places a higher value on packed area connectivity than coverage and indoor connectivity. Guo et al. [22] have proposed a robust sum mean square error scheme for designing precoders, which shows a considerable improvement in the form of reduced bit error rate for low transmission power. Shahjehan et al. [23] thoroughly analyzed the basis of scattering, network, link, and codebook for the design optimization of mmWave MIMO systems.

When it comes to meeting increased data demands, mmWave spectrum can be quite helpful. However, even if mmWave 5G can provide high-end experiences, its capacity to generate income must also be taken into account. Kabalci et al. [24] have proposed a decomposition technique based on the geometric mean, which provides a significant improvement in the performance of mmWave MIMO systems.

Jafri et al. [25] have proposed hybrid beam formers based on asynchronous distribution to reduce transmission power, even when channel state information is uncertain at the transmission end.

SE is the effective transmission of the data using the allocated bandwidth. Bandwidth is a valuable resource, and all commercial wireless applications must adhere to its usage as per the guidelines set by the International Telecommunication Union. In a mmWave MIMO system, the spatial multiplexing provides the flexibility of using the same bandwidth by multiple users. If the information of such users is separated with minimum and acceptable errors, the throughput is said to be achieved. This results in a spectrally efficient system. Hence, the SE is a crucial factor examined in this work for the effective reception of signals. Since a lesser number of RF chains is employed compared to the number of users, efficient precoding techniques are required to achieve the SE, which in turn provides massive connectivity.

The increase in SE would benefit practical applications that require high capacity and very low latency, such as mobile communication scenarios, where the actual network is connected to small cells. Furthermore, indoor wireless access can be provided with very high data rates, which can match the data rates of wired communication systems. Furthermore, vehicle-to-vehicle communication, where line-of-sight propagation is essential, is a key outcome of the mmWave MIMO system with high SE.

The objectives of the study:

1. The SE examination of the mmWave MIMO system has been done for the following precoding techniques:

a. Analog beam steering, which utilizes a single RF chain and achieves the least SE.

b. ZF hybrid and fully digital precoding, which is suited for low SNR but suffers from noise amplification.

c. MMSE hybrid and fully digital precoding, which outperforms both even at high SNR.

2. SE variation with SNR has been examined for a variable user base and number of transmit and receive antennas.

The simulations in this study limit the number of users to 10, while varying the initial value from 5 for the fixed number of transmit and receive antennas. The number of transmit and receive antennas is fixed to 16.

Thereafter, the number of transmit and receive antennas is varied from 25 to 36 for a fixed number of users. Further, the CE employed in this work is based on directional cosine vectors at the transmitter and receiver ends. A significantly less sparse channel is considered, which results in fewer multipath components.

Contributions of the paper:

1. An improvement in the SE of mmWave MIMO systems has been presented, considering both hybrid and fully digital precoders.

2. The effect of the variation in the user base on the SE has been analyzed and shown for a fixed number of transmit and receive antennas.

3. The effect of varying the number of transmit and receive antennas on the SE has been analyzed and illustrated for a fixed number of users.

The mmWave MIMO system promises ultra-high-speed communication in the backhaul, where a highly dense and cluttered environment is considered. ZF and MMSE precoding techniques have been considered in both fully digital and hybrid modes. The performance of these precoders is compared with that of a single user with no interference. Furthermore, the performance of the standalone analog beamforming technique is also demonstrated. It has been shown that for a fixed number of users, the SE of the fully digital MMSE precoder is better than that of its ZF counterpart. However, due to practical implementation constraints, hybrid MMSE can be employed when the number of transmit and receive antennas increases. However, when the user base increases with the fixed number of transmit and receive antennas, it has been shown that overall SE reduces. The key outcome is the improvement in the performance of the hybrid MMSE precoder over the fully digital ZF precoder, leading to a further increase in the user base.