In recent times, the landscape of wireless communication systems has witnessed significant expansion, driven by escalating demands for enhanced data rates, heightened reliability, and seamless connectivity. Noteworthy emphasis has been placed on pivotal technologies underpinning 5G and Beyond 5G systems, encompassing small cells, device-centric architectures, beamforming, full-duplex technology, massive multiple-input multiple-output (MIMO), millimeter (mm), and terahertz (THz) waves, non-orthogonal multiple access (NOMA), and reconfigurable intelligent surfaces (RIS). Notably, the bandwidth potential offered by millimeter waves surpasses tenfold the cumulative bandwidth of the entire 4G cellular spectrum [1].
The advent of 5G communication systems has catalysed an upsurge in demand for efficient multiple access (MA) techniques capable of sustaining high data rates and accommodating burgeoning user populations [2]. Across successive generations of mobile communication, MA strategies empower users to concurrently leverage time and/or spectrum resources while preserving the quality-of-service (QoS) they receive. These strategies encompass a spectrum of approaches, including frequency-division multiple access (FDMA), time-division multiple access (TDMA), orthogonal frequency-division multiple access (OFDMA), code-division multiple access (CDMA), and non-orthogonal multiple access (NOMA), among others, deployed across different wireless communication epochs to allocate spectra to diverse users. Figure 1 provides a comparative delineation of distinct MA techniques across varied wireless generations [2–4].
Orthogonal frequency-division multiplexing (OFDM) has long been the preferred modulation scheme due to its resilience against frequency-selective fading channels. However, OFDM faces challenges like high out-of-band emissions and suboptimal spectral efficiency, particularly in scenarios with numerous users and limited spectrum availability. To address these challenges, various non-orthogonal multiple access (NOMA) schemes have been proposed, categorized into power-domain (PD) NOMA and code-domain (CD) NOMA. Unlike traditional approaches in 3G and 4G cellular networks, NOMA eliminates orthogonality, enabling multiple users to share frequency and time resources within the same spatial layer through power or code domain multiplexing. The concept of power-domain NOMA was introduced by Y. Saito et al [5].
PD-NOMA, a prominent multiple access technique in 5G networks, enables multiple users to utilize the same radio resource in the power domain. Resource allocation is critical in PD-NOMA, involving the optimal assignment of available radio resources to users. Various studies have proposed different resource allocation schemes for PD-NOMA. For example, authors in [5] suggested a joint power and subcarrier allocation scheme maximizing the sum rate while meeting users' quality of service (QoS) requirements. Similarly, [6] proposed a user grouping and resource allocation scheme considering both channel quality and QoS requirements to enhance PD-NOMA performance.
In PD-NOMA, users are segregated based on their power levels, with each user assigned a unique power level corresponding to their channel quality. Users with superior channel conditions receive lower power levels, while those with inferior conditions are allocated higher power levels. Various researchers, including Y. Saito et al. [5], Z. Ding et al. [6], have explored the implementation of power-domain NOMA. Additionally, the spectral efficiency gains of NOMA have been well established.
Furthermore, recent works have introduced innovative concepts to enhance device connectivity in NOMA systems. For instance, Z. Yuan et al. [7] proposed methods to increase the number of connected devices, while Saito et al. [17] demonstrated the applicability of basic NOMA with a Successive Interference Cancellation (SIC) receiver in the uplink. Linglong Dai et al. [8] compared various NOMA schemes for 5G and suggested that the number of connected devices could be increased at the expense of receiver complexity. Hence, NOMA emerges as a promising choice to achieve higher spectral efficiency and facilitate massive connectivity in 5G networks.
Power Domain Non-Orthogonal Multiple Access (PD-NOMA) emerges as a promising approach for 5G networks, offering potential improvements in overall system capacity and user quality of service, especially in densely populated urban areas with high data traffic demands. In the framework of PD-NOMA, key functions at the transmitter end involve power allocation and superposition coding, while Successive Interference Cancellation (SIC) plays a crucial role at the receiver end. Let's consider a scenario of downlink transmission from the Base Station (BS) to two separate users, where d1 and d2 denote the distances of these users from the BS. The BS transmits two distinct messages, x1 and x2, intended for user 1 (far user) and user 2 (near user) respectively. Power allocation factors α1 and α2 (with α1 + α2 = 1) are assigned to user 1 and user 2, while fading coefficients h1 and h2 characterize the channel from the BS to each user. To ensure fairness among users, a strategy known as fair power allocation (PA), based on channel state information (CSI), was introduced [9–10]. In case of multiple user, many time, the fair power allocation scheme is not optimum due to dynamics of wireless channels.
After the superposition coding, transmitted NOMA signal by the BS is,
$$x=\sqrt P (\sqrt {{\alpha _1}} {x_1}+\sqrt {{\alpha _2}} {x_2})$$
1
Where P is the transmitted power received at the user1 is given by
The received signal at the user2 after propagating through Rayleigh fading channel h2 is given by
Where w1 and w2 represent white Gaussian noises. The signals y1 and y2 are the received signals. The message signals are extracted using the SIC, followed by the calculation of the BER. The BER is a function of signal power and fading through the channel. The scheme of SIC is shown in Fig. 3.
The capacity achieved by SIC for 2 users in PD-NOMA is depends on power assigned and the wireless channel. For far user the capacity Rf is given by [11]
\({R}_{f}={log}_{2}(1+\frac{{\left|{h}_{f}\right|}^{2}P{\alpha }_{f}}{{\left|{h}_{f}\right|}^{2}P{\alpha }_{n}+{\sigma }^{2}}\) ) (4)
Where P is the total power. Normally higher power (P. αf ) assigned to far user compensate the attenuation of the wireless channel. For near user the capacity Rn is given by
\({R}_{n}={log}_{2}(1+\frac{{\left|{h}_{n}\right|}^{2}P{\alpha }_{n}}{{\sigma }^{2}}\) ) (5)
Where, αf and αn are the power allocation coefficient for far and near user respectively. The sum of these coefficients is 1. Hence the optimum power allocation is essential for faithful communication. In this paper, PD-UFMC with optimum power allocation is simulated for performance analysis.