Non-Orthogonal Multiple Access (NOMA)

The base station transmits a composite signal that is a weighted sum of individual user signals. Users with poorer channel conditions are allocated higher power levels, while users with stronger channel conditions receive lower power. This intentional power imbalance creates a structured interference pattern that can be systematically resolved at the receiver, maximizing the total sum-rate capacity of the cell.

The cornerstone decoding technique that makes NOMA practical. The receiver treats stronger signals as noise while decoding the highest-power user first. Once decoded, that user's signal is subtracted from the composite waveform, and the process repeats for the next strongest signal. This iterative peeling of interference layers allows each user to recover its intended data despite the non-orthogonal superposition.

A variant designed for massive machine-type communications (mMTC) where devices transmit sporadically with small payloads. Grant-free NOMA eliminates the traditional handshaking and scheduling request process, allowing devices to transmit immediately on shared resources. The base station uses compressed sensing and multi-user detection to identify active devices and decode their overlapping transmissions, dramatically reducing latency and signaling overhead for billions of IoT endpoints.

Exploits the fact that strong-channel users in a NOMA cluster already decode the messages of weaker users during the SIC process. The strong user can act as a decode-and-forward relay, re-transmitting the weak user's data to improve reliability. This creates a virtual distributed antenna system without additional infrastructure, enhancing cell-edge throughput and outage probability through user cooperation.

The fundamental receiver-side signal processing technique that makes NOMA practical. SIC decodes the strongest user's signal first, subtracts it from the received composite waveform, then proceeds to decode the next strongest signal. This iterative subtraction requires precise channel estimation and is computationally intensive, but it is the mechanism that unlocks the capacity gains of power-domain multiplexing. Without SIC, NOMA collapses to a standard interference-limited system.

The core multiplexing dimension exploited by NOMA, distinct from time, frequency, or code. Users are assigned different transmission power levels based on their channel conditions: cell-edge users receive higher power allocations, while cell-center users transmit at lower power. This intentional power imbalance creates the signal-to-interference-plus-noise ratio (SINR) disparity that SIC exploits. Power allocation algorithms are the central optimization problem in NOMA system design.

The baseline against which NOMA is measured. OMA schemes—including OFDMA in 4G LTE and 5G NR—allocate orthogonal time-frequency resource blocks to users, ensuring no intra-cell interference. While simpler to implement, OMA is suboptimal in terms of spectral efficiency and user fairness. NOMA's theoretical advantage over OMA is most pronounced when users have highly asymmetric channel gains.

An uplink transmission paradigm where users transmit data without first requesting and receiving a scheduling grant from the base station. This eliminates the latency overhead of the handshake process, making it ideal for massive Machine-Type Communications (mMTC) with sporadic small packets. Grant-free NOMA relies on advanced multi-user detection and contention resolution algorithms at the receiver to separate overlapping transmissions.

A code-domain NOMA variant that maps user bit streams directly to multi-dimensional sparse codewords from a predefined codebook. The sparsity of the codewords reduces the complexity of message passing algorithm (MPA) detection at the receiver. SCMA offers shaping gain in addition to the overloading gain of NOMA, making it a strong candidate for 5G and beyond.

A spatial-domain multiplexing technique that serves multiple users on the same time-frequency resource using multiple antennas and precoding. When combined with NOMA, the resulting MIMO-NOMA architecture clusters users into groups—NOMA is applied within each cluster, while spatial multiplexing separates clusters. This hybrid approach multiplies spectral efficiency gains but demands highly accurate channel state information at the transmitter (CSIT).