Cooperative NOMA: Enhancing Wireless Networks

• Cooperative NOMA is an advanced multiple access scheme that uses superposition coding and successive interference cancellation, allowing strong users to act as relays.
• It employs both decode-and-forward and amplify-and-forward relay strategies to provide full-diversity gains and enhance reliability across varying channel conditions.
• The scheme significantly improves outage probability and sum-rate performance, making it ideal for applications in 5G, 6G, IoT, and ultra-reliable low-latency communications.

Cooperative Non-Orthogonal Multiple Access (NOMA) is an advanced multiple-access scheme for wireless communication systems that leverages superposition coding and user cooperation to enhance spectral efficiency, reliability, and fairness. By exploiting the hierarchical decoding capability through successive interference cancellation (SIC) at users with stronger channel conditions, cooperative NOMA enables these users to act as relays, forwarding information to weaker users and thereby providing full-diversity gain across the network. This paradigm has direct relevance to 5G, 6G, massive IoT, and future ultra-reliable low-latency communication (URLLC) scenarios, where efficient and robust multiple access is crucial (Ding et al., 2014, Wan et al., 2018, Salim et al., 22 May 2025).

1. System Model and Cooperative Transmission Protocols

In a canonical downlink cooperative NOMA system, a base station communicates with $K$ users whose channel gains are ordered as $|h_1|^2 \leq |h_2|^2 \leq \dots \leq |h_K|^2$. Transmission employs power-domain superposition coding: $x = \sum_{m=1}^K \sqrt{p_m^2 P} \cdot s_m,$ where $p_1^2 \geq p_2^2 \geq \cdots \geq p_K^2$ and $\sum_{m=1}^K p_m^2 = 1$ allocate increasing power to weaker users. Reception at each user follows: $$y_{1,k} = h_k x + n_k, \quad n_k \sim \mathcal{CN}(0,1).$$ User $K$ (strongest) uses SIC to sequentially decode $s_m$ ($m=1$ to $K$).

Cooperative NOMA augments this with a relay phase where stronger users, after decoding, serve as decode-and-forward relays for weaker users. The general protocol comprises:

• Phase 1: Direct transmission from the base station (all users listen; strongest use SIC).
• Phase 2: $|h_1|^2 \leq |h_2|^2 \leq \dots \leq |h_K|^2$0 short-range cooperative slots, with each of the $|h_1|^2 \leq |h_2|^2 \leq \dots \leq |h_K|^2$1 strongest users relaying the undecoded messages of weaker users (using their own power allocation vectors), so that each weak user obtains $|h_1|^2 \leq |h_2|^2 \leq \dots \leq |h_K|^2$2 independent spatial paths (Ding et al., 2014).

Associated relaying protocols include:

• Decode-and-Forward (DF): Relay decodes all intended messages via SIC, re-encodes and forwards them.
• Amplify-and-Forward (AF): Relay amplifies and retransmits its noisy observation; SIC is performed at the user (Wan et al., 2018, Salim et al., 22 May 2025).

Variations such as full-duplex relay operation, wireless-powered relays, user pairing, and selection strategies further expand the architecture.

2. Diversity, Outage Probability, and Sum-Rate Analysis

The diversity order is a central metric quantifying reliability improvements. In canonical cooperative NOMA: $|h_1|^2 \leq |h_2|^2 \leq \dots \leq |h_K|^2$3 where $|h_1|^2 \leq |h_2|^2 \leq \dots \leq |h_K|^2$4 is the outage probability for user $|h_1|^2 \leq |h_2|^2 \leq \dots \leq |h_K|^2$5 and $|h_1|^2 \leq |h_2|^2 \leq \dots \leq |h_K|^2$6 is transmit SNR.

Fundamental results for $|h_1|^2 \leq |h_2|^2 \leq \dots \leq |h_K|^2$7 users with Rayleigh fading show that every user achieves diversity order $|h_1|^2 \leq |h_2|^2 \leq \dots \leq |h_K|^2$8—full system diversity—under the cooperative protocol (Ding et al., 2014). In contrast, non-cooperative NOMA only yields diversity equal to the user's decoding index. For $|h_1|^2 \leq |h_2|^2 \leq \dots \leq |h_K|^2$9 and ideal conditions, cooperative NOMA attains a slope-2 reduction in outage probability (in dB) scaling, representing substantial reliability improvement over both OMA and non-cooperative NOMA schemes.

Outage probability for decoding at user $x = \sum_{m=1}^K \sqrt{p_m^2 P} \cdot s_m,$0 at rate $x = \sum_{m=1}^K \sqrt{p_m^2 P} \cdot s_m,$1 (target SNR $x = \sum_{m=1}^K \sqrt{p_m^2 P} \cdot s_m,$2) can be approximated, in high SNR, as

$x = \sum_{m=1}^K \sqrt{p_m^2 P} \cdot s_m,$3

under appropriate power split constraints and error thresholds (Ding et al., 2014). This scaling is robust under extensions to more general fading models, e.g., Fisher–Snedecor F or $x = \sum_{m=1}^K \sqrt{p_m^2 P} \cdot s_m,$4–$x = \sum_{m=1}^K \sqrt{p_m^2 P} \cdot s_m,$5 fading (Rabie et al., 2020, Kumar et al., 2019).

Sum-rate analysis reveals that, when properly pairing users (strongest with weakest), cooperative NOMA can nearly double the spectral efficiency gap versus OMA at high SNR (e.g., at 15 dB and $x = \sum_{m=1}^K \sqrt{p_m^2 P} \cdot s_m,$6 outage: OMA $x = \sum_{m=1}^K \sqrt{p_m^2 P} \cdot s_m,$7 BPCU, NOMA $x = \sum_{m=1}^K \sqrt{p_m^2 P} \cdot s_m,$8 BPCU, cooperative NOMA $x = \sum_{m=1}^K \sqrt{p_m^2 P} \cdot s_m,$9 BPCU) (Ding et al., 2014).

3. User Pairing, Scheduling, and Resource Optimization

Given the exponential overhead of applying cooperative NOMA across all $p_1^2 \geq p_2^2 \geq \cdots \geq p_K^2$0 users, practical systems adopt user grouping or pairing.

• Strongest–weakest pairing maximizes the SE gain (sum-rate gap scales with $p_1^2 \geq p_2^2 \geq \cdots \geq p_K^2$1 for a pair $p_1^2 \geq p_2^2 \geq \cdots \geq p_K^2$2).
• Hybrid NOMA: The network forms multiple groups (typically pairs), applies cooperative NOMA within the group, and allocates orthogonal resources between groups (e.g., TDMA/FDMA) (Ding et al., 2014, Khraimech et al., 2022).

Resource allocation is a highly nontrivial problem due to the need to jointly determine power coefficients $p_1^2 \geq p_2^2 \geq \cdots \geq p_K^2$3, group assignments, and relay scheduling (within practical constraints on total transmit and relay power). For composite (uplink+downlink) architectures, a hybrid power allocation strategy is effective:

• Offline (long-term): Find the optimal source power split based on statistical CSI.
• Online (per slot): Adapt relay power split to instantaneous CSI, reducing CSI overhead while retaining most of the sum-rate of full-ICSI designs (Wan et al., 2018).

Spectral efficiency is maximized when user pairs have large channel asymmetry, while fairness and reliability are optimized by targeting min–max achievable rates and fine-tuning power allocation within each group.

4. Relay Strategies: Physical Layer Paradigms and Practical Extensions

Cooperative NOMA realizes different relaying strategies depending on network architecture and node capability:

• DF relaying: Canonical in most cooperative NOMA works (Ding et al., 2014, Wan et al., 2018). Requires each relay/user to decode all prior messages, generally via SIC.
• AF relaying: Useful when user-to-user links are absent or unreliable. An independent relay node forwards signals, with users using combining schemes (e.g., selection combining or MRC) for detection (Tran et al., 2018, Khraimech et al., 2022).
• Full-duplex multi-antenna relaying: Offers spectral efficiency doubling, with zero-forcing or maximal-ratio beamforming at relay for SI suppression, and careful spatial user selection for diversity enhancement (Mobini et al., 2017, Salim et al., 22 May 2025).
• Energy-harvesting and SWIPT-based relays: Users harvest energy from the base station via power splitting or time switching, then serve as relays for far users (Liu et al., 2015). Energy-neutral operation can be achieved without compromising diversity order, provided user selection adapts to spatial locations.
• Backscatter cooperation: Passive backscatter devices re-transmit part of a decoded signal to provide additional diversity to the weak user without extra RF transmit cost, offering both diversity and throughput improvements (Chen et al., 2020).
• Asynchronous cooperative NOMA (C-ANOMA): Intentionally introduces symbol timing mismatch and receiver oversampling, leveraging sampling diversity and reducing required transmit power for given QoS constraints (Zou et al., 2019).

5. Design Trade-offs, Performance Bottlenecks, and Optimization

While cooperative NOMA yields strong performance gains, several challenges and trade-offs govern system design:

• SIC complexity and error propagation: High-order SIC at relays/users increases hardware and algorithmic complexity; decoding errors in SIC chains can cascade (Wan et al., 2018).
• CSI acquisition and feedback: Full CSI is often impractical; hybrid or statistical CSI-based protocols can mitigate feedback overhead while maintaining near-optimal rates (Wan et al., 2018, Salim et al., 22 May 2025).
• Power allocation: Proper split between users is critical; in some scenarios, reversing the usual NOMA split (strong user receives more power) minimizes mutual outage probability while maintaining full diversity (Riaz et al., 2020).
• Security and privacy: Cooperative (especially relay-aided and full-duplex) operation raises susceptibility to eavesdropping unless combined with secure beamforming and jamming strategies (Wan et al., 2018, Salim et al., 22 May 2025).
• Cross-layer and cognitive integration: Merging cooperative NOMA with energy-harvesting, cognitive radio, RIS, and semantic communication broadens the design space and can utilize deep learning for adapting scheduling, beamforming, and resource allocation in real-time (Salim et al., 22 May 2025).
• User selection and pairing: Near–far pairing is optimal for SE but needs dynamic adaptation to channel variations, user mobility, and topology (Khraimech et al., 2022, Liu et al., 2015).

Table: Summary of Core Benefits and Challenges

Aspect	Cooperative NOMA	Key Challenges
Diversity	Full diversity order for all users	CSI, relay/channel feedback overhead
Spectral efficiency	Approaches multiplexing bound with group/pairing	Complexity in pairing/relay scheduling
Reliability	Drastic outage reduction, especially at cell edge	SIC error propagation
Energy efficiency	Enhanced under wireless-powered/harvesting scenarios	Relay energy causality and randomness
Security	Prone to eavesdropping in public relay phases	Physical-layer security/jamming

6. Integration with Emerging Wireless Technologies

Cooperative NOMA is synergistic with several advanced paradigms:

• Energy harvesting and SWIPT: Near-user relays can be powered by base station transmissions (Liu et al., 2015, Salim et al., 22 May 2025). Analytical results confirm that the use of SWIPT does not jeopardize the diversity gain and enables batteryless operation.
• Reconfigurable intelligent surfaces (RIS): RIS-assisted C-NOMA enables channel reconfiguration, further enhancing coverage and link reliability via programmable reflection and passive beamforming (Salim et al., 22 May 2025).
• Space–air–ground integrated networks (SAGIN): Multi-layered relaying, possibly with UAVs or high-altitude platforms, in conjunction with NOMA, extends coverage and supports diverse QoS (Salim et al., 22 May 2025).
• Device-to-device (D2D) overlays: D2D connections facilitate direct low-latency cooperation, improving sum-rate and fairness (Uddin et al., 2018).
• Semantic communication and integrated sensing (ISAC): Multicast or broadcast tasks employing C-NOMA can optimize both channel capacity and application-layer performance (Salim et al., 22 May 2025).
• Machine learning for resource allocation: Reinforcement learning approaches, such as deep Q-networks or actor-critic models, are employed for dynamic control of power, user pairing, beamforming, and trajectory optimization (Salim et al., 22 May 2025).

7. Perspectives and Research Frontiers

Cooperative NOMA remains a rapidly evolving domain with open challenges:

• Robust protocol design under imperfect/incomplete CSI: Stochastic optimization for robust rates under hardware limitations and fast fading remains central (Wan et al., 2018).
• Physical layer security: Integrating secure beamforming, cooperative jamming, and trusted relay selection in a cross-layer secure C-NOMA protocol (Wan et al., 2018, Salim et al., 22 May 2025).
• Near-optimal complexity solutions for large K: Hybrid scheduling, AI-driven pairing, and scalable SIC architectures reduce the complexity bottleneck as user count scales.
• Joint cross-layer optimization: Coordinating PHY-MAC–semantic layers, especially in energy-constrained, latency-sensitive, and service-differentiated environments (Salim et al., 22 May 2025).
• Quantum coexistence and semantic-security: As 6G and beyond evolve, protocols guarding against quantum-level attacks, semantic inference, and RIS reconfiguration security become relevant.
• Standardization and interoperability: The absence of formal C-NOMA specifications in 3GPP NR hinders broad deployment; future standards will need to accommodate integration with RIS/ISAC, privacy assurance, and native AI.

In summary, cooperative NOMA is a theoretically mature and practically attractive solution for next-generation wireless networks, enabling high spectral and energy efficiency, full-diversity operation, and extensive flexibility via advanced relaying, cross-layer integration, and emerging hardware platforms (Ding et al., 2014, Salim et al., 22 May 2025).