Grant-Free NOMA: A Low-Complexity Power Control through User Clustering.

Area of Science:

• Wireless Communications
• Network Engineering
• Signal Processing

Background:

• Non-orthogonal multiple access (NOMA) enhances spectral efficiency for massive connectivity.
• Traditional NOMA requires grant-based operation, hindering massive machine-type communications.
• Existing NOMA schemes face challenges with channel-state information and power control.

Purpose of the Study:

• Propose synchronous grant-free NOMA (GF-NOMA) frameworks for massive machine-type communications.
• Integrate user equipment (UE) clustering and low-complexity power control within GF-NOMA.
• Facilitate power-reception disparity essential for power-domain NOMA.

Main Methods:

• Develop single-level GF-NOMA (SGF-NOMA) with identical transmit power for all UEs.
• Introduce multi-level GF-NOMA (MGF-NOMA) grouping UEs by signal strength into partitions with distinct power levels.
• Implement dynamic UE clustering based on objectives (max-sum/max-min rate), UE count, and resource blocks (RBs).

Main Results:

• GF-NOMA frameworks compute clusters in milliseconds for hundreds of UEs.
• MGF-NOMA achieves 96-99% of the optimal max-sum rate; SGF-NOMA reaches 87% at similar power.
• SGF-NOMA demonstrates superior energy consumption fairness and network lifetime due to equal UE power usage.

Conclusions:

• GF-NOMA frameworks offer efficient solutions for massive connectivity in IoT.
• MGF-NOMA balances performance with resource utilization, while SGF-NOMA prioritizes energy fairness.
• The proposed GF-NOMA schemes significantly improve upon traditional NOMA limitations for machine-type communications.