D2D Mobile Relaying Meets NOMA -Part I:A Biform Game Analysis

Area of Science:

• Mobile Communications
• Wireless Networking
• Game Theory

Background:

• Device-to-Device (D2D) relaying is crucial for enhancing mobile network performance.
• Decentralized decision-making in D2D networks can lead to performance degradation.
• Self-organizing systems are needed for efficient D2D communication.

Purpose of the Study:

• To analyze the performance of D2D relaying in mobile networks.
• To model device behavior using game theory and predict system performance.
• To implement Reinforcement Learning (RL) for distributed self-organization in D2D networks.

Main Methods:

• Derivation of outage probability for cellular and D2D links.
• Biform game analysis to determine pure and mixed Nash equilibria.
• Implementation of two RL algorithms for distributed strategy learning.

Main Results:

• D2D relaying improves per-device throughput by offloading the network.
• The study provides a framework to analyze and predict system performance.
• RL algorithms enable devices to learn equilibrium strategies in a distributed manner.

Conclusions:

• Decentralized D2D relaying, optimized by RL, enhances mobile network efficiency.
• Game theory provides valuable insights into device behavior and system performance.
• Self-organization through RL is a viable approach for future mobile networks.