Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Game Theoretic Semi-Distributed D2D Resource Allocation Underlaying an LTE Network

Anushree Neogi

TL;DR

The paper tackles D2D resource allocation under LTE with unknown DCSI-R at the BS by introducing two semi-distributed, game-theoretic algorithms that maximize social utility while guaranteeing CU SINR and D2D QoS. It models D2D actions as lists of candidate CUs and uses a frame-based, mood-driven learning mechanism to reach Pareto-optimal allocations without inter-D2D interference. The first algorithm performs poorly under channel randomness and mobility, while the second, a threshold-based variant, remains robust to shadowing, fast fading, and CU mobility, as shown by simulations. The approach reduces control overhead and computation at the BS and is suitable for local D2D transmissions in LTE underlay scenarios, with demonstrated Pareto-optimality and resilience to channel variations.

Abstract

To devise D2D resource allocation algorithms in underlay D2D communications the channel state information (CSI) between the D2D transmitters and the BS and the D2D receiver CSI (DCSI-R) needs to be transmitted to the BS. However, this increases the control overhead and power wastage which increases with a fast fading channel since the CSI needs to be transmitted in every time slot. Most of the existing works assume DCSI-R availability at the BS. However, a few works assume its unavailability and determine the Nash equilibrium which may not be Pareto optimal. We address this problem and within a game theoretic framework propose a suboptimal semi-distributed D2D resource allocation algorithm. We consider the channel to exhibit path loss. Our goal is to maximize the social utility of the D2D users while meeting their utility requirements and the signal-to-interference-plus-noise ratio (SINR) requirements of the CUs to reach a Pareto optimal solution. Next, we consider shadowing, fast fading and mobility of CUs and propose another algorithm which is a modification of our first proposed algorithm. Through simulations we observe that the first algorithm does not perform well practically but the second algorithm is very robust to channel randomness and CU mobility.

Game Theoretic Semi-Distributed D2D Resource Allocation Underlaying an LTE Network

TL;DR

The paper tackles D2D resource allocation under LTE with unknown DCSI-R at the BS by introducing two semi-distributed, game-theoretic algorithms that maximize social utility while guaranteeing CU SINR and D2D QoS. It models D2D actions as lists of candidate CUs and uses a frame-based, mood-driven learning mechanism to reach Pareto-optimal allocations without inter-D2D interference. The first algorithm performs poorly under channel randomness and mobility, while the second, a threshold-based variant, remains robust to shadowing, fast fading, and CU mobility, as shown by simulations. The approach reduces control overhead and computation at the BS and is suitable for local D2D transmissions in LTE underlay scenarios, with demonstrated Pareto-optimality and resilience to channel variations.

Abstract

To devise D2D resource allocation algorithms in underlay D2D communications the channel state information (CSI) between the D2D transmitters and the BS and the D2D receiver CSI (DCSI-R) needs to be transmitted to the BS. However, this increases the control overhead and power wastage which increases with a fast fading channel since the CSI needs to be transmitted in every time slot. Most of the existing works assume DCSI-R availability at the BS. However, a few works assume its unavailability and determine the Nash equilibrium which may not be Pareto optimal. We address this problem and within a game theoretic framework propose a suboptimal semi-distributed D2D resource allocation algorithm. We consider the channel to exhibit path loss. Our goal is to maximize the social utility of the D2D users while meeting their utility requirements and the signal-to-interference-plus-noise ratio (SINR) requirements of the CUs to reach a Pareto optimal solution. Next, we consider shadowing, fast fading and mobility of CUs and propose another algorithm which is a modification of our first proposed algorithm. Through simulations we observe that the first algorithm does not perform well practically but the second algorithm is very robust to channel randomness and CU mobility.

Paper Structure

This paper contains 20 sections, 1 theorem, 5 equations, 5 figures, 3 algorithms.

Table of Contents

  1. Introduction
  2. Background
  3. Motivation
  4. Related Work
  5. Contributions
  6. System Model
  7. Network Model
  8. Game Theoretic Model
  9. Resource Allocation Algorithm
  10. List Selection
  11. Resource Allocation Algorithm at the BS
  12. Utility Calculation
  13. Mood Calculation
  14. Concept Illustration
  15. Shadowing, Fast Fading and Mobility of CUs
  16. ...and 5 more sections

Key Result

Theorem 1

The stochastically stable states of a regular perturbed DTMC are the states $s^* \in \mathcal{S}$, which satisfy the following conditions: (1) The action profile $l(n)$ maximizes the social utility $W_{l}=\sum_{d\in\mathcal{N_D}}r_d(l)$ while satisfying the utility constraints of the D2D players and

Figures (5)

  • Figure 1: D2D players' sequencing and the selection of CUs from the lists of the D2D players in the subframes of a frame.
  • Figure 2: Example demonstrating the optimal mapping of CUs and the D2D pairs.
  • Figure 3: Social utility of the D2D players over frames for the concept illustration example.
  • Figure 4: Social utility of the D2D players over frames with Algorithm 1.
  • Figure 5: Social utility of the D2D players over superframes with Algorithm 3.

Theorems & Definitions (5)

  • Definition 1
  • Definition 2
  • Definition 3
  • Theorem 1
  • proof