Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Optimal Sensing Policy With Interference-Model Uncertainty

Vincent Corlay, Jean-Christophe Sibel, Nicolas Gresset

TL;DR

This letter considers a half-duplex scenario where an interferer behaves according to a parametric model but the values of the model parameters are unknown, and shows that both the optimal open-loop and optimal closed-loop policies can be determined with reduced computational complexity compared to the standard backward-induction algorithm.

Abstract

This paper considers a half-duplex scenario where an interferer behaves according to a parametric model but the values of the model parameters are unknown. We explore the necessary number of sensing steps to gather sufficient knowledge about the interferer's behavior. With more sensing steps, the reliability of the model-parameter estimates is improved, thereby enabling more effective link adaptation. However, in each time slot, the communication system experiencing interference must choose between sensing and communication. Thus, we propose to investigate the optimal policy for maximizing the expected sum communication data rate over a finite-time communication. This approach contrasts with most studies on interference management in the literature, which assume that the parameters of the interference model are perfectly known. We begin by showing that the problem under consideration can be modeled within the framework of a Markov decision process (MDP). Following this, we demonstrate that both the optimal open-loop and optimal closed-loop policies can be determined with reduced computational complexity compared to the standard backward-induction algorithm.

Optimal Sensing Policy With Interference-Model Uncertainty

TL;DR

This letter considers a half-duplex scenario where an interferer behaves according to a parametric model but the values of the model parameters are unknown, and shows that both the optimal open-loop and optimal closed-loop policies can be determined with reduced computational complexity compared to the standard backward-induction algorithm.

Abstract

This paper considers a half-duplex scenario where an interferer behaves according to a parametric model but the values of the model parameters are unknown. We explore the necessary number of sensing steps to gather sufficient knowledge about the interferer's behavior. With more sensing steps, the reliability of the model-parameter estimates is improved, thereby enabling more effective link adaptation. However, in each time slot, the communication system experiencing interference must choose between sensing and communication. Thus, we propose to investigate the optimal policy for maximizing the expected sum communication data rate over a finite-time communication. This approach contrasts with most studies on interference management in the literature, which assume that the parameters of the interference model are perfectly known. We begin by showing that the problem under consideration can be modeled within the framework of a Markov decision process (MDP). Following this, we demonstrate that both the optimal open-loop and optimal closed-loop policies can be determined with reduced computational complexity compared to the standard backward-induction algorithm.
Paper Structure (17 sections, 4 theorems, 16 equations, 2 figures, 1 table, 2 algorithms)

This paper contains 17 sections, 4 theorems, 16 equations, 2 figures, 1 table, 2 algorithms.

Table of Contents

  1. Introduction
  2. Interference model, channel model, and link-adaptation strategies
  3. Interference and sensing models
  4. Interference characterization
  5. Considered channel
  6. Link adaptation
  7. Optimal sensing strategy
  8. Modeling as a MDP
  9. Review of MDP and notations
  10. Similarities with bandit problems
  11. Efficiently finding the optimal open and closed-loop policies
  12. Expected communication data rate after $i$ sensing steps (open loop)
  13. The on/off property of the problem
  14. The optimal open-loop policy
  15. The optimal closed-loop policy
  16. ...and 2 more sections

Key Result

Lemma 1

For any $j\geq 0$, $g(k,n,j+1) \geq g(k,n,j)$.

Figures (2)

  • Figure 1: Value of the state $s(k=0,n=0,i=0)$ under several policies and with $T=1000$. For the closed-loop policy the x-axis (number of sensing steps) represents the maximum number of allowed sensing steps. O-L refers to the open-loop policy while C-L refers to the closed-loop policy.
  • Figure 2: Example of evolution of a system over time with $T=1000$. Only the first 60 slots are displayed.

Theorems & Definitions (7)

  • Lemma 1
  • proof
  • Lemma 2
  • proof
  • Lemma 3
  • Corollary 1
  • proof