Scintillation Minimization versus Intensity Maximization in Optimal Beams

Basic Information

• Paper ID: 2211.09265
• Title: Scintillation Minimization versus Intensity Maximization in Optimal Beams
• Authors: Anjali Nair, Qin Li, Samuel N Stechmann (University of Wisconsin-Madison)
• Classification: physics.optics, math.OC, math.SG, physics.comp-ph
• Publication Date: April 28, 2023 (arXiv v2)
• Paper Link: https://arxiv.org/abs/2211.09265

Abstract

In free-space optical communication and other applications, designing optical beams with reduced or minimized scintillation is critically important. However, the optimization problem of minimizing scintillation is challenging, and optimal solutions are rarely found. This paper investigates the general optimization problem of scintillation minimization and formulates it as a convex optimization problem. An analytical solution is obtained, proving that the beam minimizing scintillation is incoherent light (i.e., spatially uncorrelated). Furthermore, numerical solutions indicate that beams minimizing scintillation yield very low intensity at the receiver. To counteract this effect, a new convex cost function balancing scintillation and intensity is studied. Numerical experiments demonstrate that minimizers of this cost function reduce scintillation while maintaining significantly higher intensity levels at the receiver.

Research Background and Motivation

Problem Definition

When optical beams propagate through the atmosphere, they experience unwanted distortions, including intensity reduction, beam drift, or scintillation at the receiver. While fully coherent beams can maximize the total average intensity at the receiver, partially coherent beams are known to exhibit reduced scintillation effects.

Research Significance

1. Practical Application Needs: Free-space optical communication systems require stable signal transmission
2. Technical Challenges: Atmospheric turbulence-induced scintillation severely affects communication quality
3. Optimization Difficulties: The mathematical optimization problem of scintillation minimization is extremely complex

Limitations of Existing Methods

1. Previous work was primarily limited to special classes of beams (e.g., Gaussian-Schell model beams)
2. Lack of general computational methods for the mutual intensity function J
3. Convexity of the optimization problem is unclear, making it difficult to guarantee global optimality

Research Motivation

This paper aims to overcome these difficulties by reformulating the scintillation minimization problem as a convex optimization problem to find the optimal mutual intensity function J for general beams.

Core Contributions

1. Theoretical Contribution: First proof that the scintillation minimization problem can be formulated as a convex optimization problem
2. Analytical Solution Discovery: Proof that the optimal beam minimizing scintillation is completely incoherent light (white noise)
3. Numerical Method Innovation: Development of an efficient numerical algorithm based on randomized SVD solvers
4. Balanced Optimization Framework: Proposal of a novel convex cost function simultaneously considering scintillation and intensity
5. Practical Analysis: Revelation of the contradiction between pure scintillation minimization and practical application requirements

Methodology Details

Task Definition

Input: Light source φ(X) in transmission area A Output: Optimal mutual intensity function J(X₁,X₂) Objective: Minimize scintillation index S = VarI/EI² Constraints: J must be a positive semi-definite Hermitian function

Mathematical Framework

Basic Physical Model

The relationship between received field U and light source φ is:

U(X') = ∫[X∈A] h(X,X')φ(X)dX

where h(X,X') is the propagation function.

Scintillation Index Definition

S = Var[I]/E[I]² = E[I²]/E[I]² - 1

Convex Optimization Reformulation

The key innovation is reformulating the non-convex scintillation minimization problem as:

min J*AJ  s.t.  Tr(E[H]J^T) = 1, J ≽ 0

where A = EH(X₁,X₂)H*(X₃,X₄) is a four-dimensional tensor.

Technical Innovations

1. Convexity Proof

By analogy with the Rayleigh quotient in linear algebra, the reformulated problem is proven to be convex, thus guaranteeing the existence of a global optimal solution.

2. Randomized SVD Solver

To address computational and storage challenges for high-dimensional tensor A:

• Reduces cubic complexity to quadratic
• Avoids pre-construction of the complete tensor A
• Only requires computing the action of A on vectors

3. Analytical Solution Derivation

For the single-phase screen model, the propagation function is:

h(X,X') = h₀(X,X')e^(iψ(X))

It is proven that when the mutual intensity function is a Dirac delta function, scintillation S = 0.

Balanced Optimization Method

A new objective function is proposed:

min S(J) + λQ(J)

where Q(J) = ||I₀/EI - 1||² measures intensity efficiency, and λ is a balance parameter.

Experimental Setup

Numerical Simulation Parameters

1. Parabolic Wave Equation (PWE) Model:
   • Frequency: k = 2π × 10⁶ rad/m
   • Propagation distance: L = 3000m
   • Random potential: V(x,z) = V₁sin(ω_x x)sin(ω_z z)
2. Multi-Phase Screen Model:
   • 2D problem: k = 2π × 10⁶ rad/m
   • Propagation distance: L = 2000m
   • Fried parameter: r₀ = 0.04m
   • 15 phase screens, point receiver

Evaluation Metrics

• Scintillation Index S: Measures beam stability
• Intensity Quotient Q: Measures transmission efficiency
• Shape of Mutual Intensity Function J: Reflects beam coherence

Implementation Details

• CVX toolkit used for solving convex optimization problems
• Splitting method employed for PWE simulation
• Monte Carlo sampling for computing expectations

Experimental Results

Main Findings

1. Optimal Solution for Scintillation Minimization

Numerical Results: Figure 1a shows optimal J exhibits delta function shape Parametric Verification: Figure 1b shows that for Gaussian parameterization J = exp(-λ²(X₁-X₂)²/2), λ → ∞

2. Analytical Verification

For approximate Dirac delta function:

J(X₁,X₂) = I_ε(X₁-X₂)/(H₀(0)ε|A|)

Calculation yields: EI = 1, EI² = 1, therefore S = 0

3. Intensity-Scintillation Trade-off

Key Finding: The received intensity of optimal scintillation-minimizing beams is EI/I₀ = O(ε), practically useless

4. Balanced Optimization Results

Figure 2 shows the trade-off for different λ values:

• λ = 0: Pure scintillation minimization, λ → ∞
• λ > 0: Finite optimal λ value
• Larger λ: Tends toward coherent beams

Numerical Experiment Verification

Figure 3 shows optimal J in the non-parametric case:

• Small λ values: J relatively diagonalized (partially coherent)
• Large λ values: J essentially coherent

Figure 4 quantifies the relationship between S and Q with respect to λ:

• Phase screen model: Good balance achieved at λ ≈ 0.25
• PDE model: Good balance achieved at λ ≈ 2

Experimental Findings

1. Theory-Numerics Consistency: Multiple models confirm that completely incoherent light is optimal for scintillation minimization
2. Practical Contradiction: Pure scintillation minimization conflicts with intensity maximization objectives
3. Balance Strategy Effectiveness: Appropriate λ values can reduce scintillation while maintaining sufficient intensity

Related Work

Main Research Directions

1. Partially Coherent Beam Research: Gbur & Wolf (2002), Dogariu & Amarande (2003) and others studied the impact of partial coherence on scintillation
2. Special Beam Optimization: Voelz & Xiao (2009), Borah & Voelz (2010) and others optimized specific beam classes
3. Atmospheric Propagation Theory: Andrews & Phillips (2005) and others established foundational theoretical frameworks

Advantages of This Work

1. Generality: Not limited to special beam classes
2. Theoretical Rigor: Provides theoretical foundation based on convex optimization
3. Computational Feasibility: Develops efficient numerical algorithms
4. Practical Considerations: Proposes methods balancing scintillation and intensity

Conclusions and Discussion

Main Conclusions

1. The Optimal Beam for Scintillation Minimization is Completely Incoherent Light: This conclusion is verified both analytically and numerically
2. Pure Scintillation Minimization is Impractical: It results in extremely low received intensity
3. Balanced Optimization is Necessary: A suitable balance point between scintillation and intensity must be found
4. Effectiveness of the Convex Optimization Framework: Provides reliable solution methods for this class of problems

Limitations

1. Computational Complexity: Despite randomized SVD, large-scale 3D problems remain challenging
2. Model Assumptions: Based on specific turbulence models; real atmospheric conditions may be more complex
3. Parameter Selection: The choice of balance parameter λ requires adjustment based on specific application scenarios

Future Directions

1. More Complex Turbulence Models: Consider more realistic atmospheric turbulence conditions
2. Adaptive Optimization: Dynamically adjust beam characteristics based on real-time environmental conditions
3. Hardware Implementation: Translate theoretical results into practical beam-shaping systems

In-Depth Evaluation

Strengths

1. Significant Theoretical Contribution: First rigorous formulation of scintillation minimization as a convex optimization problem
2. Strong Method Innovation: Randomized SVD solver effectively addresses computational challenges
3. Insightful Results: The optimality of completely incoherent light is surprising and has profound physical significance
4. High Practical Value: Balanced optimization framework provides feasible solutions for practical applications
5. Comprehensive Experimental Verification: Multiple models and methods validate theoretical results

Weaknesses

1. Computational Scale Limitations: Computational cost remains high for large-scale 3D problems
2. Model Simplification: Single and multi-phase screen models are simplified compared to real atmosphere
3. Parameter Sensitivity: Lack of systematic guidance for selecting balance parameter λ
4. Absence of Hardware Implementation: Lacks experimental verification on actual optical systems

Impact

1. Academic Impact: Provides new theoretical framework for optical propagation and optimization theory
2. Application Prospects: Significant implications for free-space optical communication, LIDAR, and other fields
3. Methodological Contribution: Convex optimization methods can be generalized to other optical optimization problems

Applicable Scenarios

1. Free-space optical communication system design
2. Laser atmospheric propagation optimization
3. Atmospheric disturbance compensation in astronomical observation
4. Beam quality control in industrial laser applications

References

1. T. J. Schulz, "Optimal beams for propagation through random media," Opt. Lett. 30, 1093 (2005)
2. L. C. Andrews and R. L. Phillips, "Laser Beam Propagation Through Random Media," Second Edition (2005)
3. G. Gbur and E. Wolf, "Spreading of partially coherent beams in random media," J. Opt. Soc. Am. A 19, 1592 (2002)

Overall Assessment: This is a high-quality research paper combining theory and computation, making important contributions to optical propagation optimization. The paper not only resolves a long-standing theoretical problem but also provides practical solutions, possessing significant academic value and application prospects.