Hierarchical Functionality Prioritization in Multicast ISAC: Optimal Admission Control and Discrete-Phase Beamforming

Basic Information

• Paper ID: 2501.00201
• Title: Hierarchical Functionality Prioritization in Multicast ISAC: Optimal Admission Control and Discrete-Phase Beamforming
• Authors: Luis F. Abanto-Leon, Setareh Maghsudi
• Classification: eess.SP (Signal Processing), cs.IT (Information Theory), math.IT (Mathematical Information Theory)
• Publication Date: December 31, 2024
• Paper Link: https://arxiv.org/abs/2501.00201

Abstract

This paper investigates the joint design of user admission control and discrete-phase multicast beamforming in integrated sensing and communication (ISAC) systems, where sensing and communication functionalities possess different hierarchical priority levels. The system first allocates resources to high-priority functionalities, then opportunistically utilizes remaining resources to support low-priority functionalities. This resource allocation problem formulates as a non-convex mixed-integer nonlinear programming (MINLP) problem. The authors propose an exact mixed-integer linear programming (MILP) reformulation method to obtain globally optimal solutions. Furthermore, three baseline methods are implemented for comparison, with the proposed method achieving performance improvements exceeding 39% over baselines.

Research Background and Motivation

Core Problems

This paper addresses the hierarchical resource allocation problem for sensing and communication functionalities in ISAC systems, specifically including:

1. Beamforming Design under Discrete-Phase Constraints: Accounting for practical hardware limitations of analog beamformers
2. User Admission Control: Selecting a subset of users to serve under limited resources
3. Functionality Priority Management: Establishing a strict hierarchical structure between communication and sensing functionalities

Research Significance

1. Practical Requirements: Analog beamformers are more practical in high-frequency ISAC systems due to cost and complexity considerations
2. Multicast Applications: Applicable to scenarios serving multiple users simultaneously, such as live sports broadcasts and concerts
3. Resource Scarcity: Severe path loss in millimeter-wave bands makes resource allocation optimization critical

Limitations of Existing Methods

1. Phase Modeling: Existing research primarily focuses on continuous phases, overlooking practical discrete-phase constraints
2. Optimization Approaches: Lack of globally optimal solutions tailored for ISAC systems
3. Functionality Balancing: Traditional weighting methods struggle to maintain preset functionality priorities under parameter variations

Core Contributions

1. First to propose an optimization framework jointly addressing user admission control and discrete-phase multicast beamforming in ISAC systems
2. Establish a strict functionality hierarchy through carefully designed weights ensuring communication functionality always takes priority over sensing
3. Develop an exact MINLP-to-MILP transformation method enabling globally optimal solutions
4. Account for target angle uncertainty improving robustness of sensing performance
5. Provide comprehensive performance evaluation demonstrating significant advantages over baseline methods

Methodology Details

Task Definition

Inputs:

• Base station configuration: N transmit and receive antennas
• User parameters: Channel information of U single-antenna users
• System parameters: Transmit power P_tx, noise power, SNR threshold Γ_th
• Discrete phase set: S = {δe^{jφ₁}, ..., δe^{jφ_L}}

Outputs:

• Multicast beamforming vector w ∈ C^{N×1}
• User admission decisions μ = μ₁, ..., μ_U^T
• Sensing SNR lower bound τ

Constraints:

• Discrete phase constraint: w_n ∈ S, ∀n ∈ N
• User admission constraint: μ_u ∈ {0,1}, ∀u ∈ U
• Communication SNR constraint: w^H H̃_u w ≥ μ_u · Γ_th, ∀u ∈ U
• Sensing SNR constraint: w^H G̃(θ)w ≥ τ, ∀θ ∈ Θ

Model Architecture

1. Communication Model

The signal received by user u is:

y_{com,u} = h_u^H w z + η_{com,u}

where the communication SNR is:

SNR_{com,u}(w) = w^H H̃_u w, H̃_u = (h_u h_u^H)/σ²_{com}

2. Sensing Model

Employing a single-static radar model with target response matrix:

G(θ) = α a(θ) a^H(θ)

The sensing SNR is:

SNR_{sen}(w,θ) = w^H G̃(θ) w, G̃(θ) = G(θ)/σ²_{sen}

3. Hierarchical Weight Design

To ensure communication functionality priority, the following weight design is adopted (Lemma 1):

• ρ_ = 1 (handling the integer part of the objective function)
• ρ_ = σ²_/(2αNP_) (handling the fractional part)

This design ensures that the communication component dominates in f(μ,τ) = ρ_·f_(μ) + ρ_·f_(τ).

Technical Innovations

1. Exact MINLP-to-MILP Transformation

The original problem is transformed through five propositions:

Proposition 1: Discrete phase constraint transformation

[w]_n ∈ S ⇔ {
    [x_n]_l ∈ {0,1}, ∀n∈N, l∈L
    1^T x_n = 1, ∀n∈N  
    [w]_n = s^T x_n, ∀n∈N
}

Propositions 2-5: Stepwise linearization of W = ww^H constraints

• Introduce auxiliary variable W replacing quadratic terms
• Exploit special properties of binary variables
• Convert product constraints into linear constraint combinations

2. Target Angle Uncertainty Modeling

Model target angle θ uncertainty as interval θ-Δ, θ+Δ, discretized into C sampling points:

Θ = {θ̄ | θ̄ = θ - Δ + (2Δc)/(C-1)}, c = 0,...,C-1

Experimental Setup

System Parameters

• Frequency: f_c = 71 GHz (millimeter-wave band)
• Number of Antennas: N = 10
• Number of Users: U = 5
• Transmit Power: P_tx = 36 dBm
• Noise Power: σ²_ = σ²_ = -84 dBm
• Quantization Bits: Q = 3 (8 discrete phases)
• Channel Model: UMa model with Rician fading

Performance Metrics

• Communication Performance: f_(μ) = (1^T μ)/U (number of admitted users)
• Sensing Performance: f_(τ) = τ (minimum sensing SNR)
• Overall Performance: f(μ,τ) = ρ_·f_(μ) + ρ_·f_(τ)

Comparison Methods

1. BL1: Semidefinite relaxation (SDR)-based method using randomization and phase projection
2. BL2: Conservative convex inequality method based on inner approximation
3. BL3: Successive convex approximation (SCA)-based method

Implementation Details

• Solver: CVX + MOSEK
• Hardware: 16GB RAM, Intel Core i7@1.8GHz
• Randomization Iterations: 10⁴ (for baseline methods)

Experimental Results

Main Results

1. Impact of Number of Antennas and Quantization Bits (Scenario I)

• Increasing antenna count N significantly improves sensing performance (enhanced directivity)
• Performance improvement from Q=3 to Q=5 is limited (<6.5%)
• Average runtime for Q=3 is 0.31 seconds; Q=5 requires 6.98 seconds

2. Impact of SNR Threshold and Angle Uncertainty (Scenario II)

• Higher Γ_th requires more power to satisfy communication demands, delaying full user service
• At Γ_th=30, P_tx=28dBm can serve all users
• At Γ_th=60, P_tx=32dBm is required to serve all users
• Increased angle uncertainty Δ disperses sensing power, reducing sensing performance

3. Adaptive Beam Pattern (Scenario III)

• As Γ_th increases, more power is allocated toward user directions
• When Γ_th is too high to satisfy, the system automatically switches to pure sensing mode
• Demonstrates the adaptability and robustness of the method

4. Performance Comparison (Scenario IV)

Within distance range 10,66 meters, average performance improvements of the proposed method over baselines:

• vs. BL1: 59%
• vs. BL2: 39%
• vs. BL3: 47%

Runtime comparison:

• OPT: 0.39 seconds
• BL1: 0.30 seconds
• BL2: 0.19 seconds
• BL3: 0.24 seconds

Experimental Findings

1. Effectiveness of Hierarchical Priority: Communication functionality is always prioritized
2. Advantages of Global Optimality: Significantly outperforms heuristic methods
3. Practical Balance: Q=3 provides good trade-off between performance and complexity
4. Adaptive Capability: System automatically switches between different modes based on conditions

Related Work

ISAC Beamforming Research

• Continuous Phase Methods: Works 3-5 focus on theoretically optimal but complex implementations
• Discrete Phase Methods: Few works like 6, but lacking globally optimal solutions
• Multicast ISAC: Works like 9 do not consider discrete-phase constraints

Admission Control Research

• Non-ISAC Systems: Works 8,10 apply to traditional communication systems
• ISAC Systems: Research gap in this field; this paper is the first to introduce it

Optimization Methods

• SDR Methods: Widely applied but solutions may not be rank-one
• SCA Methods: Iterative optimization but may converge to local optima
• MILP Methods: First application to ISAC discrete-phase optimization in this paper

Conclusions and Discussion

Main Conclusions

1. First to achieve globally optimal design of discrete-phase multicast beamforming in ISAC systems
2. Successfully establish a hierarchical resource allocation framework prioritizing communication
3. Effectively handle target angle uncertainty's impact on sensing performance
4. Significantly outperform existing heuristic methods with performance improvements exceeding 39%

Limitations

1. Computational Complexity: While superior to exhaustive search, still higher than heuristic methods
2. Single Target Assumption: Currently considers only single sensing target
3. Perfect CSI Assumption: Does not account for channel estimation errors
4. Static Scenario: Does not address online optimization in dynamic environments

Future Directions

1. Multi-Target Sensing: Extend to multi-target ISAC scenarios
2. Imperfect CSI: Robust design considering channel estimation errors
3. Dynamic Optimization: Develop online algorithms for time-varying environments
4. Hardware Implementation: Validate algorithm performance on practical ISAC systems

In-Depth Evaluation

Strengths

1. Problem Importance: Addresses practical engineering problems in ISAC systems
2. Methodological Innovation: MINLP-to-MILP transformation method has theoretical value
3. Comprehensive Experiments: Multi-scenario simulations validate method effectiveness
4. Writing Clarity: Rigorous mathematical derivations with clear exposition

Weaknesses

1. Insufficient Theoretical Analysis: Lacks convergence and complexity analysis
2. Parameter Sensitivity: Insufficient analysis of key parameters' sensitivity to performance
3. Missing Practical Validation: Only simulation results; lacks real system verification
4. Scalability Concerns: Performance in large-scale systems remains unknown

Impact

1. Academic Contribution: Provides new theoretical framework for ISAC system optimization
2. Practical Value: Offers guidance for 6G millimeter-wave ISAC system design
3. Reproducibility: Detailed method description facilitates reproduction and extension

Applicable Scenarios

1. Millimeter-Wave ISAC Systems: Particularly suitable for high-frequency applications
2. Multicast Service Scenarios: Such as UAV swarm control and vehicular network broadcasting
3. Resource-Constrained Environments: Scenarios with limited power and hardware costs
4. Sensing-Aided Communication: Applications where communication is primary and sensing is auxiliary

References

The paper cites 18 relevant references covering key works in ISAC system design, beamforming optimization, and multicast communication, providing a solid theoretical foundation for the research.