Introduction
Integrated Sensing and Communications (ISAC) is a cornerstone of 6G research, promising to unify data transmission and environmental sensing within the same hardware and spectral footprint. However, transitioning ISAC from theory to practice involves overcoming significant hardware and operational hurdles. Most theoretical models assume "continuous phase" beamformers, which are prohibitively expensive for mass deployment. In reality, cost-effective analog beamformers utilize discrete phases, introducing non-convexity and discrete constraints that complicate optimization. Furthermore, when resources are limited, systems must decide how to prioritize communication quality against sensing accuracy.
This paper introduces a novel framework that addresses these challenges through a hierarchical prioritization strategy. By reformulating a complex, non-convex optimization problem into a Mixed-Integer Linear Program (MILP), the authors provide a globally optimal solution for joint user admission and beamforming, ensuring that communication performance is strictly prioritized over sensing.
The Challenge: Hardware Constraints and Functional Conflict
The research identifies two primary obstacles in current ISAC development:
- Hardware Realism: Standard optimization techniques often assume high-precision phase shifters. Practical systems use discrete-phase hardware, which makes the underlying mathematical problem a Non-Convex Mixed-Integer Nonlinear Program (MINLP)—a class of problems that is typically NP-hard.
- Resource Prioritization: Traditional ISAC research often uses weighted sums to balance communication and sensing. However, these weights can be unpredictable; a change in transmit power might inadvertently degrade a critical communication link to slightly improve radar resolution.
The authors propose a "comms-first" hierarchy, where the system first maximizes the number of admitted users and their Quality of Service (QoS) before allocating any remaining degrees of freedom to optimize sensing Signal-to-Noise Ratio (SNR).
Methodology: From MINLP to MILP
The core technical contribution is the transformation of the "nasty" MINLP into a solvable MILP. This process involves several key steps:
- Linearization: Through a series of mathematical reformulations (Propositions 1–5), the authors replace the products of binary variables with new matrix variables. They apply constraints that enforce the "rank-one" and "constant-modulus" properties essential for valid beamforming.
- Hierarchical Weighting Strategy: The authors derive a specific set of weights (detailed in Lemma 1) that mathematically guarantees the desired hierarchy. This ensures the optimizer treats the number of admitted users as the primary objective and sensing SNR as the secondary objective, regardless of external parameter fluctuations.
- Robust Admission Control: Unlike models that assume all users must be served, this framework includes an admission control mechanism. This allows the network to gracefully degrade by dropping lower-priority users when the environment becomes too crowded or the signal path too weak, rather than failing entirely.
Performance and Key Insights
The proposed optimal approach (OPT) was evaluated using 3GPP UMa (Urban Micro) channel models and Rician fading at 71 GHz. The results demonstrate significant advantages over common heuristic baselines, such as Semidefinite Relaxation (SDR) and Successive Convex Approximation (SCA):
- Global Optimality: While heuristics are computationally faster, they frequently converge on local optima. By using a branch-and-cut solver on the MILP formulation, the authors achieve the global best solution, resulting in a performance gain of 39% to 59% over baselines in terms of the communication operating region.
- Dynamic Adaptation: Simulations show that the system intelligently adapts its beampattern. When communication requirements are low, the beam focuses on the sensing target. As requirements increase, the system dynamically widens or shifts the beam to maintain user connections, only sacrificing sensing quality when absolutely necessary to preserve the communication hierarchy.
- Robustness: The model accounts for angular uncertainty in target location, ensuring that sensing remains effective even when the target's exact position is not perfectly known.
Conclusion and Future Directions
This research provides a rigorous mathematical foundation for deploying ISAC on realistic, low-cost hardware. By guaranteeing a strict hierarchy of service, it aligns ISAC operations with the reliability requirements of industrial and commercial telecommunications.
While the MILP approach ensures optimality, its computational complexity remains a challenge for real-time implementation in massive MIMO arrays with hundreds of antennas. Future work will likely focus on scaling these optimal solutions or developing "optimality-aware" heuristics that can approach these performance gains with lower latency. Additionally, expanding the model to include multi-group multicasting and dynamic, mobile environments will be essential for full 6G integration.
Summary updated