A Flexible Design for Beam Squint Effect Suppression in IRS-Aided THz Communications

Technical Summary: A Flexible Design for Beam Squint Effect Suppression in IRS-Aided THz Communications

Problem Statement
Terahertz (THz) communications offer vast bandwidth for next-generation networks but suffer from severe path loss and atmospheric absorption. While Intelligent Reflecting Surfaces (IRS) and Movable Antennas (MAs) have been proposed individually to enhance performance, their joint deployment in wideband THz systems faces a critical challenge: the double beam squint effect. In wideband systems, subcarrier frequencies deviate significantly from the reference (central) frequency. This causes the analog beamforming gains at both the Base Station (BS) and the IRS to degrade and fluctuate across the spectrum. Due to the multiplicative nature of the BS-IRS-user link, these individual degradations compound, leading to severe signal amplitude fluctuations across subcarriers, known as the double beam squint effect. Existing literature has explored MAs for beam squint mitigation in single-hop systems and IRS for THz enhancement, but the specific strategy of deploying movable components on both the BS and the IRS to combat this double effect in wideband THz MISO systems remains largely uninvestigated.

Methodology
The authors propose a system model where a BS equipped with a 2D planar array of MAs serves a user via an IRS composed of movable subarrays. The core objective is to maximize the minimal received power across the entire THz frequency band by jointly optimizing the positions of the MAs at the BS and the subarrays on the IRS surface.

The optimization problem is formulated as a non-convex problem (P1) due to the complex relationship between antenna positions and received signal amplitude, as well as constraints regarding feasible movement regions and minimum separation distances to avoid coupling/collision. To solve this, the authors develop an iterative algorithm based on the Block Coordinate Descent (BCD) method combined with Majorization-Minimization (MM):

1. Problem Transformation: The min-max problem is converted into a tractable form (P2) by introducing a slack variable representing the minimal received power.
2. Alternating Optimization: The algorithm alternates between optimizing the position of a single MA (while fixing others) and a single IRS subarray.
3. Convexification via MM: For the non-convex received power constraints and non-convex distance constraints:
   • The received power function is approximated using a second-order Taylor expansion, replacing the Hessian matrix with a smaller, globally valid lower-bound surrogate.
   • The non-convex distance constraints (minimum separation) are convexified using a first-order Taylor expansion, which serves as a globally concave lower bound.
4. Solution: The resulting subproblems (P4m for MAs and P6k for IRS subarrays) are convex and solved efficiently using standard numerical solvers (e.g., CVX).

Key Contributions

• Novel System Architecture: This work presents the first study on suppressing the double beam squint effect in wideband THz systems by simultaneously deploying an MA array at the BS and movable subarrays on the IRS.
• Joint Position Optimization: The paper formulates and solves a challenging joint optimization problem to configure the physical locations of movable components to maximize the worst-case received power across the spectrum.
• Algorithm Development: A low-complexity algorithm is developed using the MM framework to handle the non-convexities inherent in the position-dependent channel model and separation constraints.

Simulation Results
Numerical results validate the proposed approach under specific THz parameters (e.g., 287–291.6 GHz).

• Convergence: The proposed algorithm demonstrates monotonic convergence, reaching a stable solution in fewer than 10 iterations.
• Beam Squint Suppression: The results show that appropriately configuring the positions of MAs and IRS subarrays can completely eliminate the double beam squint effect. In contrast, systems with fixed antennas or only one side of movable components exhibit significant amplitude fluctuations across the frequency band.
• Performance Gain: While the double beam squint effect is mitigated, the received amplitude still varies slightly across the band due to frequency-dependent path loss (higher frequencies experience higher loss), but the severe degradation caused by beam misalignment is removed.

Significance and Claims
The paper claims that the proposed flexible design offers a significant benefit for wideband THz communications by effectively mitigating the double beam squint effect, which is a major bottleneck for maintaining consistent performance across wide bandwidths. The authors emphasize that this is the first work to investigate this specific dual-movable-component strategy. The study suggests that utilizing movable components on both the BS and IRS is a viable and effective solution to enhance array gain and communication reliability in future THz networks, provided that the positions are delicately configured to counteract frequency-dependent beam deviations. The work maintains a focus on the theoretical and algorithmic contributions to position optimization, supported by numerical simulations, without proposing specific hardware prototypes or broader commercial applications beyond the scope of the communication link performance.