Fast Fractional Programming for Multi-Cell Integrated Sensing and Communications

Yannan Chen,Yi Feng,Xiaoyang Li,Licheng Zhao,Kaiming Shen

2024-06-16

Abstract:This paper concerns the coordinate multi-cell beamforming design for integrated sensing and communications (ISAC). In particular, we assume that each base station (BS) has massive antennas. The optimization objective is to maximize a weighted sum of the data rates (for communications) and the Fisher information (for sensing). We first show that the conventional beamforming method for the multiple-input multiple-output (MIMO) transmission, i.e., the weighted minimum mean square error (WMMSE) algorithm, has a natural extension to the ISAC problem scenario from a fractional programming (FP) perspective. However, the extended WMMSE algorithm requires computing the $N\times N$ matrix inverse extensively, where $N$ is proportional to the antenna array size, so the algorithm becomes quite costly when antennas are massively deployed. To address this issue, we develop a nonhomogeneous bound and use it in conjunction with the FP technique to solve the ISAC beamforming problem without the need to invert any large matrices. It is further shown that the resulting new FP algorithm has an intimate connection with gradient projection, based on which we can accelerate the convergence via Nesterov's gradient extrapolation.

Information Theory,Signal Processing

What problem does this paper attempt to address?

This paper attempts to address the problem of optimizing antenna beamforming design in multi-cell Integrated Sensing and Communication (ISAC) systems. Specifically, the paper focuses on how to maximize the weighted sum of data transmission rate (for communication) and Fisher information (for sensing) when each base station (BS) is equipped with a large-scale antenna array. The main problems and solutions of the paper are as follows:

### Problem Description
1. **Background**:
- Integrated Sensing and Communication (ISAC) is an emerging wireless technology that simultaneously performs communication and sensing by reusing network infrastructure and radio signals, thereby reducing infrastructure costs and improving spectrum efficiency.
- This work particularly focuses on ISAC systems with large-scale antenna arrays, aiming to achieve system-level optimization through coordinated antenna beamformers across multiple cells.

2. **Main Challenges**:
- Traditional Multiple Input Multiple Output (MIMO) transmission beamforming methods, such as the Weighted Minimum Mean Square Error (WMMSE) algorithm, are computationally expensive when extended to ISAC scenarios due to the frequent need to compute the inverse of large-scale matrices.
- The deployment of large-scale antenna arrays makes beamforming design very complex, especially in multi-cell environments.

### Solutions
1. **Extending the WMMSE Algorithm**:
- The paper first demonstrates that the WMMSE algorithm can be extended to the ISAC problem scenario through Fractional Programming (FP) techniques, but this method is computationally expensive under large-scale antenna arrays.

2. **Application of Non-Homogeneous Bound**:
- To eliminate the computation of large-scale matrix inverses, the paper introduces a non-homogeneous bound and combines it with FP techniques to solve the ISAC beamforming problem without computing any large-scale matrix inverses.

3. **Gradient Projection and Accelerated Convergence**:
- The study finds that the above method is closely related to gradient projection, thus Nesterov's gradient extrapolation scheme can be used to accelerate convergence.

### Main Results
1. **Extending the WMMSE Algorithm**:
- Through FP techniques, the classical WMMSE algorithm can be extended to ISAC beamforming, but it still requires the computation of large-scale matrix inverses.

2. **Non-Homogeneous FP Method**:
- By introducing a non-homogeneous bound, the computation of large-scale matrix inverses is eliminated, and a non-homogeneous FP method is proposed, significantly reducing computational complexity.

3. **Accelerated Convergence**:
- The non-homogeneous FP method is interpreted as a gradient projection method, and combined with Nesterov's gradient extrapolation scheme, further accelerates the convergence speed of ISAC beamforming optimization.

### Summary
By introducing non-homogeneous bounds and FP techniques, the paper successfully addresses the computational complexity problem of ISAC beamforming under large-scale antenna arrays. Additionally, the use of gradient projection and Nesterov's gradient extrapolation scheme further improves the convergence speed of the algorithm. These methods provide effective solutions for the practical application of ISAC systems.