RIS-Assisted Sensing: A Nested Tensor Decomposition-Based Approach

Kenneth Benício ∗† , Fazal-E-Asim ∗ , Bruno Sokal ∗ , André L. F. de Almeida ∗ , Behrooz Makki ⋆ , Gabor Fodor ⋆ , and A. Lee Swindlehurst † · 2024

Read RIS-Assisted Sensing: A Nested Tensor Decomposition-Based Approach by Kenneth Benício ∗† , Fazal-E-Asim ∗ , Bruno Sokal ∗ , André L. F. de Almeida ∗ , Behrooz Makki ⋆ , Gabor Fodor ⋆ , and A. Lee Swindlehurst † (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-02778.txt.

Introduction

Kenneth Benício and co-authors sense targets with a smart reflecting surface and tensor math in two thousand twenty-four. A reconfigurable intelligent surface, a wall of tunable reflectors, can aid a radar that both sends and listens. The echo carries the target's delay, speed shift, and angles tangled together.

Their model treats the echo as a tensor, a multi-way array, with nested structure. The sensing task becomes a nested factorization of that array. A two-stage solver first fits by alternating least squares, then reads the target traits by a rotation-based trick. Simulations show sharp estimates at low cost. The method suits sixth-generation networks that merge sensing with communication.

Catalog record

CREATORKenneth Benício ∗† , Fazal-E-Asim ∗ , Bruno Sokal ∗ , André L. F. de Almeida ∗ , Behrooz Makki ⋆ , Gabor Fodor ⋆ , and A. Lee Swindlehurst †
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE40 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2412-02778

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

RIS-Assisted Sensing: A Nested Tensor Decomposition-Based Approach Thanks: This work was partially supported by the Ericsson Research, Sweden, and Ericsson Innovation Center, Brazil, under UFC.51. The authors acknowledge the partial support of Fundação Cearense de Apoio ao Desenvolvimento Científico e Tecnológico (FUNCAP) under grants FC3-00198-00056.01.00/22 and ITR-0214-00041.01.00/23, the National Institute of Science and Technology (INCT-Signals) sponsored by Brazil’s National Council for Scientific and Technological Development (CNPq) under grant 406517/2022-3, and the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - Brazil (CAPES) - Finance Code 001. This work is also partially supported by CNPq under grants 312491/2020-4 and 443272/2023-9. G. Fodor was supported by the EU Horizon Europe Program, Grant No: 101139176 6G-MUSICAL. Kenneth Benício ∗† , Fazal-E-Asim ∗ , Bruno Sokal ∗ , André L. F. de Almeida ∗ , Behrooz Makki ⋆ , Gabor Fodor ⋆ , and A. Lee Swindlehurst † Abstract We study a monostatic multiple-input multiple-output sensing scenario assisted by a reconfigurable intelligent surface using tensor signal modeling. We propose a method that exploits the intrinsic multidimensional structure of the received echo signal, allowing us to recast the target sensing problem as a nested tensor-based decomposition problem to jointly estimate the delay, Doppler, and angular information of the target. We derive a two-stage approach based on the alternating least squares algorithm followed by the estimation of the signal parameters via rotational invariance techniques to extract the target parameters. Simulation results show that the proposed tensor-based algorithm yields accurate estimates of the sensing parameters with low complexity. RIS reconfigurable intelligent surface ALS alternating least squares CE channel estimation RF radio-frequency THz Terahertz communication EVD eigenvalue decomposition CRB Cramér-Rao lower bound CSI channel state information ESPRIT estimation of the signal parameters via rotational invariance techniques BS base station LS least squares SISO single-input single-output MIMO multiple-input multiple-output NMSE normalized mean squared error 2G Second Generation 3G 3 {}^{\text{rd}} Generation 3GPP 3 {}^{\text{rd}} Generation Partnership Project 4G 4 {}^{\text{th}} Generation 5G 5 {}^{\text{th}} Generation 6G 6 {}^{\text{th}} generation E-TALS enhanced TALS ULA uniform linear array UT user terminal UTs users terminal LS least squares KRF Khatri-Rao factorization KF Kronecker factorization MU-MIMO multi-user multiple-input multiple-output MU-MISO multi-user multiple-input single-output MU multi-user NTBS nested tensor decomposition based sensing SER symbol error rate SNR signal-to-noise ratio SVD singular value decomposition OFDM orthogonal frequency division multiplexing EoA elevation angle of arrival AoA azimuth angle of arrival EoD elevation angle of departure AoD azimuth angle of departure AWGN additive white Gaussian noise RCS radar cross section ROT Rank-One-Tensor HOSVD higher-order singular value decomposition LOS line-of-sight NLOS non-line-of-sight GLRT generalized likelihood ratio test CRLB Cramér–Rao lower bound ISAC integrated sensing and communications DFT discrete Fourier transform RMSE root mean squared error I Introduction Reconfigurable intelligent surfaces (RISs) have grown rapidly as a transformative technology and one of the most promising advancements for the next generation of wireless communication systems. RIS s are important for mitigating coverage holes in wireless networks and overcoming blockages, especially for higher frequency bands [ 1 ] . RIS s also have the potential to improve sensing/tracking tasks performed by the BS ( BS ) that do not fall under the conventional communications-only umbrella of the BS . Therefore, RIS s have the potential to be an integral component in supporting ISAC ( ISAC ) technology [ 2 ] . In [ 3 ] , the authors considered an RIS to improve target detection in MIMO ( MIMO ) radar systems by properly adjusting the RIS phase shifts and the radar waveform. The authors in [ 4 ] deal with how to deploy an RIS to improve the spatial diversity of a radar system by creating an additional virtual LOS ( LOS ) path between the radar and the desired target. The authors in [ 5 ] investigate an RIS -assisted NLOS ( NLOS ) sensing scenario and propose both an active and passive beamforming design that minimizes the Cramér–Rao lower bound. The work in [ 6 ] proposes a scenario with a target-mounted RIS that aims to enhance the sensing performance for the desired targets while preventing the detection of the target by eavesdroppers. Finally, [ 7 ] proposes a parameter estimation method for an RIS -assisted monostatic scenario in which a generalized likelihood ratio test is designed. The approach involves a four-dimensional search over the delay-Doppler-azimuth-elevation parameters that uses a repetitive RIS phase-shift profile to decompose the four-dimensional search into separate delay-Doppler and azimuth-elevation search problems. Our previous work in [ 8 ] addresses the problem of parameter estimation for a SISO ( SISO ) sensing system by exploiting the intrinsic multidimensional structure of the received echo signal at the BS . In this paper, we generalize the idea to the MIMO case. To cope with this more challenging scenario, we propose a NTBS ( NTBS ) algorithm for RIS -assisted monostatic sensing [ 9 ] . Our algorithm estimates the desired parameters by exploiting the inherent geometrical structure of the problem and recasting the received echo signal as a fourth-order nested tensor model. The NTBS method is a two-stage algorithm based on ALS ( ALS ), which initially solves the parameter estimation problem iteratively and, in the final step, employs the well-known ESPRIT ( ESPRIT ) algorithm [ 10 , 11 ] to extract the delay, Doppler, and angle estimates. Our simulation results show the impact of the system parameters on estimation performance, as well as computational complexity. Notation : Scalars, vectors, matrices, and tensors are respectively represented as a,\boldsymbol{a},\boldsymbol{A} , and \boldsymbol{\mathcal{A}} . Also, \boldsymbol{A}^{*} , \boldsymbol{A}^{\text{T}} , \boldsymbol{A}^{\text{H}} , and \boldsymbol{A}^{\dagger} respectively stand for the conjugate, transpose, Hermitian transpose, and pseudoinverse of a matrix \boldsymbol{A} . The j th column of \boldsymbol{A}\in\mathbb{C}^{I\times J} is denoted by \boldsymbol{a}_{j}\in\mathbb{C}^{I\times 1} . The operator D (\cdot) converts a vector into a diagonal matrix, \text{D}_{j}(\boldsymbol{B}) forms a diagonal matrix R\times R out of the j th row of \boldsymbol{B}\in\mathbb{C}^{J\times R} . The matrix \boldsymbol{I}_{N} denotes an identity matrix of size N\times N . The n th mode product between a tensor and a matrix is given by \boldsymbol{\mathcal{A}}\times_{n}\boldsymbol{B}=\boldsymbol{B}\left[\boldsymbol{\mathcal{A}}\right]_{(n)} . The symbols \otimes , \diamond , and \odot indicate the Kronecker, Khatri-Rao, and Hadamard products, respectively. Element-wise division is indicated by \oslash . II System Model RIS BS \cdots Target Fig. 1 : RIS-assisted MIMO Monostatic Sensing \boldsymbol{w}_{1} \boldsymbol{w}_{1} \cdots \boldsymbol{w}_{1} \boldsymbol{w}_{2} \boldsymbol{w}_{2} \cdots \boldsymbol{w}_{2} \boldsymbol{w}_{K} \boldsymbol{w}_{K} \cdots \boldsymbol{w}_{K} Block size ( QM ) Total of KQM OFDM symbols 1 st block 2 nd block K th block Fig. 2 : Time-domain protocol of the proposed RIS design. Consider a downlink monostatic RIS scenario, as shown in Figure 1 . The BS has multiple antennas and is assisted by an RIS with N=N_{y}N_{z} passive reflecting elements. The direct LOS between the BS and the target is blocked. The target reflection is only accessible via a virtual LOS path via the RIS . Furthermore, the BS transmits an OFDM ( OFDM ) waveform towards the target and listens to the echo backscattered from the desired target via the RIS . The BS is equipped with an L -element uniformly-spaced antenna array and transmits an OFDM pulse with Q subcarriers and M symbols, structured as shown in Figure 2 . The complex

Read the complete text