pith. sign in

arxiv: 1906.08970 · v1 · pith:H2OBUIQ7new · submitted 2019-06-21 · 📡 eess.SP

Analog Beamforming for Active Imaging using Sparse Arrays

Robin Rajam\"aki , Sundeep Prabhakar Chepuri , Visa Koivunen This is my paper

Pith reviewed 2026-05-25 19:08 UTC · model grok-4.3

classification 📡 eess.SP
keywords analog beamformingactive imagingsparse arraysimage additionpoint spread functiongradient descentmillimeter-wave radarultrasound imaging
0
0 comments X

The pith

Adding multiple analog beamformed component images can synthesize the point spread function of a fully-digital beamformer while enabling sparse arrays.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

Fully-digital beamformers need a separate radio-frequency chain for every array element, which raises costs sharply in active imaging such as millimeter-wave radar or ultrasound. The paper demonstrates that analog beamformers, which use only inexpensive phase shifters attached to a single chain, can reach equivalent image quality by acquiring and summing several lower-resolution component images formed with different phase settings. This image-addition approach also permits the use of sparse arrays with fewer sensors. The authors pose an optimization problem whose solution, found by gradient descent, minimizes the number of component images needed to meet a target point spread function and prove an upper bound guaranteeing that the fully-digital performance is always reachable with a finite number of images.

Core claim

Image addition synthesizes a high-resolution image by adding together several lower-resolution component images obtained via analog beamforming with phase shifters. This enables the use of sparse arrays. A gradient descent algorithm finds a locally optimal solution that minimizes the number of component images subject to achieving a desired point spread function, and an upper bound is derived on the number needed to achieve the traditional fully-digital beamformer solution.

What carries the argument

Image addition, the mechanism of summing multiple phase-shifter-controlled component images to approximate the point spread function of a fully-digital beamformer.

If this is right

  • Sparse arrays become usable, lowering the total number of sensors required.
  • Hardware cost drops because only one RF-IF chain is needed instead of one per element.
  • Image acquisition time stays practical because the number of component images is minimized by the optimization.
  • The achieved point spread function matches that of the traditional fully-digital beamformer.
  • A finite upper bound always exists on the number of component images required to reach the digital solution.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The approach could be tested directly on existing analog hardware to quantify real-world side-lobe levels under motion or clutter.
  • Dynamic scenes would require the component images to be acquired fast enough that the target does not move appreciably between them.
  • The gradient-descent procedure might be initialized with closed-form phase patterns from classical array theory to improve convergence speed.
  • The same addition principle could be applied to passive sensing if suitable reference signals are available.

Load-bearing premise

Summing a finite number of analog beamformed component images controlled only by phase shifters can achieve a point spread function comparable to a fully-digital beamformer without being limited by noise, mutual coupling, or hardware imperfections.

What would settle it

Measure the point spread function obtained from the summed analog component images in a physical array experiment and compare it to the theoretical fully-digital point spread function; a significant mismatch would falsify the claim.

read the original abstract

This paper studies analog beamforming in active sensing applications, such as millimeter-wave radar or ultrasound imaging. Analog beamforming architectures employ a single RF-IF chain connected to all array elements via inexpensive phase shifters. This can drastically lower costs compared to fully-digital beamformers having a dedicated RF-IF chain for each sensor. However, controlling only the element phases may lead to elevated side-lobe levels and degraded image quality. We address this issue by image addition, which synthesizes a high resolution image by adding together several lower resolution component images. Image addition also facilitates the use of sparse arrays, which can further reduce array costs. To limit the image acquisition time, we formulate an optimization problem for minimizing the number of component images, subject to achieving a desired point spread function. We propose a gradient descent algorithm for finding a locally optimal solution to this problem. We also derive an upper bound on the number of component images needed for achieving the traditional fully-digital beamformer solution.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 0 minor

Summary. The paper proposes image addition—synthesizing a target point-spread function (PSF) as a linear combination of multiple analog-beamformed component images acquired via phase-shifter-controlled sparse arrays—as a means to achieve high-resolution active imaging (e.g., mm-wave radar, ultrasound) at lower hardware cost than fully-digital beamforming. It formulates an optimization problem that minimizes the number of component images subject to a desired PSF, supplies a gradient-descent procedure for a locally optimal solution, and derives an explicit upper bound guaranteeing that the fully-digital beamformer solution can be recovered.

Significance. If the optimization, algorithm, and bound are shown to work in practice, the work would offer a concrete route to cost reduction in active sensing arrays while preserving PSF quality and providing a theoretical guarantee of equivalence to the digital case; the bound derivation and the explicit formulation of the minimization problem are strengths that could be cited by subsequent hardware-oriented papers.

major comments (1)
  1. [Abstract and optimization sections] The abstract and method description present the gradient-descent solver and the upper bound on the number of component images, yet the manuscript supplies no numerical simulations, Monte-Carlo error analysis, or direct comparison against a fully-digital beamformer to confirm that the obtained solution actually meets the target PSF. This evidentiary gap is load-bearing for the central claim that the proposed procedure is practically useful.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the detailed review and the recommendation for major revision. We address the single major comment below.

read point-by-point responses

  1. Referee: [Abstract and optimization sections] The abstract and method description present the gradient-descent solver and the upper bound on the number of component images, yet the manuscript supplies no numerical simulations, Monte-Carlo error analysis, or direct comparison against a fully-digital beamformer to confirm that the obtained solution actually meets the target PSF. This evidentiary gap is load-bearing for the central claim that the proposed procedure is practically useful.

    Authors: We agree that the absence of numerical validation constitutes a significant evidentiary gap for claims of practical utility. The present manuscript is primarily theoretical, concentrating on the optimization formulation, the gradient-descent procedure, and the derivation of the upper bound that guarantees recovery of the fully-digital solution. In the revised manuscript we will add a dedicated numerical-results section containing (i) direct comparisons of the synthesized PSF against the target fully-digital PSF for representative array geometries, (ii) Monte-Carlo trials that quantify the deviation under phase-shifter quantization and additive noise, and (iii) timing and hardware-cost metrics that illustrate the savings relative to fully-digital beamforming. revision: yes

Circularity Check

0 steps flagged

No significant circularity

full rationale

The paper's central construction—formulating image addition as a linear combination of analog beamformed responses to synthesize a target PSF, posing an optimization problem to minimize the number of such responses, proposing a gradient-descent solver, and deriving an explicit upper bound guaranteeing equivalence to the fully-digital beamformer—is presented as a self-contained mathematical framework. No step reduces by definition to a fitted parameter, renames a prior result, or relies on a load-bearing self-citation whose content is itself unverified within the paper. The optimization objective and bound are stated independently of any author-specific prior equations or fitted inputs, making the derivation chain non-circular.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The approach rests on standard assumptions about the ability of phase-only control plus linear image addition to synthesize arbitrary PSFs and on the applicability of gradient descent to the resulting non-convex problem; no new physical entities or fitted constants are introduced.

axioms (2)
  • domain assumption Gradient descent finds a locally optimal solution to the formulated minimization problem.
    Invoked when proposing the algorithm without convergence guarantees or global optimality claims.
  • domain assumption A finite linear combination of analog beam patterns can approximate any desired point spread function.
    Underlying premise of the image-addition method and the optimization objective.

pith-pipeline@v0.9.0 · 5703 in / 1317 out tokens · 36782 ms · 2026-05-25T19:08:35.923094+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

29 extracted references · 29 canonical work pages

  1. [1]

    For ex- ample, 3D ultrasound imaging typically uses hundreds of sen- sors, each with a dedicated transceiver chain

    INTRODUCTION The use of high frequencies enables small array form factors by packing many elements into a small physical area. For ex- ample, 3D ultrasound imaging typically uses hundreds of sen- sors, each with a dedicated transceiver chain. Although the resulting large electrical aperture improves the array’s r esolu- tion, the hardware cost, number of ...

    work page

  2. [2]

    As shown in Fig

    SIGNAL MODEL AND DEFINITIONS Consider a sensor array with Nt transmit (Tx) and Nr receive (Rx) elements. As shown in Fig. 1, each array element is con- 1We address the more general case of hybrid beamforming with quantized phase shifts in the longer journal version of this paper [14] . Nt H Scattering scene Analog beamforming Digital processing DAC Tx cha...

    work page

  3. [3]

    BOUNDS ON NO. OF COMPONENT IMAGES Q Next, we derive an upper and lower bound on the number of component images Q required by an analog beamformer for factorizing any co-array matrix W∈ CNr×Nt as in (4). In the case of fully-digital beamforming, SVD can be used to decompose W as in (3) using Qd = rank(W)≤ min(Nr, Nt) component images [4]. Any analog factor...

    work page

  4. [4]

    Here ∡ , cos−1, and|·|are applied elementwise

    mod 2 . Here ∡ , cos−1, and|·|are applied elementwise. Proof sketch (see [14] for details). By [6, Theorem 1], any wx = ∑2 m=1 cx,mfx,m. Consequently, wrwT t = ∑4 m=1 cr,ict,l fr,if T t,l, where i and l are functions of the summation index m. As W is a sum of Qd rank-1 matrices, we have Q = 4Qd. The solution given by Theorem 1 is actually non-unique, and ...

    work page

  5. [5]

    Assuming that the PSF is evalu- ated for a set of V discrete target directions {vi}V i=1, we may express the desired PSF as ψ ∈ CV and the realized PSF as Avec(W)

    PROBLEM FORMULA TION The goal of the optimization problem formulated in this pa- per is to minimize the number of component images Q, while achieving a desired PSF. Assuming that the PSF is evalu- ated for a set of V discrete target directions {vi}V i=1, we may express the desired PSF as ψ ∈ CV and the realized PSF as Avec(W). The ith row of measurement m...

    work page

  6. [6]

    Note that in practice, the maximum value of Q is determined by Theorem 1, or by a design constraint on the minimum imaging frame rate

    (P2) If we can solve (P2), we can easily recover the solution to (P1) by finding the smallest Q for which the objective of (P2) does not exceed ε2 max. Note that in practice, the maximum value of Q is determined by Theorem 1, or by a design constraint on the minimum imaging frame rate

    work page

  7. [7]

    We start by noting that the optimal value of c in (P2) is the least-squares solution c = ( A(Ft⋄ Fr))†ψ, where† denotes the pseudo-inverse

    GRADIENT DESCENT ALGORITHM In this section, we present a simple gradient descent method for solving (P2). We start by noting that the optimal value of c in (P2) is the least-squares solution c = ( A(Ft⋄ Fr))†ψ, where† denotes the pseudo-inverse. We also write the analog weight matrix Fx directly as a function of the unknown phase matrix Φx∈ RNx×Q, i.e., F...

    work page

  8. [8]

    NUMERICAL EXAMPLES Next, we compare the PSFs of two analog beamformers with linear array geometries (Fig. 2). In particular, we conside r a uniform linear array (ULA) with N = 11 elements, and a minimum-redundancy array (MRA) [16, 17] with N = 7 ele- ments [18]. Both arrays span an aperture of 10λ/2, where λ/2 is the smallest inter-element spacing. Assumi...

    work page

  9. [9]

    We proposed a gradient descent al- gorithm for jointly finding the transmit and receive element weights that achieve a desired PSF using as few transmissions as possible

    CONCLUSIONS This paper considered active imaging using phased arrays with an analog beamforming architecture consisting of con- tinuous phase shifters. We proposed a gradient descent al- gorithm for jointly finding the transmit and receive element weights that achieve a desired PSF using as few transmissions as possible. Moreover, we derived a bound on the...

    work page

  10. [10]

    The unifying role of the coarray in aperture synthesis for coherent and incoher- ent imaging,

    R. T. Hoctor and S. A. Kassam, “The unifying role of the coarray in aperture synthesis for coherent and incoher- ent imaging,” Proceedings of the IEEE , vol. 78, no. 4, pp. 735–752, Apr 1990

    work page 1990

  11. [11]

    Nested arrays: A novel approach to array processing with enhanced degrees of freedom,

    P . Pal and P . P . V aidyanathan, “Nested arrays: A novel approach to array processing with enhanced degrees of freedom,” IEEE Transactions on Signal Processing , vol. 58, no. 8, pp. 4167–4181, Aug 2010

    work page 2010

  12. [12]

    Coarrays, MUSIC, and the Cram´ er-Rao bound,

    M. Wang and A. Nehorai, “Coarrays, MUSIC, and the Cram´ er-Rao bound,”IEEE Transactions on Signal Pro- cessing, vol. 65, no. 4, pp. 933–946, Feb 2017

    work page 2017

  13. [13]

    Linear imaging with sensor arrays on convex polygonal boundaries,

    R. J. Kozick and S. A. Kassam, “Linear imaging with sensor arrays on convex polygonal boundaries,” IEEE Transactions on Systems, Man, and Cybernetics, vol. 21, no. 5, pp. 1155–1166, Sep 1991

    work page 1991

  14. [14]

    Hybrid beamforming for massive MIMO: A survey,

    A. F. Molisch, V . V . Ratnam, S. Han, Z. Li, S. L. H. Nguyen, L. Li, and K. Haneda, “Hybrid beamforming for massive MIMO: A survey,” IEEE Communications Magazine, vol. 55, no. 9, pp. 134–141, September 2017

    work page 2017

  15. [15]

    V ariable- phase-shift-based rf-baseband codesign for MIMO an- tenna selection,

    X. Zhang, A. F. Molisch, and S.-Y . Kung, “V ariable- phase-shift-based rf-baseband codesign for MIMO an- tenna selection,” IEEE Transactions on Signal Process- ing, vol. 53, no. 11, pp. 4091–4103, Nov 2005

    work page 2005

  16. [16]

    On achieving optimal rate of digital precoder by rf-baseband codesign for MIMO sys- tems,

    E. Zhang and C. Huang, “On achieving optimal rate of digital precoder by rf-baseband codesign for MIMO sys- tems,” in IEEE 80th V ehicular T echnology Conference (VTC2014-Fall), Sept 2014, pp. 1–5

    work page 2014

  17. [17]

    Alternating minimization algorithms for hybrid precoding in mil- limeter wave MIMO systems,

    X. Y u, J. Shen, J. Zhang, and K. B. Letaief, “Alternating minimization algorithms for hybrid precoding in mil- limeter wave MIMO systems,” IEEE Journal of Selected T opics in Signal Processing, vol. 10, no. 3, pp. 485–500, April 2016

    work page 2016

  18. [18]

    Hybrid pre- coding for millimeter wave MIMO systems: A matrix factorization approach,

    J. Jin, Y . R. Zheng, W . Chen, and C. Xiao, “Hybrid pre- coding for millimeter wave MIMO systems: A matrix factorization approach,” IEEE Transactions on Wireless Communications, vol. 17, no. 5, pp. 3327–3339, May 2018

    work page 2018

  19. [19]

    Beam-pattern design for hybrid beamforming using wirtinger flow,

    A. Koochakzadeh and P . Pal, “Beam-pattern design for hybrid beamforming using wirtinger flow,” inIEEE 19th International W orkshop on Signal Processing Advances in Wireless Communications (SPAWC), 2018, pp. 1–5

    work page 2018

  20. [20]

    Hybrid digital and analog beam- forming design for large-scale antenna arrays,

    F. Sohrabi and W . Y u, “Hybrid digital and analog beam- forming design for large-scale antenna arrays,” IEEE Journal of Selected T opics in Signal Processing, vol. 10, no. 3, pp. 501–513, April 2016

    work page 2016

  21. [21]

    On the number of rf chains and phase shifters, and scheduling design with hybrid analogdigital beam- forming,

    T. E. Bogale, L. B. Le, A. Haghighat, and L. V anden- dorpe, “On the number of rf chains and phase shifters, and scheduling design with hybrid analogdigital beam- forming,” IEEE Transactions on Wireless Communica- tions, vol. 15, no. 5, pp. 3311–3326, May 2016

    work page 2016

  22. [22]

    MIMO precoding and combining solutions for millimeter-wave systems,

    A. Alkhateeb, J. Mo, N. Gonz´ alez-Prelcic, and R. W . Heath, “MIMO precoding and combining solutions for millimeter-wave systems,”IEEE Communications Mag- azine, vol. 52, no. 12, pp. 122–131, December 2014

    work page 2014

  23. [23]

    Hybrid beamforming for active sensing using sparse arrays,

    R. Rajam¨ aki, S. P . Chepuri, and V . Koivunen, “Hybrid beamforming for active sensing using sparse arrays,” Manuscript in preparation, 2019

    work page 2019

  24. [24]

    Complex-valued ma- trix differentiation: Techniques and key results,

    A. Hjorungnes and D. Gesbert, “Complex-valued ma- trix differentiation: Techniques and key results,” IEEE Transactions on Signal Processing , vol. 55, no. 6, pp. 2740–2746, June 2007

    work page 2007

  25. [25]

    Minimum-redundancy linear arrays,

    A. Moffet, “Minimum-redundancy linear arrays,” IEEE Transactions on Antennas and Propagation , vol. 16, no. 2, pp. 172–175, Mar 1968

    work page 1968

  26. [26]

    Array redundancy for active line arrays,

    R. T. Hoctor and S. A. Kassam, “Array redundancy for active line arrays,” IEEE Transactions on Image Pro- cessing, vol. 5, no. 7, pp. 1179–1183, Jul 1996

    work page 1996

  27. [27]

    A meet-in-the-middle algorithm for find- ing extremal restricted additive 2-bases,

    J. Kohonen, “A meet-in-the-middle algorithm for find- ing extremal restricted additive 2-bases,” Journal of In- teger Sequences, vol. 17, no. 6, 2014

    work page 2014

  28. [28]

    A current distribution for broadside arra ys which optimizes the relationship between beam width and side-lobe level,

    C. L. Dolph, “A current distribution for broadside arra ys which optimizes the relationship between beam width and side-lobe level,” Proceedings of the IRE , vol. 34, no. 6, pp. 335–348, June 1946

    work page 1946

  29. [29]

    Advanced algebraic concepts for efficient multi-channel signal processing,

    F. Roemer, “Advanced algebraic concepts for efficient multi-channel signal processing,” Ph.D. dissertation, Il - menau University of Technology, 2013

    work page 2013