Physics-grounded Mechanism Design for Spectrum Sharing between Passive and Active Users

Jiguang Yu, Nicholas Brendle, Joel T. Johnson, David Starobinski · 2026

Read Physics-grounded Mechanism Design for Spectrum Sharing between Passive and Active Users by Jiguang Yu, Nicholas Brendle, Joel T. Johnson, David Starobinski (2026) 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-2603-18776.txt.

Introduction

Yu, Brendle, Johnson, and Starobinski design a market for sharing spectrum between passive and active users. Radiometers need quiet time-frequency tiles to keep retrieval error low. Their Vickrey-Clarke-Groves auction buys quiet tiles from active users. Payment follows each tile's marginal cut in error variance. The design secures efficient allocation and honest bidding. Exact auctions are too heavy on large grids, so a submodularity-based approximation steps in. Simulations with AMSR-2 data cut procurement costs by about 60 percent versus a fixed-band plan, while accuracy targets still hold.

Catalog record

CREATORJiguang Yu, Nicholas Brendle, Joel T. Johnson, David Starobinski
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE44980
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.18776

Browse the full library · About the Tower of Babel

Full text

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

Physics-grounded Mechanism Design for Spectrum Sharing between Passive and Active Users Thanks: This work was supported in part by the NSF, under grants AST-2229103 and AST-2229104. Jiguang Yu, Nicholas Brendle, Joel T. Johnson, David Starobinski Abstract We propose a physics-grounded mechanism design for dynamic spectrum sharing that bridges the gap between radiometric retrieval constraints and economic incentives. We formulate the active and passive users coexistence problem as a Vickrey-Clarke-Groves (VCG) auctions mechanism, where the radiometer dynamically procures ”quiet” time-frequency tiles from active users based on the marginal reduction in retrieval error variance. This approach ensures allocative efficiency and dominant-strategy incentive compatibility (DSIC). To overcome the computational intractability of exact VCG on large grids, we derive an approximation algorithm by using the monotone submodularity induced by the radiometer equation. AMSR-2–based simulations show that the approach avoids high-cost tiles by aggregating low-cost spectrum across time and frequency. In an interference-trap case study, the proposed framework reduces procurement costs by about 60% over a fixed-band baseline while satisfying accuracy targets. I Introduction Passive microwave radiometers aboard the Earth Exploration-Satellite Service (EESS) provide key observations of geophysical variables such as integrated water vapor, near-surface wind speed, and sea-surface temperature [ 11 , 21 ] . These instruments operate in frequency bands that increasingly overlap with active terrestrial services. With radio spectrum becoming increasingly congested, the traditional model of passive sensing (large frequency bands are set aside exclusively and permanently) is untenable, creating friction between scientific observation and commercial use [ 22 ] . Fig. 1: Comparison of coexistence paradigms on a discretized time–frequency grid with three channels. (Top) Regime 1: Static Partitioning. The 23.8 GHz band (Ch 1) is permanently allocated to passive sensing (blue), resulting in the complete exclusion of active users in this band regardless of their costs. (Middle) Regime 2: Time-Dynamic Sharing. Spectrum access is regulated by a fixed integration window \tau . All channels are silenced simultaneously during the overpass, blocking high-value active users in Ch 2 and Ch 3 while failing to use low-cost periods outside the window. (Bottom) Regime 3: Fully Flexible Sharing (Proposed). The market-based mechanism dynamically selects a subset of tiles with the lowest opportunity costs (lightest red background) scattered across the grid. This trade-off lead to a progression toward flexible coexistence. Fig. 1 summarizes these sharing regimes over a radiometric integration window. While static partitioning globally excludes active systems and yields poor spectral efficiency, fully flexible time–frequency sharing allows the radiometer to dynamically procure measurement passbands based on the instantaneous opportunity costs of active users [ 3 ] . We formalize this flexibility using the basic physics of measurement. During a single integration window \tau , the time-frequency plane is discretized into tiles of bandwidth \delta f_{x} and duration \delta t_{x}\leq\tau . Each tile x has a clean duty cycle \alpha_{x}\in[0,1] , capturing partial-time muting or hardware latencies. For a procured set of quiet tiles S temporally coincident with \tau [ 12 ] , the effective clean bandwidth B_{j}(S) for channel j is the baseline protected bandwidth plus the normalized allocated spectral volume [ 25 ] . The radiometer equation dictates that thermal-noise variance \sigma_{j}^{2} is inversely proportional to the time-bandwidth product B_{j}(S)\tau [ 13 , 12 ] . This allows quiet spectrum to be flexibly aggregated across time and frequency. Nontrivial cross-channel elasticity ensures comparable product accuracy can be maintained under varying allocations [ 2 ] . Multi-channel retrievals map these spectral resources directly to mission performance. Because geophysical products are typically estimated as weighted linear combinations of channel brightness temperatures [ 27 ] , the total retrieval error variance is the weighted sum of uncorrelated channel noise variances [ 14 ] . This linkage quantifies the marginal value of spectrum as the retrieval-error reduction per additional Hz of effective bandwidth. Treating quiet spectrum as a dynamically procured commodity rather than a permanent entitlement raises three fundamental questions: 1. How should the radiometer quantify its valuation for quiet tiles given multi-channel coupling and accuracy constraints? 2. How can a mechanism elicit private cost information from active users to maximize social welfare while ensuring incentive compatibility? 3. Can scalable algorithms approximate optimal allocations on large time–frequency grids? This work addresses these questions through the following contributions: 1. Physics-Driven Valuation Model: We derive the radiometer’s utility explicitly from retrieval physics [ 27 , 15 ] , establishing a closed-form expression for the marginal value density of clean bandwidth based on mask constraints and retrieval errors. 2. VCG-Based Procurement Formulation: We frame the problem as a multi-unit reverse auction with a tailored VCG mechanism, guaranteeing dominant-strategy incentive compatibility (DSIC) and allocative efficiency [ 29 , 5 , 9 ] . 3. Scalable Approximation via Submodularity: To bypass the computationally complex VCG benchmark, we leverage the radiometer equation [ 13 ] to show that the variance-reduction objective is monotone submodular ( \sigma^{2}\propto 1/B ) [ 23 ] . We design a scalable greedy posted-price algorithm with an analytical approximation bound. Numerically, this achieves a 4.45% average optimality gap and reduces procurement costs by approximately 60% in an “interference trap” scenario compared to static allocation. We envision that regulators would procure EESS users with credits or a virtual currency to participate in the bids. Active users winning the auction could be rewarded with subsidies or tax rebates from the government [ 16 ] . Design of such a reward system requires care, and should be considered as part of future work. TABLE I: Definition of Parameters and Variables Symbol Definition Unit / Domain Sets and Indices \mathcal{N},i Set of active users (sellers) and user index \{1,\dots,N\} \mathcal{J},j Set of radiometer channels and channel index \{1,\dots,J\} \Omega,x Set of time–frequency tiles and tile index Set of available time-frequency tiles S Allocation set (subset of tiles designated as “quiet”) S\subseteq\Omega \mathcal{A} Set of feasible allocations satisfying accuracy constraints \mathcal{A}\subseteq 2^{\Omega} Physics and Retrieval Model \tau Radiometric integration window duration s \delta f_{x},\delta t_{x} Frequency width and time duration of tile x Hz, s \alpha_{x} Duty cycle (fraction of quiet time) for tile x [0,1] B_{j}(S) Effective clean bandwidth for channel j Hz \kappa_{j} Radiometer system noise constant ( \propto T_{\mathrm{sys}}^{2} ) K {}^{2}\cdot Hz \cdot s \sigma_{j}^{2} Total measurement error variance in channel j K 2 c_{k,j} Retrieval sensitivity coefficient for product k , channel j Unit k / K \mathrm{Var}[\hat{y}_{k}] Retrieval error variance for geophysical product k Unit {}_{k}^{2} \varepsilon_{k}^{2} Maximum permissible error variance (Mission Constraint) Unit {}_{k}^{2} \phi_{j}(M_{j}) Residual RFI penalty function K 2 Mechanism Design and Optimization v_{0}(S) Radiometer (Buyer) valuation for allocation S Currency ($) C_{i}(S_{i}) Opportunity cost function for seller i Currency ($) W(S) Social welfare function Currency ($) p_{i} VCG payment to seller i Currency ($) \lambda_{k} Dual variable (shadow price) for constraint k $ / Unit {}_{k}^{2} v_{j} Marginal value density of bandwidth in channel j $ / Hz F(S) Discrete submodular utility function Unitless / Scaled II Related Work This work bridges passive microwave remote sensing, dynamic spectrum access (DSA), mechanism design, and submodular optimization. We review related work to highlight the specific research gaps addressed by our framework. II-A Passive Microwave Radiometry and Retrieval Modeling Passive microwave radiometry provides global atmospheric and surface observations [ 28 ] , employing variational or regression-based

Read the complete text