arXiv subjects

Neal Patwari

Publications and source records attributed to Neal Patwari.

At least 19 recordsLinked to original sources

arXiv2026

Evaluating Bias in Phoneme-Based Automatic Speech Recognition Systems: An Analysis of IPA Transcription Models

The popularization of automatic speech recognition (ASR) systems has increased exploration of the demographic biases related to race, age, gender, and accent, often formed from imbalanced training data. Most of these studies focused on standard grapheme-based ASR systems with comparatively little emphasis on phoneme-based systems, such as models that produce International Phonetic Alphabet (IPA) representations. As ASR systems shift toward multilingual support and low-resource language modeling, IPA-based layers serve as a critical, language-agnostic foundation. In this study, we evaluate the performance of two state-of-the-art open-source ASR systems, WhisperIPA and ZIPA, that generate IPA transcriptions across diverse accents and language sources. Our evaluation includes existing multilingual speech corpora and demographically annotated English-language corpora. We measure model performance by comparing model-generated IPA transcriptions against grapheme-to-phoneme (G2P) systems using both standard phoneme error rate (PER) and a proposed Soft PER metric that tolerates linguistically similar phoneme substitutions. Our analysis examines how performance varies across languages and demographic groups such as gender, accent, ethnicity, and age, revealing persistent disparities even after accounting for acceptable phonemic variation. These findings provide insight into potential sources of bias and inform the development of more inclusive and linguistically robust phoneme-based ASR systems. Our code and data will be made publicly available to the community.

cs.CL↗

arXiv2025

RFI Detection and Identification at OVRO Using Pseudonymetry

Protecting passive radio astronomy observatories from unintended radio-frequency interference (RFI) is increasingly challenging as wireless activity expands near protected bands. While radio quiet zones, database-driven coordination, and post-processing mitigation can reduce interference risk, they often lack the ability to attribute detected RFI to a specific transmitter, particularly in low signal-to-noise ratio (SNR) regimes where conventional demodulation is infeasible. This paper presents the first over-the-air field demonstration of Pseudonymetry at the Owens Valley Radio Observatory (OVRO), evaluating an accountable coexistence approach between heterogeneous systems: an SDR-based narrowband OFDM transmitter and a wideband radio telescope backend. The transmitter embeds a pseudonym watermark on a dedicated OFDM subcarrier using coded power modulation, while OVRO passively extracts the watermark from standard backend spectrogram (power) products without IQ access. We develop a spectrogram-only receiver that performs correlation-based packet alignment, compensates timing resolution mismatch via resampling, and decodes pseudonym bits using energy-domain template matching. Field results across -20 to -5 dB SNR show that pseudonym watermarks can be recovered at low SNR, enabling practical transmitter attribution using only passive backend measurements. These findings suggest that observatories can support lightweight accountability mechanisms that complement dynamic protection and enforcement-oriented spectrum sharing frameworks.

eess.SP↗

arXiv2025

Sensing and Stopping Interfering Secondary Users: Validation of an Efficient Spectrum Sharing System

We present the design and validation of Stoppable Secondary Use (StopSec), a privacy-preserving protocol with the capability to identify a secondary user (SU) causing interference to a primary user (PU) and to act quickly to stop the interference. All users are served by a database that provides a feedback mechanism from a PU to an interfering SU. We introduce a new lightweight and robust method to watermark an SU's OFDM packet. Through extensive over-the-air real-time experiments, we evaluate StopSec in terms of interference detection, identification, and stopping latency, as well as impact on SUs. We show that the watermarking method avoids negative impact to the secondary data link and is robust to real-world time-varying channels. Interfering SUs can be stopped in under 150 milliseconds, and when multiple users are simultaneously interfering, they can all be stopped. Even when the interference is 10 dB lower than the noise power, StopSec successfully stops interfering SUs within a few seconds of their appearance in the channel. StopSec can be an effective spectrum sharing protocol for cases when interference to a PU must be quickly and automatically stopped.

eess.SP↗

arXiv2023

Channel Estimation via Loss Field: Accurate Site-Trained Modeling for Shadowing Prediction

Future mobile ad hoc networks will share spectrum between many users. Channels will be assigned on the fly to guarantee signal and interference power requirements for requested links. Channel losses must be re-estimated between many pairs of users as they move and as environmental conditions change. Computational complexity must be low, precluding the use of some accurate but computationally intensive site-specific channel models. Channel model errors must be low, precluding the use of standard statistical channel models. We propose a new channel model, CELF, which uses channel loss measurements from a deployed network in the area and a Bayesian linear regression method to estimate a site-specific loss field for the area. The loss field is explainable as the site's 'shadowing' of the radio propagation across the area of interest, but it requires no site-specific terrain or building information. Then, for any arbitrary pair of transmitter and receiver positions, CELF sums the loss field near the link line to estimate its channel loss. We use extensive measurements to show that CELF lowers the variance of channel estimates by up to 56%. It outperforms 3 popular machine learning methods in variance reduction and training efficiency.

eess.SP↗

arXiv2022

Two Measure is Two Know: Calibration-free Full Duplex Monitoring for Software Radio Platforms

Future virtualized radio access network (vRAN) infrastructure providers (and today's experimental wireless testbed providers) may be simultaneously uncertain what signals are being transmitted by their base stations and legally responsible for their violations. These providers must monitor the spectrum of transmissions and external signals without access to the radio itself. In this paper, we propose FDMonitor, a full-duplex monitoring system attached between a transmitter and its antenna to achieve this goal. Measuring the signal at this point on the RF path is necessary but insufficient since the antenna is a bidirectional device. FDMonitor thus uses a bidirectional coupler, a two-channel receiver, and a new source separation algorithm to simultaneously estimate the transmitted signal and the signal incident on the antenna. Rather than requiring an offline calibration, we also adaptively estimate the linear model for the system on the fly. FDMonitor has been running on a real-world open wireless testbed, monitoring 19 SDR platforms controlled (with bare metal access) by outside experimenters over a seven month period, sending alerts whenever a violation is observed. Our experimental results show that FDMonitor accurately separates signals across a range of signal parameters. Over more than 7 months of observation, it achieves a positive predictive value of 97%, with a total of 20 false alerts.

eess.SP↗

arXiv2022

Racial Disparities in Pulse Oximetry Cannot Be Fixed With Race-Based Correction

Studies have shown pulse oximeter measurements of blood oxygenation have statistical bias that is a function of race, which results in higher rates of occult hypoxemia, i.e., missed detection of dangerously low oxygenation, in patients of color. This paper further characterizes the statistical distribution of pulse ox measurements, showing they also have a higher variance for patients racialized as Black, compared to those racialized as white. We show that no single race-based correction factor will provide equal performance in the detection of hypoxemia. The results have implications for racially equitable pulse oximetry.

eess.SP↗

arXiv2022

Global Performance Disparities Between English-Language Accents in Automatic Speech Recognition

Past research has identified discriminatory automatic speech recognition (ASR) performance as a function of the racial group and nationality of the speaker. In this paper, we expand the discussion beyond bias as a function of the individual national origin of the speaker to look for bias as a function of the geopolitical orientation of their nation of origin. We audit some of the most popular English language ASR services using a large and global data set of speech from The Speech Accent Archive, which includes over 2,700 speakers of English born in 171 different countries. We show that, even when controlling for multiple linguistic covariates, ASR service performance has a statistically significant relationship to the political alignment of the speaker's birth country with respect to the United States' geopolitical power. This holds for all ASR services tested. We discuss this bias in the context of the historical use of language to maintain global and political power.

cs.CL↗

arXiv2021

Precarity: Modeling the Long Term Effects of Compounded Decisions on Individual Instability

When it comes to studying the impacts of decision making, the research has been largely focused on examining the fairness of the decisions, the long-term effects of the decision pipelines, and utility-based perspectives considering both the decision-maker and the individuals. However, there has hardly been any focus on precarity which is the term that encapsulates the instability in people's lives. That is, a negative outcome can overspread to other decisions and measures of well-being. Studying precarity necessitates a shift in focus - from the point of view of the decision-maker to the perspective of the decision subject. This centering of the subject is an important direction that unlocks the importance of parting with aggregate measures to examine the long-term effects of decision making. To address this issue, in this paper, we propose a modeling framework that simulates the effects of compounded decision-making on precarity over time. Through our simulations, we are able to show the heterogeneity of precarity by the non-uniform ruinous aftereffects of negative decisions on different income classes of the underlying population and how policy interventions can help mitigate such effects.

cs.AI↗

arXiv2020

Collision Prediction from UWB Range Measurements

The ability to predict, and thus react to, oncoming collisions among a set of mobile agents is a fundamental requirement for safe autonomous movement, both human and robotic. This paper addresses systems that use range measurements between mobile agents for the purpose of collision prediction, which involves prediction of the agents' future paths to know if they will collide at any time. One straightforward system would use known-location static anchors to estimate agent coordinates over time, and use the track to predict collision. Fundamentally, no fixed coordinate system is required for collision prediction, so using only the pairwise range between two agents can be used to predict collision. We present lower bound analysis which shows the limitations of this pairwise method. As an alternative anchor-free method, we propose the friend-based autonomous collision prediction and tracking (FACT) method that uses all measured ranges between nearby (unknown location mobile) agents, in a distributed algorithm, to estimate their relative locations and velocities and predict future collisions between agents. Using analysis and simulation, we show the potential for FACT to achieve equal or better collision detection performance compared to other methods, while avoiding the need for anchors. We then build a network of $N$ ultra wideband (UWB) devices and an efficient multi-node protocol which allows all ${\cal O}(N^2)$ pairwise ranges to be measured in $N$ slots. We run experiments with up to six independent robot agents moving and colliding in a 2D plane and up to four anchor nodes to compare the performance of the collision prediction methods. We show that the FACT method can perform better than either other method but without the need for a fixed infrastructure of anchor nodes.

eess.SP↗

arXiv2020

Quantifying Interference-Assisted Signal Strength Surveillance of Sound Vibrations

A malicious attacker could, by taking control of internet-of-things devices, use them to capture received signal strength (RSS) measurements and perform surveillance on a person's vital signs, activities, and sound in their environment. This article considers an attacker who looks for subtle changes in the RSS in order to eavesdrop sound vibrations. The challenge to the adversary is that sound vibrations cause very low amplitude changes in RSS, and RSS is typically quantized with a significantly larger step size. This article contributes a lower bound on an attacker's monitoring performance as a function of the RSS step size and sampling frequency so that a designer can understand their relationship. Our bound considers the little-known and counter-intuitive fact that an adversary can improve their sinusoidal parameter estimates by making some devices transmit to add interference power into the RSS measurements. We demonstrate this capability experimentally. As we show, for typical transceivers, the RSS surveillance attacker can monitor sound vibrations with remarkable accuracy. New mitigation strategies will be required to prevent RSS surveillance attacks.

eess.SP↗

arXiv2019

Sitara: Spectrum Measurement Goes Mobile Through Crowd-sourcing

Software-defined radios (SDRs) are often used in the experimental evaluation of next-generation wireless technologies. While crowd-sourced spectrum monitoring is an important component of future spectrum-agile technologies, there is no clear way to test it in the real world, i.e., with hundreds of users each with an SDR in their pocket participating in RF experiments controlled by, and data uploaded to, the cloud. Current fully functional SDRs are bulky, with components connected via wires, and last at most hours on a single battery charge. To address the needs of such experiments, we design and develop a compact, portable, untethered, and inexpensive SDR we call Sitara. Our SDR interfaces with a mobile device over Bluetooth 5 and can function standalone or as a client to a central command and control server. The Sitara offers true portability: it operates up to one week on battery power, requires no external wired connections and occupies a footprint smaller than a credit card. It transmits and receives common waveforms, uploads IQ samples or processed receiver data through a mobile device to a server for remote processing and performs spectrum sensing functions. Multiple Sitaras form a distributed system capable of conducting experiments in wireless networking and communication in addition to RF monitoring and sensing activities. In this paper, we describe our design, evaluate our solution, present experimental results from multi-sensor deployments and discuss the value of this system in future experimentation.

cs.NI↗

arXiv2019

Quantifying an Interference-Assisted Signal Strength Breathing Surveillance Attack

A malicious attacker could, by taking control of internet-of-things devices, use them to capture received signal strength (RSS) measurements and perform surveillance on a person's vital signs, activities, audio in their environment, and other RF sensing capabilities. This paper considers an attacker who looks for periodic changes in the RSS in order to surveil a person's breathing rate. The challenge to the attacker is that a person's breathing causes a low amplitude change in RSS, and RSS is typically quantized with a significantly larger step size. This paper contributes a lower bound on an attacker's breathing monitoring performance as a function of the RSS step size and sampling frequency so that a designer can understand their relationship. Our bound considers the little-known and counter-intuitive fact that an adversary can improve their sinusoidal parameter estimates by making some devices transmit to add interference power into the RSS measurements. We demonstrate this capability experimentally. As we show, for typical transceivers, the RSS surveillance attack can monitor RSS with remarkable accuracy. New mitigation strategies will be required to prevent RSS surveillance attacks.

eess.SP↗

arXiv2018

Never Use Labels: Signal Strength-Based Bayesian Device-Free Localization in Changing Environments

Device-free localization (DFL) methods use measured changes in the received signal strength (RSS) between many pairs of RF nodes to provide location estimates of a person inside the wireless network. Fundamental challenges for RSS DFL methods include having a model of RSS measurements as a function of a person's location, and maintaining an accurate model as the environment changes over time. Current methods rely on either labeled empty-area calibration or labeled fingerprints with a person at each location. Both need to be frequently recalibrated or retrained to stay current with changing environments. Other DFL methods only localize people in motion. In this paper, we address these challenges by, first, introducing a new mixture model for link RSS as a function of a person's location, and second, providing the framework to update model parameters without ever being provided labeled data from either empty-area or known-location classes. We develop two new Bayesian localization methods based on our mixture model and experimentally validate our system at three test sites with seven days of measurements. We demonstrate that our methods localize a person with non-degrading performance in changing environments, and, in addition, reduce localization error by 11-51% compared to other DFL methods.

cs.NI↗

arXiv2018

Save Our Spectrum: Contact-Free Human Sensing Using Single Carrier Radio

Recent research has demonstrated new capabilities in radio frequency (RF) sensing that apply to health care, smart home, and security applications. However, previous work in RF sensing requires heavy utilization of the radio spectrum, for example, transmitting thousands of WiFi packets per second. In this paper, we present a device-free human sensing system based on received signal strength (RSS) measurements from a low-cost single carrier narrowband radio transceiver. We test and validate its performance in three different applications: real-time heart rate monitoring, gesture recognition, and human speed estimation. The challenges in these applications stem from the very low signal-to-noise ratio and the use of a single-dimensional measurement of the channel. We apply a combination of linear and non-linear filtering, and time-frequency analysis, and develop new estimators to address the challenges in the particular applications. Our experimental results indicate that RF sensing based on single-carrier magnitude measurements performs nearly as well as the state-of-the-art while utilizing three orders of magnitude less bandwidth.

eess.SP↗

arXiv2018

Comparing Respiratory Monitoring Performance of Commercial Wireless Devices

This paper addresses the performance of systems which use commercial wireless devices to make bistatic RF channel measurements for non-contact respiration sensing. Published research has typically presented results from short controlled experiments on one system. In this paper, we deploy an extensive real-world comparative human subject study. We observe twenty patients during their overnight sleep (a total of 160 hours), during which contact sensors record ground-truth breathing data, patient position is recorded, and four different RF breathing monitoring systems simultaneously record measurements. We evaluate published methods and algorithms. We find that WiFi channel state information measurements provide the most robust respiratory rate estimates of the four RF systems tested. However, all four RF systems have periods during which RF-based breathing estimates are not reliable.

eess.SP↗

arXiv2017

RSS Models for Respiration Rate Monitoring

Received signal strength based respiration rate monitoring is emerging as an alternative non-contact technology. These systems make use of the radio measurements of short-range commodity wireless devices, which vary due to the inhalation and exhalation motion of a person. The success of respiration rate estimation using such measurements depends on the signal-to-noise ratio, which alters with properties of the person and with the measurement system. To date, no model has been presented that allows evaluation of different deployments or system configurations for successful breathing rate estimation. In this paper, a received signal strength model for respiration rate monitoring is introduced. It is shown that measurements in linear and logarithmic scale have the same functional form, and the same estimation techniques can be used in both cases. The implications of the model are validated under varying signal-to-noise ratio conditions using the performances of three estimators: batch frequency estimator, recursive Bayesian estimator, and model-based estimator. The results are in coherence with the findings, and they imply that different estimators are advantageous in different signal-to-noise ratio regimes.

eess.SP↗

arXiv2017

EpiFi: An In-Home Sensor Network Architecture for Epidemiological Studies

We design and build a system called EpiFi, which allows epidemiologists to easily design and deploy experiments in homes. The focus of EpiFi is reducing the barrier to entry for deploying and using an in-home sensor network. We present a novel architecture for in-home sensor networks configured using a single configuration file and provide: a fast and reliable method for device discovery when installed in the home, a new mechanism for sensors to authenticate over the air using a subject's home WiFi router, and data reliability mechanisms to minimize loss in the network through a long-term deployment. We work collaboratively with pediatric asthma researchers to design three studies and deploy EpiFi in homes.

cs.NI↗

arXiv2016

Detector Based Radio Tomographic Imaging

Received signal strength based radio tomographic imaging is a popular device-free indoor localization method which reconstructs the spatial loss field of the environment using measurements from a dense wireless network. Existing methods solve an associated inverse problem using algebraic or compressed sensing reconstruction algorithms. We propose an alternative imaging method that reconstructs spatial field of occupancy using a back-projection based reconstruction algorithm. The introduced system has the following advantages over the other imaging based methods: i.) significantly lower computational complexity such that no floating point multiplication is required; ii.) each link's measured data is compressed to a single bit, providing improved scalability; iii.) physically significant and repeatable parameters. The proposed method is validated using measurement data. Results show that the proposed method achieves the above advantages without loss of accuracy compared to the other available methods.

cs.ET↗