Showing 1–11 of 11 results for author: Caulfield, A

Big Bird: Privacy Budget Management for W3C's Privacy-Preserving Attribution API

Authors: Pierre Tholoniat, Alison Caulfield, Giorgio Cavicchioli, Mark Chen, Nikos Goutzoulias, Benjamin Case, Asaf Cidon, Roxana Geambasu, Mathias Lécuyer, Martin Thomson

Abstract: Privacy-preserving advertising APIs like Privacy-Preserving Attribution (PPA) are designed to enhance web privacy while enabling effective ad measurement. PPA offers an alternative to cross-site tracking with encrypted reports governed by differential privacy (DP), but current designs lack a principled approach to privacy budget management, creating uncertainty around critical design decisions. We… ▽ More Privacy-preserving advertising APIs like Privacy-Preserving Attribution (PPA) are designed to enhance web privacy while enabling effective ad measurement. PPA offers an alternative to cross-site tracking with encrypted reports governed by differential privacy (DP), but current designs lack a principled approach to privacy budget management, creating uncertainty around critical design decisions. We present Big Bird, a privacy budget manager for PPA that clarifies per-site budget semantics and introduces a global budgeting system grounded in resource isolation principles. Big Bird enforces utility-preserving limits via quota budgets and improves global budget utilization through a novel batched scheduling algorithm. Together, these mechanisms establish a robust foundation for enforcing privacy protections in adversarial environments. We implement Big Bird in Firefox and evaluate it on real-world ad data, demonstrating its resilience and effectiveness. △ Less

Submitted 5 June, 2025; originally announced June 2025.

On the Verification of Control Flow Attestation Evidence

Authors: Adam Caulfield, Norrathep Rattanavipanon, Ivan De Oliveira Nunes

Abstract: Remote run-time attestation methods, including Control Flow Attestation (CFA) and Data Flow Attestation (DFA), have been proposed to generate precise evidence of execution's control flow path (in CFA) and optionally execution data inputs (in DFA) on a remote and potentially compromised embedded device, hereby referred to as a Prover (Prv). Recent advances in run-time attestation architectures are… ▽ More Remote run-time attestation methods, including Control Flow Attestation (CFA) and Data Flow Attestation (DFA), have been proposed to generate precise evidence of execution's control flow path (in CFA) and optionally execution data inputs (in DFA) on a remote and potentially compromised embedded device, hereby referred to as a Prover (Prv). Recent advances in run-time attestation architectures are also able to guarantee that a remote Verifier (Vrf) reliably receives this evidence from Prv, even when Prv's software state is fully compromised. This, in theory, enables secure "run-time auditing" in addition to best-effort attestation, i.e., it guarantees that Vrf can examine execution evidence to identify previously unknown compromises as soon as they are exploited, pinpoint their root cause(s), and remediate them. However, prior work has for the most part focused on securely implementing Prv's root of trust (responsible for generating authentic run-time evidence), leaving Vrf 's perspective in this security service unexplored. In this work, we argue that run-time attestation and auditing are only truly useful if Vrf can effectively analyze received evidence. From this premise, we characterize different types of evidence produced by existing run-time attestation/auditing architectures in terms of Vrf 's ability to detect and remediate (previously unknown) vulnerabilities. As a case study for practical uses of run-time evidence by Vrf, we propose SABRE: a Security Analysis and Binary Repair Engine. SABRE showcases how Vrf can systematically leverage run-time evidence to detect control flow attacks, pinpoint corrupted control data and specific instructions used to corrupt them, and leverage this evidence to automatically generate binary patches to buffer overflow and use-after-free vulnerabilities without source code knowledge. △ Less

Submitted 28 April, 2025; v1 submitted 16 November, 2024; originally announced November 2024.

TRACES: TEE-based Runtime Auditing for Commodity Embedded Systems

Authors: Adam Caulfield, Antonio Joia Neto, Norrathep Rattanavipanon, Ivan De Oliveira Nunes

Abstract: Control Flow Attestation (CFA) offers a means to detect control flow hijacking attacks on remote devices, enabling verification of their runtime trustworthiness. CFA generates a trace (CFLog) containing the destination of all branching instructions executed. This allows a remote Verifier (Vrf) to inspect the execution control flow on a potentially compromised Prover (Prv) before trusting that a va… ▽ More Control Flow Attestation (CFA) offers a means to detect control flow hijacking attacks on remote devices, enabling verification of their runtime trustworthiness. CFA generates a trace (CFLog) containing the destination of all branching instructions executed. This allows a remote Verifier (Vrf) to inspect the execution control flow on a potentially compromised Prover (Prv) before trusting that a value/action was correctly produced/performed by Prv. However, while CFA can be used to detect runtime compromises, it cannot guarantee the eventual delivery of the execution evidence (CFLog) to Vrf. In turn, a compromised Prv may refuse to send CFLog to Vrf, preventing its analysis to determine the exploit's root cause and appropriate remediation actions. In this work, we propose TRACES: TEE-based Runtime Auditing for Commodity Embedded Systems. TRACES guarantees reliable delivery of periodic runtime reports even when Prv is compromised. This enables secure runtime auditing in addition to best-effort delivery of evidence in CFA. TRACES also supports a guaranteed remediation phase, triggered upon compromise detection to ensure that identified runtime vulnerabilities can be reliably patched. To the best of our knowledge, TRACES is the first system to provide this functionality on commodity devices (i.e., without requiring custom hardware modifications). To that end, TRACES leverages support from the ARM TrustZone-M Trusted Execution Environment (TEE). To assess practicality, we implement and evaluate a fully functional (open-source) prototype of TRACES atop the commodity ARM Cortex-M33 micro-controller unit. △ Less

Submitted 27 September, 2024; originally announced September 2024.

SpecCFA: Enhancing Control Flow Attestation/Auditing via Application-Aware Sub-Path Speculation

Authors: Adam Caulfield, Liam Tyler, Ivan De Oliveira Nunes

Abstract: At the edge of modern cyber-physical systems, Micro-Controller Units (MCUs) are responsible for safety-critical sensing/actuation. However, MCU cost constraints rule out the usual security mechanisms of general-purpose computers. Thus, various low-cost security architectures have been proposed to remotely verify MCU software integrity. Control Flow Attestation (CFA) enables a Verifier (Vrf) to rem… ▽ More At the edge of modern cyber-physical systems, Micro-Controller Units (MCUs) are responsible for safety-critical sensing/actuation. However, MCU cost constraints rule out the usual security mechanisms of general-purpose computers. Thus, various low-cost security architectures have been proposed to remotely verify MCU software integrity. Control Flow Attestation (CFA) enables a Verifier (Vrf) to remotely assess the run-time behavior of a prover MCU (Prv), generating an authenticated trace of all of Prv control flow transfers (CFLog). Further, Control Flow Auditing architectures augment CFA by guaranteeing the delivery of evidence to Vrf. Unfortunately, a limitation of existing CFA lies in the cost to store and transmit CFLog, as even simple MCU software may generate large traces. Given these issues, prior work has proposed static (context-insensitive) optimizations. However, they do not support configurable program-specific optimizations. In this work, we note that programs may produce unique predictable control flow sub-paths and argue that program-specific predictability can be leveraged to dynamically optimize CFA while retaining all security guarantees. Therefore, we propose SpecCFA: an approach for dynamic sub-path speculation in CFA. SpecCFA allows Vrf to securely speculate on likely control flow sub-paths for each attested program. At run-time, when a sub-path in CFLog matches a pre-defined speculation, the entire sub-path is replaced by a reserved symbol. SpecCFA can speculate on multiple variable-length control flow sub-paths simultaneously. We implement SpecCFA atop two open-source control flow auditing architectures: one based on a custom hardware design and one based on a commodity Trusted Execution Environment (ARM TrustZone-M). In both cases, SpecCFA significantly lowers storage/performance costs that are critical to resource-constrained MCUs. △ Less

Submitted 26 September, 2024; originally announced September 2024.

SoK: Runtime Integrity

Authors: Mahmoud Ammar, Adam Caulfield, Ivan De Oliveira Nunes

Abstract: This paper provides a systematic exploration of Control Flow Integrity (CFI) and Control Flow Attestation (CFA) mechanisms, examining their differences and relationships. It addresses crucial questions about the goals, assumptions, features, and design spaces of CFI and CFA, including their potential coexistence on the same platform. Through a comprehensive review of existing defenses, this paper… ▽ More This paper provides a systematic exploration of Control Flow Integrity (CFI) and Control Flow Attestation (CFA) mechanisms, examining their differences and relationships. It addresses crucial questions about the goals, assumptions, features, and design spaces of CFI and CFA, including their potential coexistence on the same platform. Through a comprehensive review of existing defenses, this paper positions CFI and CFA within the broader landscape of runtime defenses, critically evaluating their strengths, limitations, and trade-offs. The findings emphasize the importance of further research to bridge the gaps in CFI and CFA and thus advance the field of runtime defenses. △ Less

Submitted 21 October, 2024; v1 submitted 19 August, 2024; originally announced August 2024.

Wavefront Threading Enables Effective High-Level Synthesis

Authors: Blake Pelton, Adam Sapek, Ken Eguro, Daniel Lo, Alessandro Forin, Matt Humphrey, Jinwen Xi, David Cox, Rajas Karandikar, Johannes de Fine Licht, Evgeny Babin, Adrian Caulfield, Doug Burger

Abstract: Digital systems are growing in importance and computing hardware is growing more heterogeneous. Hardware design, however, remains laborious and expensive, in part due to the limitations of conventional hardware description languages (HDLs) like VHDL and Verilog. A longstanding research goal has been programming hardware like software, with high-level languages that can generate efficient hardware… ▽ More Digital systems are growing in importance and computing hardware is growing more heterogeneous. Hardware design, however, remains laborious and expensive, in part due to the limitations of conventional hardware description languages (HDLs) like VHDL and Verilog. A longstanding research goal has been programming hardware like software, with high-level languages that can generate efficient hardware designs. This paper describes Kanagawa, a language that takes a new approach to combine the programmer productivity benefits of traditional High-Level Synthesis (HLS) approaches with the expressibility and hardware efficiency of Register-Transfer Level (RTL) design. The language's concise syntax, matched with a hardware design-friendly execution model, permits a relatively simple toolchain to map high-level code into efficient hardware implementations. △ Less

Submitted 10 June, 2024; v1 submitted 29 May, 2024; originally announced May 2024.

Comments: Accepted to PLDI'24

FASTFLOW: Flexible Adaptive Congestion Control for High-Performance Datacenters

Authors: Tommaso Bonato, Abdul Kabbani, Daniele De Sensi, Rong Pan, Yanfang Le, Costin Raiciu, Mark Handley, Timo Schneider, Nils Blach, Ahmad Ghalayini, Daniel Alves, Michael Papamichael, Adrian Caulfield, Torsten Hoefler

Abstract: The increasing demand of machine learning (ML) workloads in datacenters places significant stress on current congestion control (CC) algorithms, many of which struggle to maintain performance at scale. These workloads generate bursty, synchronized traffic that requires both rapid response and fairness across flows. Unfortunately, existing CC algorithms that rely heavily on delay as a primary conge… ▽ More The increasing demand of machine learning (ML) workloads in datacenters places significant stress on current congestion control (CC) algorithms, many of which struggle to maintain performance at scale. These workloads generate bursty, synchronized traffic that requires both rapid response and fairness across flows. Unfortunately, existing CC algorithms that rely heavily on delay as a primary congestion signal often fail to react quickly enough and do not consistently ensure fairness. In this paper, we propose FASTFLOW, a streamlined sender-based CC algorithm that integrates delay, ECN signals, and optional packet trimming to achieve precise, real-time adjustments to congestion windows. Central to FASTFLOW is the QuickAdapt mechanism, which provides accurate bandwidth estimation at the receiver, enabling faster reactions to network conditions. We also show that FASTFLOW can effectively enhance receiver-based algorithms such as EQDS by improving their ability to manage in-network congestion. Our evaluation reveals that FASTFLOW outperforms cutting-edge solutions, including EQDS, Swift, BBR, and MPRDMA, delivering up to 50% performance improvements in modern datacenter networks. △ Less

Submitted 20 September, 2024; v1 submitted 2 April, 2024; originally announced April 2024.

Penalties and Rewards for Fair Learning in Paired Kidney Exchange Programs

Authors: Margarida Carvalho, Alison Caulfield, Yi Lin, Adrian Vetta

Abstract: A kidney exchange program, also called a kidney paired donation program, can be viewed as a repeated, dynamic trading and allocation mechanism. This suggests that a dynamic algorithm for transplant exchange selection may have superior performance in comparison to the repeated use of a static algorithm. We confirm this hypothesis using a full scale simulation of the Canadian Kidney Paired Donation… ▽ More A kidney exchange program, also called a kidney paired donation program, can be viewed as a repeated, dynamic trading and allocation mechanism. This suggests that a dynamic algorithm for transplant exchange selection may have superior performance in comparison to the repeated use of a static algorithm. We confirm this hypothesis using a full scale simulation of the Canadian Kidney Paired Donation Program: learning algorithms, that attempt to learn optimal patient-donor weights in advance via dynamic simulations, do lead to improved outcomes. Specifically, our learning algorithms, designed with the objective of fairness (that is, equity in terms of transplant accessibility across cPRA groups), also lead to an increased number of transplants and shorter average waiting times. Indeed, our highest performing learning algorithm improves egalitarian fairness by 10% whilst also increasing the number of transplants by 6% and decreasing waiting times by 24%. However, our main result is much more surprising. We find that the most critical factor in determining the performance of a kidney exchange program is not the judicious assignment of positive weights (rewards) to patient-donor pairs. Rather, the key factor in increasing the number of transplants, decreasing waiting times and improving group fairness is the judicious assignment of a negative weight (penalty) to the small number of non-directed donors in the kidney exchange program. △ Less

Submitted 23 September, 2023; originally announced September 2023.

Comments: Shorter version accepted in WINE 2023

DiCA: A Hardware-Software Co-Design for Differential Checkpointing in Intermittently Powered Devices

Authors: Antonio Joia Neto, Adam Caulfield, Chistabelle Alvares, Ivan De Oliveira Nunes

Abstract: Intermittently powered devices rely on opportunistic energy-harvesting to function, leading to recurrent power interruptions. This paper introduces DiCA, a proposal for a hardware/software co-design to create differential check-points in intermittent devices. DiCA leverages an affordable hardware module that simplifies the check-pointing process, reducing the check-point generation time and energy… ▽ More Intermittently powered devices rely on opportunistic energy-harvesting to function, leading to recurrent power interruptions. This paper introduces DiCA, a proposal for a hardware/software co-design to create differential check-points in intermittent devices. DiCA leverages an affordable hardware module that simplifies the check-pointing process, reducing the check-point generation time and energy consumption. This hardware module continuously monitors volatile memory, efficiently tracking modifications and determining optimal check-point times. To minimize energy waste, the module dynamically estimates the energy required to create and store the check-point based on tracked memory modifications, triggering the check-pointing routine optimally via a nonmaskable interrupt. Experimental results show the cost-effectiveness and energy efficiency of DiCA, enabling extended application activity cycles in intermittently powered embedded devices. △ Less

Submitted 25 August, 2023; v1 submitted 24 August, 2023; originally announced August 2023.

Comments: 8 pages and 7 figures. To be published at IEEE/ACM International Conference on Computer-Aided Design (ICCAD) 2023

ACFA: Secure Runtime Auditing & Guaranteed Device Healing via Active Control Flow Attestation

Authors: Adam Caulfield, Norrathep Rattanavipanon, Ivan De Oliveira Nunes

Abstract: Low-end embedded devices are increasingly used in various smart applications and spaces. They are implemented under strict cost and energy budgets, using microcontroller units (MCUs) that lack security features available in general-purpose processors. In this context, Remote Attestation (RA) was proposed as an inexpensive security service to enable a verifier (Vrf) to remotely detect illegal modif… ▽ More Low-end embedded devices are increasingly used in various smart applications and spaces. They are implemented under strict cost and energy budgets, using microcontroller units (MCUs) that lack security features available in general-purpose processors. In this context, Remote Attestation (RA) was proposed as an inexpensive security service to enable a verifier (Vrf) to remotely detect illegal modifications to a software binary installed on a low-end prover MCU (Prv). Since attacks that hijack the software's control flow can evade RA, Control Flow Attestation (CFA) augments RA with information about the exact order in which instructions in the binary are executed, enabling detection of control flow attacks. We observe that current CFA architectures can not guarantee that Vrf ever receives control flow reports in case of attacks. In turn, while they support exploit detection, they provide no means to pinpoint the exploit origin. Furthermore, existing CFA requires either binary instrumentation, incurring significant runtime overhead and code size increase, or relatively expensive hardware support, such as hash engines. In addition, current techniques are neither continuous (only meant to attest self-contained operations) nor active (offer no secure means to remotely remediate detected compromises). To jointly address these challenges, we propose ACFA: a hybrid (hardware/software) architecture for Active CFA. ACFA enables continuous monitoring of all control flow transfers in the MCU and does not require binary instrumentation. It also leverages the recently proposed concept of Active Roots-of-Trust to enable secure auditing of vulnerability sources and guaranteed remediation when a compromise is detected. We provide an open-source reference implementation of ACFA on top of a commodity low-end MCU (TI MSP430) and evaluate it to demonstrate its security and cost-effectiveness. △ Less

Submitted 19 October, 2023; v1 submitted 28 March, 2023; originally announced March 2023.

Comments: In 29th USENIX Security Symposium (USENIX Security 23), 2023

ASAP: Reconciling Asynchronous Real-Time Operations and Proofs of Execution in Simple Embedded Systems

Authors: Adam Caulfield, Norrathep Rattanavipanon, Ivan De Oliveira Nunes

Abstract: Embedded devices are increasingly ubiquitous and their importance is hard to overestimate. While they often support safety-critical functions (e.g., in medical devices and sensor-alarm combinations), they are usually implemented under strict cost/energy budgets, using low-end microcontroller units (MCUs) that lack sophisticated security mechanisms. Motivated by this issue, recent work developed ar… ▽ More Embedded devices are increasingly ubiquitous and their importance is hard to overestimate. While they often support safety-critical functions (e.g., in medical devices and sensor-alarm combinations), they are usually implemented under strict cost/energy budgets, using low-end microcontroller units (MCUs) that lack sophisticated security mechanisms. Motivated by this issue, recent work developed architectures capable of generating Proofs of Execution (PoX) for the correct/expected software in potentially compromised low-end MCUs. In practice, this capability can be leveraged to provide "integrity from birth" to sensor data, by binding the sensed results/outputs to an unforgeable cryptographic proof of execution of the expected sensing process. Despite this significant progress, current PoX schemes for low-end MCUs ignore the real-time needs of many applications. In particular, security of current PoX schemes precludes any interrupts during the execution being proved. We argue that lack of asynchronous capabilities (i.e., interrupts within PoX) can obscure PoX usefulness, as several applications require processing real-time and asynchronous events. To bridge this gap, we propose, implement, and evaluate an Architecture for Secure Asynchronous Processing in PoX (ASAP). ASAP is secure under full software compromise, enables asynchronous PoX, and incurs less hardware overhead than prior work. △ Less

Submitted 6 June, 2022; originally announced June 2022.

Comments: 2022 59th ACM/IEEE Design Automation Conference (DAC)