Philip Levis
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Philip Levis has a total of 23 co-authors. Here are the top co-authors by number of shared publications:
- Chinmayee Shah
Stanford University
3 shared publications - Hang Qu
Stanford University
3 shared publications - Omid Mashayekhi
Stanford University
2 shared publications - Alberto Cerpa
University of California, Merced, Merced, CA, USA
1 shared publication - Amit Levy
Stanford University Stanford University Stanford University Stanford University Princeton University Stanford University
1 shared publication Andrew Lim
Stanford University
1 shared publication
Latest Publications
- 19 November 2025
- Published by Springer Nature in Optimization and Engineering
- Vol. 27 (2) , 1303-1326
Aging-aware battery control via convex optimization
- 19 November 2025
- Published by Springer Nature in Optimization and Engineering
- Vol. 27 (2) , 1303-1326
Aging-aware battery control via convex optimization
- Open Access
- 1 February 2020
- Published by Wiley in Computer Graphics Forum
- Vol. 39 (1) , 375-388
Accelerating Distributed Graphical Fluid Simulations with Micro-partitioning
- 1 February 2020
- Published by Wiley in Computer Graphics Forum
- Vol. 39 (1) , 375-388
Open AccessAccelerating Distributed Graphical Fluid Simulations with Micro-partitioning
- Open Access
Design Considerations for Low Power Internet Protocols
Over the past 10 years, low-power wireless networks have transitioned to supporting IPv6 connectivity through 6LoWPAN, a set of standards which specify how to aggressively compress IPv6 packets over low-power wireless links such as 802.15.4. We find that different low-power IPv6 stacks are unable to communicate using 6LoWPAN, and therefore IP, due to design tradeoffs between code size and energy efficiency. We argue that applying traditional protocol design principles to low-power networks is responsible for these failures, in part because receivers must accommodate a wide range of senders. Based on these findings, we propose three design principles for Internet protocols on low-power networks. These principles are based around the importance of providing flexible tradeoffs between code size and energy efficiency. We apply these principles to 6LoWPAN and show that the resulting design of the protocol provides developers a wide range of tradeoff points while allowing implementations with different choices to seamlessly communicate.
Open AccessDesign Considerations for Low Power Internet Protocols
Over the past 10 years, low-power wireless networks have transitioned to supporting IPv6 connectivity through 6LoWPAN, a set of standards which specify how to aggressively compress IPv6 packets over low-power wireless links such as 802.15.4. We find that different low-power IPv6 stacks are unable to communicate using 6LoWPAN, and therefore IP, due to design tradeoffs between code size and energy efficiency. We argue that applying traditional protocol design principles to low-power networks is responsible for these failures, in part because receivers must accommodate a wide range of senders. Based on these findings, we propose three design principles for Internet protocols on low-power networks. These principles are based around the importance of providing flexible tradeoffs between code size and energy efficiency. We apply these principles to 6LoWPAN and show that the resulting design of the protocol provides developers a wide range of tradeoff points while allowing implementations with different choices to seamlessly communicate.
- Open Access
- 27 May 2018
- Published by Association for Computing Machinery (ACM)
- Vol. 2018, 1301-1316
Fonduer
We focus on knowledge base construction (KBC) from richly formatted data. In contrast to KBC from text or tabular data, KBC from richly formatted data aims to extract relations conveyed jointly via textual, structural, tabular, and visual expressions. We introduce Fonduer, a machine-learning-based KBC system for richly formatted data. Fonduer presents a new data model that accounts for three challenging characteristics of richly formatted data: (1) prevalent document-level relations, (2) multimodality, and (3) data variety. Fonduer uses a new deep-learning model to automatically capture the representation (i.e., features) needed to learn how to extract relations from richly formatted data. Finally, Fonduer provides a new programming model that enables users to convert domain expertise, based on multiple modalities of information, to meaningful signals of supervision for training a KBC system. Fonduer-based KBC systems are in production for a range of use cases, including at a major online retailer. We compare Fonduer against state-of-the-art KBC approaches in four different domains. We show that Fonduer achieves an average improvement of 41 F1 points on the quality of the output knowledge base---and in some cases produces up to 1.87x the number of correct entries---compared to expert-curated public knowledge bases. We also conduct a user study to assess the usability of Fonduer's new programming model. We show that after using Fonduer for only 30 minutes, non-domain experts are able to design KBC systems that achieve on average 23 F1 points higher quality than traditional machine-learning-based KBC approaches.
- 27 May 2018
- Published by Association for Computing Machinery (ACM)
- Vol. 2018, 1301-1316
Open AccessFonduer
We focus on knowledge base construction (KBC) from richly formatted data. In contrast to KBC from text or tabular data, KBC from richly formatted data aims to extract relations conveyed jointly via textual, structural, tabular, and visual expressions. We introduce Fonduer, a machine-learning-based KBC system for richly formatted data. Fonduer presents a new data model that accounts for three challenging characteristics of richly formatted data: (1) prevalent document-level relations, (2) multimodality, and (3) data variety. Fonduer uses a new deep-learning model to automatically capture the representation (i.e., features) needed to learn how to extract relations from richly formatted data. Finally, Fonduer provides a new programming model that enables users to convert domain expertise, based on multiple modalities of information, to meaningful signals of supervision for training a KBC system. Fonduer-based KBC systems are in production for a range of use cases, including at a major online retailer. We compare Fonduer against state-of-the-art KBC approaches in four different domains. We show that Fonduer achieves an average improvement of 41 F1 points on the quality of the output knowledge base---and in some cases produces up to 1.87x the number of correct entries---compared to expert-curated public knowledge bases. We also conduct a user study to assess the usability of Fonduer's new programming model. We show that after using Fonduer for only 30 minutes, non-domain experts are able to design KBC systems that achieve on average 23 F1 points higher quality than traditional machine-learning-based KBC approaches.
- Open Access
Scalable, Fast Cloud Computing with Execution Templates
Large scale cloud data analytics applications are often CPU bound. Most of these cycles are wasted: benchmarks written in C++ run 10-51 times faster than frameworks such as Naiad and Spark. However, calling faster implementations from those frameworks only sees moderate (3-5x) speedups because their control planes cannot schedule work fast enough. This paper presents execution templates, a control plane abstraction for CPU-bound cloud applications, such as machine learning. Execution templates leverage highly repetitive control flow to cache scheduling decisions as {\it templates}. Rather than reschedule hundreds of thousands of tasks on every loop execution, nodes instantiate these templates. A controller's template specifies the execution across all worker nodes, which it partitions into per-worker templates. To ensure that templates execute correctly, controllers dynamically patch templates to match program control flow. We have implemented execution templates in Nimbus, a C++ cloud computing framework. Running in Nimbus, analytics benchmarks can run 16-43 times faster than in Naiad and Spark. Nimbus's control plane can scale out to run these faster benchmarks on up to 100 nodes (800 cores).
Open AccessScalable, Fast Cloud Computing with Execution Templates
Large scale cloud data analytics applications are often CPU bound. Most of these cycles are wasted: benchmarks written in C++ run 10-51 times faster than frameworks such as Naiad and Spark. However, calling faster implementations from those frameworks only sees moderate (3-5x) speedups because their control planes cannot schedule work fast enough. This paper presents execution templates, a control plane abstraction for CPU-bound cloud applications, such as machine learning. Execution templates leverage highly repetitive control flow to cache scheduling decisions as {\it templates}. Rather than reschedule hundreds of thousands of tasks on every loop execution, nodes instantiate these templates. A controller's template specifies the execution across all worker nodes, which it partitions into per-worker templates. To ensure that templates execute correctly, controllers dynamically patch templates to match program control flow. We have implemented execution templates in Nimbus, a C++ cloud computing framework. Running in Nimbus, analytics benchmarks can run 16-43 times faster than in Naiad and Spark. Nimbus's control plane can scale out to run these faster benchmarks on up to 100 nodes (800 cores).
- Open Access
Distributed Graphical Simulation in the Cloud
Graphical simulations are a cornerstone of modern media and films. But existing software packages are designed to run on HPC nodes, and perform poorly in the computing cloud. These simulations have complex data access patterns over complex data structures, and mutate data arbitrarily, and so are a poor fit for existing cloud computing systems. We describe a software architecture for running graphical simulations in the cloud that decouples control logic, computations and data exchanges. This allows a central controller to balance load by redistributing computations, and recover from failures. Evaluations show that the architecture can run existing, state-of-the-art simulations in the presence of stragglers and failures, thereby enabling this large class of applications to use the computing cloud for the first time.
Open AccessDistributed Graphical Simulation in the Cloud
Graphical simulations are a cornerstone of modern media and films. But existing software packages are designed to run on HPC nodes, and perform poorly in the computing cloud. These simulations have complex data access patterns over complex data structures, and mutate data arbitrarily, and so are a poor fit for existing cloud computing systems. We describe a software architecture for running graphical simulations in the cloud that decouples control logic, computations and data exchanges. This allows a central controller to balance load by redistributing computations, and recover from failures. Evaluations show that the architecture can run existing, state-of-the-art simulations in the presence of stragglers and failures, thereby enabling this large class of applications to use the computing cloud for the first time.
- 1 April 2007
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 21-30
Improving Wireless Simulation Through Noise Modeling
We propose modeling environmental noise in order to efficiently and accurately simulate wireless packet delivery. We measure noise traces in many different environments and propose three algorithms to simulate noise from these traces. We evaluate applying these algorithms to signal-to-noise curves in comparison to existing simulation approaches used in EmStar, TOSSIM, and ns2. We measure simulation accuracy using the Kantorovich-Wasserstein distance on conditional packet delivery functions. We demonstrate that using a closest-fit pattern matching (CPM) noise model can capture complex temporal dynamics which existing approaches do not, increasing packet simulation fidelity by a factor of 2 for good links, a factor of 1.5 for bad links, and a factor of 5 for intermediate links. As our models are derived from real-world traces, they can be generated for many different environments.
- 1 April 2007
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 21-30
Improving Wireless Simulation Through Noise Modeling
We propose modeling environmental noise in order to efficiently and accurately simulate wireless packet delivery. We measure noise traces in many different environments and propose three algorithms to simulate noise from these traces. We evaluate applying these algorithms to signal-to-noise curves in comparison to existing simulation approaches used in EmStar, TOSSIM, and ns2. We measure simulation accuracy using the Kantorovich-Wasserstein distance on conditional packet delivery functions. We demonstrate that using a closest-fit pattern matching (CPM) noise model can capture complex temporal dynamics which existing approaches do not, increasing packet simulation fidelity by a factor of 2 for good links, a factor of 1.5 for bad links, and a factor of 5 for intermediate links. As our models are derived from real-world traces, they can be generated for many different environments.
- 1 April 2007
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 158-165
Interface Contracts for TinyOS
TinyOS applications are built with software components that communicate through narrow interfaces. Since components enable finegrained code reuse, this approach has been successful in creating applications that make very efficient use of the limited code and data memory on sensor network nodes. However, the other important benefit of components-rapid application development through black-box reuse-remains largely unrealized because in many cases interfaces have implied usage constraints that can be the source of frustrating program errors. Developers are commonly forced to read the source code for components, partially defeating the purpose of using components in the first place. Our research helps solve these problems by allowing developers to explicitly specify and enforce component interface contracts. Due to the extensive reuse of the most common interfaces, implementing contracts for a small number of frequently reused interfaces permitted us to extensively check a number of applications. We uncovered some subtle and previously unknown bugs in applications that have been in common use for years.
- 1 April 2007
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 158-165
Interface Contracts for TinyOS
TinyOS applications are built with software components that communicate through narrow interfaces. Since components enable finegrained code reuse, this approach has been successful in creating applications that make very efficient use of the limited code and data memory on sensor network nodes. However, the other important benefit of components-rapid application development through black-box reuse-remains largely unrealized because in many cases interfaces have implied usage constraints that can be the source of frustrating program errors. Developers are commonly forced to read the source code for components, partially defeating the purpose of using components in the first place. Our research helps solve these problems by allowing developers to explicitly specify and enforce component interface contracts. Due to the extensive reuse of the most common interfaces, implementing contracts for a small number of frequently reused interfaces permitted us to extensively check a number of applications. We uncovered some subtle and previously unknown bugs in applications that have been in common use for years.