John V. Guttag

Cutting the electric bill for internet-scale systems

Energy expenses are becoming an increasingly important fraction of data center operating costs. At the same time, the energy expense per unit of computation can vary significantly between two different locations. In this paper, we characterize the variation due to fluctuating electricity prices and argue that existing distributed systems should be able to exploit this variation for significant economic gains. Electricity prices exhibit both temporal and geographic variation, due to regional demand differences, transmission inefficiencies, and generation diversity. Starting with historical electricity prices, for twenty nine locations in the US, and network traffic data collected on Akamais CDN, we use simulation to quantify the possible economic gains for a realistic workload. Our results imply that existing systems may be able to save millions of dollars a year in electricity costs, by being cognizant of locational computation cost differences.

ANTS: a toolkit for building and dynamically deploying network protocols

The authors present a novel approach to building and deploying network protocols. The approach is based on mobile code, demand loading, and caching techniques. The architecture of the system allows new protocols to be dynamically deployed at both routers and end systems, without the need for coordination and without unwanted interaction between co-existing protocols. They describe the architecture and its realization in a prototype implementation. To demonstrate how to exploit the architecture, they present two simple protocols that operate within the prototype to introduce multicast and mobility services into a network that initially lacks them.

The algebraic specification of abstract data types

SummaryThere have been many recent proposals for embedding abstract data types in programming languages. In order to reason about programs using abstract data types, it is desirable to specify their properties at an abstract level, independent of any particular implementation. This paper presents an algebraic technique for such specifications, develops some of the formal properties of the technique, and shows that these provide useful guidelines for the construction of adequate specifications.

Eulerian video magnification for revealing subtle changes in the world

Our goal is to reveal temporal variations in videos that are difficult or impossible to see with the naked eye and display them in an indicative manner. Our method, which we call Eulerian Video Magnification, takes a standard video sequence as input, and applies spatial decomposition, followed by temporal filtering to the frames. The resulting signal is then amplified to reveal hidden information. Using our method, we are able to visualize the flow of blood as it fills the face and also to amplify and reveal small motions. Our technique can run in real time to show phenomena occurring at the temporal frequencies selected by the user.

Abstract data types and software validation

A data abstraction can be naturally specified using algebraic axioms. The virtue of these axioms is that they permit a representation-independent formal specification of a data type. An example is given which shows how to employ algebraic axioms at successive levels of implementation. The major thrust of the paper is twofold. First, it is shown how the use of algebraic axiomatizations can simplify the process of proving the correctness of an implementation of an abstract data type. Second, semi-automatic tools are described which can be used both to automate such proofs of correctness and to derive an immediate implementation from the axioms. This implementation allows for limited testing of programs at design time, before a conventional implementation is accomplished.

The Larch Family of Specification Languages

Larch specifications are two-tiered. Each one has a component written in an algebraic language and another tailored to a programming language.

LCLint: a tool for using specifications to check code

This paper describes LCLint, an efficient and flexible tool that accepts as input programs (written in ANSI C) and various levels of formal specification. Using this information, LCLint reports inconsistencies between a program and its specification. We also describe our experience using LCLint to help understand, document, and re-engineer legacy code.

Patient-specific seizure onset detection

This work presents an automated, patient-specific method for the detection of epileptic seizure onsets from noninvasive EEG. We adopt a patient-specific approach to exploit the consistency of an individual patients seizure and non-seizure EEG. Our method uses a wavelet decomposition to construct a feature vector that captures the morphology and spatial distribution of an EEG epoch, and then determines whether that vector is representative of a patients seizure or non-seizure EEG using the support-vector machine classification algorithm. Our completely automated method was tested on non-invasive EEG from thirty-six pediatric subjects suffering from a variety of seizure types. It detected 131 of 139 seizure events within 8.0/spl plusmn/3.2 seconds following electrographic onset, and declared 15 false-detections in 60 hours of clinical EEG. Our patient-specific method can be used to initiate delay-sensitive clinical procedures following seizure onset; for example, the injection of an imaging radiopharmaceutical or stimulation of the vagus nerve.

Detecting Pulse from Head Motions in Video

We extract heart rate and beat lengths from videos by measuring subtle head motion caused by the Newtonian reaction to the influx of blood at each beat. Our method tracks features on the head and performs principal component analysis (PCA) to decompose their trajectories into a set of component motions. It then chooses the component that best corresponds to heartbeats based on its temporal frequency spectrum. Finally, we analyze the motion projected to this component and identify peaks of the trajectories, which correspond to heartbeats. When evaluated on 18 subjects, our approach reported heart rates nearly identical to an electrocardiogram device. Additionally we were able to capture clinically relevant information about heart rate variability.

Introducing new Internet services: why and how

Active networks permit applications to inject programs into the nodes of local and, more important, wide area networks. This supports faster service innovation by making it easier to deploy new network services. In this article, we discuss both the potential impact of active network services on applications and how such services can be built and deployed. We explore the impact by suggesting sample uses and arguing how such uses would improve application performance. We explore the design of active networks by presenting a novel architecture, ANTS (active network transport system), that adds extensibility at the network layer and allows for incremental deployment of active nodes within the Internet. In doing so, ANTS tackles the challenges of ensuring that the flexibility offered by active networks does not adversely impact performance or security. Finally, we demonstrate how a new network service may be expressed in ANTS.