AT&T Labs

About: AT&T Labs is a based out in . It is known for research contribution in the topics: Network packet & The Internet. The organization has 1879 authors who have published 5595 publications receiving 483151 citations.

Threading electronic mail: a preliminary study

Abstract: Tools for processing email and other electronic messages should be able to recognize and manipulate threads, that is, conversations among two or more people carried out by exchange of messages. While user clients typically insert in messages structural information useful for recovering threads, inconsistencies between clients, loose standards, creative user behavior, and the subjective nature of conversation make threading systems based on structural information only partially successful. We propose that this situation is unlikely to change, and that threading of electronic messages be treated as a language processing task. Preliminary experiments show that a significant level of threading effectiveness can be achieved by applying standard text matching methods from information retrieval to the textual portions of messages.

Ricean $K$ -Factors in Narrow-Band Fixed Wireless Channels: Theory, Experiments, and Statistical Models

Abstract: Fixed wireless channels in suburban macrocells are subject to fading due to scattering by moving objects such as windblown trees and foliage in the environment. When, as is often the case, the fading follows a Ricean distribution, the first-order statistics of fading are completely described by the corresponding average path gain and Ricean K-factor. Because such fading has important implications for the design of both narrow-band and wideband multipoint communication systems that are deployed in such environments, it must be well characterized. We conducted a set of 1.9-GHz experiments in suburban macrocell environments to generate a collective database from which we could construct a simple model for the probability distribution of K as experienced by fixed wireless users. Specifically, we find K to be lognormal, with the median being a simple function of season, antenna height, antenna beamwidth, and distance and with a standard deviation of 8 dB. We also present plausible physical arguments to explain these observations, elaborate on the variability of K with time, frequency, and location, and show the strong influence of wind conditions on K.

Kinetic: verifiable dynamic network control

Abstract: Network conditions are dynamic; unfortunately, current approaches to configuring networks. Network operators need tools to express how a network's data-plane behavior should respond to a wide range of events and changing conditions, ranging from unexpected failures to shifting traffic patterns to planned maintenance. Yet, to update the network configuration today, operators typically rely on a combination of manual intervention and ad hoc scripts. In this paper, we present Kinetic, a domain specific language and network control system that enables operators to control their networks dynamically in a concise, intuitive way. Kinetic also automatically verifies the correctness of these control programs with respect to user-specified temporal properties. Our user study of Kinetic with several hundred network operators demonstrates that Kinetic is intuitive and usable, and our performance evaluation shows that realistic Kinetic programs scale well with the number of policies and the size of the network.

Distributed versus centralized antenna arrays in broadband wireless networks

Abstract: This paper studies the potential benefits of antenna array processing-used in conjunction with adaptive data-rate control-in broadband wireless networks. We focus on distributed antenna arrays, i.e., combining signals from a group of microcells, rather than the more conventional centralized (macrocellular) antenna, array processing. We show that distributed arrays promise significant power and capacity gains over centralized arrays. Moreover, we show that even selection combining (though less effective than coherent combining) can be very successful in this architecture, offering a promising tradeoff between performance and complexity.

The AT&T spoken language understanding system

Abstract: Spoken language understanding (SLU) aims at extracting meaning from natural language speech. Over the past decade, a variety of practical goal-oriented spoken dialog systems have been built for limited domains. SLU in these systems ranges from understanding predetermined phrases through fixed grammars, extracting some predefined named entities, extracting users' intents for call classification, to combinations of users' intents and named entities. In this paper, we present the SLU system of VoiceTone/spl reg/ (a service provided by ATT 2) extending statistical classifiers to seamlessly integrate hand crafted classification rules with the rules learned from data; and 3) developing an active learning framework to minimize the human labeling effort for quickly building the classifier models and adapting them to changes. We present an evaluation of this system using two deployed applications of VoiceTone/spl reg/.