Scott Jordan, Ph.D. - US grants

9412011 Rutledge This Equipment Grant funds the purchase of computer equipment, audio interface components and support software which is dedicated to support research in engineering. The equipment will be used for several projects, including speech processing to compensate for sensorineural hearing losses, speech recognition over the telephone, enhancement of alaryngeal speech, and audio processing based on psychoacoustics. Acquisition of this equipment will permit processing of higher quality speech and audio signals, and implementation of algorithms having much greater complexity with more efficiency. ***

This project is concerned with the allocation of network resources, specifically buffer and bandwidth, to users and to virtual paths in an Asynchronous Transfer Mode network. The research concentrates on developing a framework for analysis of this resource allocation probelm. Users are assumed to receive benefit as a function of their quality of service. measured by cell loss probability and completion time for real-time services. Network usage is measured using effective bandwidth and buffer occupancy. The paradigm of price and demand is used to represent the minimal required information transfer between user and network. These prices are considered purely as internal network indicators for the purpose of resource allocation; what users are actually charged by their service provider is a separate issue not considered here. The goal is to allocate bandwidth and buffer so as to make as many users happy as possible, and to do so efficiently. Furthermore, it is desirable that any such mechanism be distributed both hierarchically and geographically. The P.I. proposes to develop this framework to jointly allocate buffer and bandwidth, to study the use of priorities to replace explicit reservation of resources, and to analyze the resulting network dynamics.

This research is concerned with techniques for supporting integrated services over a wireless cellular network. A fundamental problem is how to allocate radio resources to different users to optimize an overall network objective criterion while satisfying Quality of Service constraints. Wireless networks based on Direct-Sequence (DS)-Code-Division Multiple Access (CDMA) networks are considered, since DS-CDMA is a strong contender for next generation cellular systems. Radio ``resources'' in a DS-CDMA network include transmitted powers, processing gains, and number of assigned codes. The focus of this research is on Dynamic Resource Allocation (DRA) in the presence of time-varying user requests. The goals are to (1) Develop techniques for DRA given multiple traffic flows with different Quality of Service (QoS) requirements, and (2) Characterize performance/complexity tradeoffs associated with
these techniques given simple traffic and propagation models.

The investigators consider a series of problems and approaches for DRA in a CDMA network based on progressive refinements of the system model. First, they consider a single class of data traffic flows within an isolated cell. Reliability is assumed to be guaranteed through retransmissions, and DRA is studied
in the presence of time-selective fading. Two traffic flows, representative of voice and data, are then introduced (in the absence of fading), and the relationship between reassignment of resources and admission control is studied. DRA in a multi-cell system is also considered taking into account measurement-based power constraints which limit interference to other cells. Finally, pricing is investigated as a method for DRA which accounts for user utilities. Methods for dynamic pricing are studied in the context of the models previously described.

The Principle Investigators of this proposal propose to organize a PI meeting for the Networking Research Programs (Networking Research Program and Special Projects in Networking) in the ANIR Division of NSF. The PI meeting will be held over a two-day period in addition to an opening night reception. The reception will be Wednesday evening, November 1, 2000 with two full days of technical sessions, Thursday, November 2, and Friday, November 3. The PI meeting will facilitate the exchange of research ideas among the PIs with active NSF awards from the Networking Research Programs and NSF program directors.

The goal of this project is the design of joint power control, rate control, and base station assignment for multimedia wireless networks. Two key aspects of the multimedia network will be considered that have been given short shrift in the past: the requirements of data and system dynamics. The project will consider 4th generation Direct-Sequence Code-Division Multiple-Access (DS-CDMA) networks. The goal is to implement dynamic assignment of downlink transmit power, data rate, and codes to each mobile in accordance with the ability of each application to use the resulting QoS. The focus is on network-wide behavior, and therefore network-wide performance measures will be considered, jointly optimized between physical and networking layers. Three main thrusts will be considered:
(a) Use of multiple state automata hybrid systems models as a novel tool to control wireless network dynamics.
(b) Design and analysis of resource allocation algorithms using novel pricing theory.
(c) Design and analysis of parameter estimation algorithms to achieve robust network behavior.
The first thrust is intended to provide a powerful tool for generating control algorithms for systems with both discrete and continuous parameters and switching costs. It will be used in this project for controlling power, data rate, and handoffs. The research in this area will extend the dimensionality of a previously studied hybrid systems approach to handoff control. The second research thrust is the construction of joint power control, data rate control, and base station assignment algorithms. This research will be based on economic theory for resource allocation, and implemented using hybrid systems. The final thrust will address the reliance of many resource allocation schemes upon unknown communication parameters such as received SINR. Three investigators will work together: an expert in hybrid systems, an expert in resource allocation, and an expert in multiuser spread spectrum wireless systems.

Since the fundamental Internet protocols were designed, the Internet has changed from a cooperative network to one in which carriers and application providers both cooperate and compete. Internet design, however, has not evolved to gracefully recognize or support this competition. Support for both cooperation and competition requires coordinated evolution of Internet architecture and communications law. Communications law that is not based on Internet architecture can not effectively regulate the converging communications marketplace. Similarly, Internet architecture that does not recognize the effect of communications law upon the marketplace can not serve as a catalyst for innovation and new applications.

Broader Impact: The focus of this project is on the interplay between Internet architecture and communications law around a recent issue called Net Neutrality. Net neutrality represents the idea that Internet users are entitled to service that does not discriminate on the basis of source, destination, or ownership of Internet traffic. Proponents of net neutrality argue that without a prohibition on discrimination, Internet Service Providers may charge application providers discriminatory prices for access to dedicated bandwidth or for quality of service, or may outright block access to certain applications or websites, and that such activity will inhibit development of new Internet applications. Opponents of net neutrality argue that there is no current problem, that competition is sufficient to ensure that commercially negotiated arrangements for bandwidth or QoS will not negatively impact consumers, and that any regulation will discourage investment in network infrastructure. Congressional action, or inaction, on net neutrality will greatly impact Internet architecture. If Congress adopts one of the more extreme versions of net neutrality, then the use of QoS could be prohibited. If Congress adopts a more moderate version of net neutrality, then QoS might be allowed but charging for QoS might be prohibited. If Congress does not act on net neutrality, then ISPs may block access to QoS mechanisms to competing application providers. Any of these actions would substantially affect the evolution of the Internet and of future Internet research.

Intellectual Merit: In this project, the issue of net neutrality will be used to examine the interplay between Internet architecture and communications law, by considering the implications of Internet architecture upon net neutrality law and by considering the implications of net neutrality law upon Internet architecture. First, a network model and terminology will be created that can be interpreted in both Internet architecture and communications law. Internet infrastructure services will be defined as functionality that must be provided within the access network, whereas Internet application services will be defined as functionality that can be provided elsewhere in the Internet. Second, to consider the implications of Internet architecture upon net neutrality law, a net neutrality law will be formulated based on a careful delineation of Internet infrastructure and Internet applications. The approach will essentially mandate the use of open interfaces between infrastructure and applications, but will not mandate full open access, which would require open interfaces at each protocol layer. Third, to consider the implications of net neutrality law upon Internet architecture, architectural requirements required to support net neutrality through an open interface will be identified. Internet architecture requirements required to gracefully support both cooperation and competition will be studied. New design principles that the FIND community should attempt to incorporate into future Internet architecture will be formulated. It is difficult to formulate

The goal of this project is to incorporate key elements of telecommunication policy and economics into Internet architecture. As a result of technical, economic and public policy forces, the Internet's original design principles -- layering and end-to-end -- are increasingly violated. Internet Service Providers are deploying quality of service mechanisms, but only allowing their use for certain applications sold to their own subscribers. Some Internet Service Providers have used deep packet inspection techniques to implement traffic management practices that throttle or block peer-to-peer applications.

To counteract this deterioration, this project will propose an interdisciplinary approach to update the Internet architectural principles to account for telecommunications policy and economics. The project will identify the flaws of the end-to-end and layering models that are not withstanding the technical, economic, and legal forces upon networking; modify these models so that they promote good technical design, respond appropriately to economic pressures, and encourage beneficial outcomes that benefit society; and validate these new models and illustrate their potential use by applying them to three case studies -- net neutrality, traffic management, and Quality of Service.

This research will have a broad impact. The P.I. is developing an undergraduate course on ?The Internet and Public Policy?. This research will help bridge the gulf that exists between communication lawmakers and networking researchers by informing staff members in the United States Congress about the technical aspects of telecommunication issues, and by developing an architectural framework for the networking research community to help them consider impacts of network economics and law.

Networking research on Internet Quality of Service techniques has resulted in standards that have been widely incorporated into routers. And yet, use of Quality of Service practices is very limited. Internet Service Providers have implemented Quality of Service practices to support their own voice and video services, but do not offer Quality of Service techniques to their subscribers for use by any other Internet applications. Internet Service Providers have not incorporated Quality of Service practices into their agreements with other Internet Service Providers, and hence Quality of Service techniques are not available across the Internet. The limited deployment of Quality of Service practices has ignited vigorous debate over Net Neutrality, one of the most contentious telecommunications public policy issues in decades.


This project will address the lack of availability of Quality of Service techniques widely across the Internet. It will adopt an interdisciplinary approach that integrates network architecture, economics, and law. The project will have three goals. First, it will create and analyze network architecture and economic models that can give insight into when an Internet Service Provider will choose to offer Quality of Service techniques to application providers other than itself. Second, it will consider how communications law may affect the development of Quality of Service practices. Third, it will illustrate potential technical and economic arrangements between users, Internet Service Providers, and application-providers that could enable widespread deployment of Quality of Service practices in a manner consistent with new laws regulating traffic management.


It is hoped that the results will encourage the widespread deployment of Quality of Service practices and thus enable further development of novel Internet applications. The research will also encourage interdisciplinary teaching of networking technology, economics, and law through further development of a novel undergraduate course on The Internet and Public Policy. The research, if successful, will inform ISPs and regulators, and potentially improve both the functional character of the Internet and the revenue opportunities of the ISPs.

Interconnection agreements between Internet Service Providers (ISPs), transit providers, and content delivery networks (CDNs) allow broadband users to transmit traffic to, and receive traffic from, everywhere on the Internet. These agreements are based on a combination of technical issues (e.g. the locations to and from which the provider agrees to route traffic and the amount of traffic exchanged in each direction) and economic issues (e.g. the perceived value of the arrangement and any fees involved). Historically, interconnection agreements consisted of settlement-free peering agreements (between network providers at the same tier) or paid transit agreements (between network providers of different tiers). However, changes in Internet topology resulting from mergers between access and backbone networks and the proliferation of CDNs have resulted in most Internet traffic now flowing through interconnection points between access networks and CDNs. These traffic flows are often subject to a new category of paid peering agreements.

In the past five years, disputes over interconnection agreements have occasionally led to failures to augment interconnection capacity, which in turn have harmed the experience of consumers. This project will construct a modern model of Internet interconnection that incorporates both the relevant technical and economic factors. The model will be used to explain and guide new forms of Internet interconnection arrangements. The project will also develop methods to illustrate when Internet interconnection agreements may be used in anticompetitive and discriminatory manners.

This project consists of two interrelated efforts. In the development of models of Internet interconnection, the project will investigate the cost factors that should affect interconnection arrangements, including the destinations to which a network operator will route traffic, the volume of traffic, the network cost along that path, and any transit payments required. The project will also investigate the value factors that may affect interconnection arrangements, including the destinations to which an interconnection partner will route traffic, the volume of traffic, and the value of that traffic based on the application. The project will then analyze the variation of the feasible ranges of prices for paid peering with network costs, the maximum that a party is willing to pay, routing and traffic ratios, and competitive pressures.

Second, in the development of analytical models of competitive versus anticompetitive behavior, the project will evaluate the characteristics of a variety of interconnection agreements, including the minimum number of interconnection points, routing requirements, capacity augmentation requirements, determinants of payment, and conditions for termination. The project will examine the impact of interconnection prices on broadband providers, content providers, and Internet users, and the resulting impact upon engineering decisions.

The project is expected to contribute substantially to the development of network modeling and analysis techniques that integrate technical and economic factors. The expected outcomes of the project will lend insight into commercial interconnection arrangements, as well as into related public policy issues when disputes harm the experience of consumers. The project will also contribute to informing policymakers and their staff about the technical aspects of telecommunication issues.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

The Great Lakes provide the largest freshwater resource in the world, and Lake Superior is at the top of this critical watershed. Recent weather fluctuations, increases in surface and lake temperatures, ice cover and lake levels have had a tremendous impact on infrastructure in the Great Lakes region. Collaboration and interdisciplinary science is necessary to understand the dynamic relationships between ecosystem health, commerce and development. The goal of this project is to conduct a conference series that will engage (1) interdisciplinary experts in coastal geography, geohazards, water quality, climate science, livelihoods, tourism, environmental history, sustainable business and adaptation planning; (2) practitioners, including local and regional planning staff, public sector stakeholders, and local and regional non-governmental organizations; and (3) industry leaders in tourism, mining, education, and real estate development.

The expected outcomes for this series include: 1) integrated stakeholder input to create multiple scenarios for sustainable coastal community development; 2) visualizations of those scenarios created by landscape architects; and 3) new ways to achieve sustainable development along the south shore of Lake Superior.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.