Projects and Research

Most of the following projects have been directed and supervised by me, while I have participated in some of them of a technical and discussion level as well.


1. Charging and Accounting

1.1 Pre-study

The TIK has written a pre-study on "Customer Care, Charging, Accounting, Billing, and Pricing" for the Swiss National Science Foundation in the Information and Communication Program (SPP-SNF). The final version may be obtained in a full version (pdf, 226 kByte) or a short version (gzipped PostScript, 25 kByte) including the summary and conclusions. This pre-study determines the state of the art as of February 1998 in the areas of customer care, charging, accounting, billing, and pricing of communication services. Particular interest has been addressed to Internet-based communication services, however, basics of ATM work is included as well. A quite comprehensive overview on projects, products, and tasks are included in the full version.

This project was performed between November 1997 and February 1998.

1.2 COINS - Charging and Accounting for Integrated Internet Services

An early TIK project called COINS concerning Charging and Acounting for Integrated Internet Services has been carried out. A basic overview of this work may be obtained in gzipped PostScript right here (72 kByte) concentrating on "Arrow - A Flexible Architecture for an Accounting and Charging Infrastructure in the Next Generation Internet".

Additional work has been performed to obtain a clear view on overhead involved in reservation-based charging and accounting mechanisms for integrated services networks. Therefore, SSP (State Setup Protocol, a first version of the Resource Reservation Protocol RSVP) has been extended by charging and accounting information, implemented on a BSD-Unix kernel system, and evaluated in a given test-bed based on Crossbow. Details are included in the paper entitled "Reservation-based Charging in an Integrated Services Network" which is available here (89 kByte, gzipped PostScript).

As an intermediate overview of the state of the art, problems, and trends in charging and accounting for integrated Internet services the TIK reported their observations to the INET'98. This paper may be obtained in gzipped PostScript right here (51 kByte).

This project was performed between February 1997 and July 1998.

1.3 Billing Systems for the Internet and Some Dedicated Services

An overview of billing systems for Internet service providers, mainly intended for the use of H.323-based IP telephony services, may be obtained on this separate page here. It shows a variety of available billing solutions for IP telephony, however, it identifies as well that general Internet service solutions are missing.

This project was performed in July 1999.

1.4 CATI - Charging and Accounting Technology for the Internet

The main goals of this SNF project cover the following ones for the work at TIK:

There is the official CATI project page (http://www.tik.ee.ethz.ch/~cati) maintained at TIK.

This project runs since July 1998 and will be finished by March 2000.

1.5 M3I - Market Managed Multi-service Internet

The project M3I aims to design, implement and trial a next-generation system that will enable Internet resource management through market forces, specifically by enabling differential charging for multiple levels of service. End-users and Internet Service Providers will have measurable improvements in terms of the ability to charge differentially for applications requiring differing QoS levels or multicast, they will have a more effective competition in a differentiated services market, and they will see real-time feedback and validation of charges.

There is the official M3I project page (http://www.m3i.org) maintained at Hewlett Packard European Labs as well as the local TIK M3I project page (http://www.tik.ee.ethzc.ch/~m3i).

This project has been started January 2000 and will finish by December 2001.

1.6 MobyDick - Mobility and Differentiated Services in a Future IP Network

The MobyDick project focusses at the development of a seamless access to existing and emerging IP-based services and applications by proposing an architecture for wireless Internet access, hand-overs, QoS support, authentication, authorization, accounting, auditing, and charging. This will enable new business opportunities for operartors, service providers, and content providers in the area of wireless access and backbone technology market.

There is the official MobyDick project page (http://www.ist-mobydick.org) maintained at T-Nova, Berlin as well as the local TIK MobyDick project page (http://www.tik.ee.ethzc.ch/~mobydick).

This project has been started January 2001 and will finish by December 2003.

1.7 ANAISOFT - Advanced Network and Agent Infrastructure for the Support of Federations of Workflow Trading Systems

The ANAISOFT project proposes to study the creation of secure and responsive Federations of Workflow Trading Systems (FWTS) using intelligent and mobile agent technology. In workflow trading systems, bidding protocols, structure of catalogues and ways to describe activities can vary from case to case, depending on the application domains. A high degree of flexibility is needed to cope with the specifics of each marketplace. Appropriate charging and accounting schemes are inherently connected with QoS-supporting communication environments of FWTS, since the selection of suitable service quality determines the communication performance of federations. It remains to find out how QoS can be provided appropriately in a dynamic and mobile environment. The investigation of charging functionality in Differentiated Services environment is under its way. Aggregated views of end-to-end flows will change paradigms applied for charging in the flow-based (or Integrated Services) access environment of the Internet.

There is the official ANISOFT project page (http://anaisoft.unige.ch) maintained at University of Geneva, Switzerland as well as the local TIK ANAISOFT project page (http://www.tik.ee.ethzc.ch/~anaisoft).

This project has been started January 2000 and will finish by December 2001.


2. AX -- Authentication, Authorization, Accounting, Auditing, and Charging in the Internet

The traditional AAA Architecture (Authentication, Authorization, and Accounting) of the lacks a degree of generality due to the fact that charging tasks are not integrated, QoS support is lacking, mobility support is provided at the edges, and policy enforcement as well as policy decision points are not well separated. Therefore, a new generic architecture is developed enabling an approach to overcome the deficits mentioned above.

This work is part of the MobyDick project and utilizes local ETH resources in addition to those EU resources. A first technical report on the basic ideas on Ax may be obtained as a TIK Report (May 2001) PDF-format (277 kByte).

For further details cf. Section 1.6 above on the MobyDick project and there is the official MobyDick project page (http://www.ist-mobydick.org) maintained at T-Nova, Berlin as well as the local TIK MobyDick project page (http://www.tik.ee.ethzc.ch/~mobydick).


3. Service Deployment

Since services are key differentiators for Internet Service Providers (ISP) in the near future, and already today, an interoperable and fast creation mechanism for Internet services on a heterogeneous network infrastructure is essential. Therefore, a novel mechanism for a scalable deployment of services over programmable networks has been developed within a colaboration project between the IBM Zürich Research Laboratory and the ETH Zürich, TIK within a supervised external Ph.D. thesis performed by Robert Haas. The work is ongoing and the elaboration of this mechanism as well as its evaluation just started.

Further details are available in the following paper from ICC'01 in Helsinki, Finland, June 2001: PDF-format (298 kByte) as well as the complete TIK Report (December 2000) PDF-format (289 kByte).


4. Adaptive End-to-end QoS in IP Networks

The emerging need for Quality-of-Service (QoS) for applications in and end-to-end fashion is the driving force for mechanisms in support of QoS in the Internet. However, the lack of end-to-end service guarantees, a difficult configuration of end-systems, and routers, as well as the lack of suitable specification models for QoS between applications and the network make clear that the Internet as itstands today is not really suitable for these emerging applications. Therefore, the work undertaken in an externally supervised Ph.D. project performed by Roman Pletka from IBM Zürich Research Laboratory and the ETH Zürich, TIK targets at the integration of lower and upper protocol layers and mechanisms to enable bandwidth and delay guarantees. This will be complemented by adaptation schemes which will enable communication devices to adapt automatically to changing networking conditions. The implementation environment is based on a programmable network, where "monitor packets" travel in the network to detect bottlenecks.

Further details on a buffer management scheme for bandwidth and delay variations using a virtual scheduler are available in the following paper from ICN'01 in Colmar, France, June 2001: PDF-format (505 kByte).


5. ATM over Shared Media Networks Communication Middleware

Providing statistical QoS guarantees on shared media networks, e.g., Ethernet, can be obtained by offering ATM Adaptation Layer Type 5 (AAL 5) services to end-systems which are connected to a local Ethernet. In this approach called ATM over Shared Media Networks - ASHMEN AAL 5 Protocol Data Units are transported in Ethernet Frames and transparently mapped onto ATM cells in a so-called Attachment Device which interconnects the local Ethernet with a Backbone or public ATM network.

Further details are available in the following paper: PDF-format (111 kByte). Another short paper is available on the resource management scheme for ASHMEN as well as some initial results on achieved delays and delay distributions in an investigated local area environment PDF-format (208 kByte).

This project runs since October 1997 and will be finished by mid 2000.


6. Communication Middleware

6.1 Da CaPo++ (Dynamic Configuration of Protocols)

A final description of the KWF/KIT-Project No. 2984.1 between ETH Zürich, SBV, and XMIT on Da CaPo++ - A Framework for Dynamic Configuration of Communication Systems. can be obtained from the paper in PDF-format (244 kByte).

This project has been finished by September 1997, while performance results have been obtained until December 1997.

6.2 Middleware Research

The investigation of currently available communication middleware and its possibilities to support Quality-of-Service features as well as charging support will be started.

Please check the section on Charging for further details on charging issues.

This project started 2000.

6.3 MCF - Multicast Communication Framework

The design and evaluation of a communication framework for multicasting and an integrated and guaranteed support of Quality-of-Service has been performed in a Ph.D. project. An overview in a paper may be obtained in PDF-format (87 kByte).

This project has been finished by 1998.


7. Tele-teaching

7.1 Telepoly

A project concerning tele-teaching issues called Telepoly (in Dr. Thomas Walter's responsibility) has been carried out at TIK. A brief overview in a short paper on the technical set up and investigations and experiences may be obtained in gzipped PostScript right here (75 kByte).

A set of practical experiences as well as didactic experiences have been collected and investigated, which I have supported. After this period of time it has been moved into productive service at the ETH and is supervised by the Didaktikzentrum of the ETH and it is regularly used for lectures as well as talks.

This project was performed between 1995 and 1999.

7.2 ET&L Easy Teach and Learn

The successor project Easy Teach & Learn® (in Dr. Thomas Walter's responsibility as well) is a project that extends the capabilities of traditional teleteaching scenarios where synchronous and interactive distance education is done between classrooms, into a virtual classroom scenario where people can access courses from everywhere, e.g., from office or from home.


8. Formal Evaluations of Multimedia Synchronization Algorithms

This project tackles the missing schemes for comparing analytically synchronization algorithms for multimedia applications. QoSy - Quality-of-Synchronization has been developed at TIK in a Ph.D. project. An overview in a paper may be obtained in gzipped PostScript right here (143 kByte).

This project was performed between July 1997 and September 1999.