Network Data and Services Research Department


Formally known as "Database Systems Research Department"

[Lucent Logo]




Richard Hull, Director


----------

The Network Data and Services Research Department is part of the Converged Networks and Services Research Center, which is part of Bell Labs Research. We are located at 600 Mountain Avenue, Murray Hill, in New Jersey, USA. Directions to Bell Labs, Murray Hill are available courtesy of MapsOnUs, a technology that originated in our department.

The department is focused on creating core technologies and software infrastructure to support data access and electronic services (e-services) in wireline and wireless data/telephony networks. Targeted services include those that service providers (SPs) may themselves need for operating their infrastructure (e.g., directory integration); those that SPs may provide in addition to simply conveying information between endpoints (e.g., value-added network services, unified messaging, presence services); and those that SP customers may offer to third parties over an SP's infrastructure (e.g., e-commerce and other Web-based services). Special attention is given to the use of data and creation of information in support of these services, and to preferences and policy management. Targeted SPs include wireless, wireline, data, voice, and converged network carriers, as well as Web hosting and content distribution companies. The infrastructure to support services is a crucial component of Lucent's future, as SP customers demand increasingly rich yet simple-to-use network-enabled capabilities.

Department activity is motivated through a variety of influences, ranging from the academic research community to current and expected Lucent product directions in the areas of mobile internet, multi-media, integration of the telephony networks with the internet, and e-services. Driven by applications, the department draws on a broad range of research specializations, including database systems, XML technology, policy management, automated reasoning, the World Wide Web and Semantic Web, artificial intelligence, distributed systems, operating systems, programming languages, and workflow systems. The department performs fundamental research in both systems and theoretical areas, and develops experimental and practical systems for demonstration, testing, and for eventual application within Lucent products and processes.

Some representative current and past project areas include the following.

Intutitive Network Applications
E-services Personalization and Customization
The department's efforts in E-services personalization and customization involve technology for personalizing and customizing telecomm and web services, including work on managing profile data in the telecom environment, and novel policy management infrastructures for managing end-user preferences. This work builds on earlier work on Vortex, a new rules-based paradigm appropriate for business and customer relationship management (CRM) applications, which is richer than existing "business rules" but more managable than traditional expert systems. Here is a press release about how this work is being incorporated into Lucent's business.
E-services Description and Composition
This effort is currently focused on two aspects of this emerging field. One aspect involves participation in the semantic web community (see here), which is developing a new generation of tools for describing and reasoning about Web content and e-services. Another aspect is developing an infrastructure for assembling and executing collections of Web-based and telephony-based e-services.
Data Grid
Mobile applications are driven by subscriber information: provisioned data (features, billing info, and preferences) and also dynamic data such as location, presence, and account status. While wireless networks have long considered the management of subscriber data to be fundamental, past network architectures have generally dealt with data on ad-hoc basis, with protocol and information models created as needed for particular network needs (e.g., call-processing), and data being made available over wireless- specific signaling interfaces. This forces new applications to either As the quantity and variety of wireless and converged applications grow, there is a pressing need for a more unified approach to subscriber data management. The new approach should allow network operators to create their own views of subscriber data that is managed across multiple network elements, and then allow applications to easily manipulate this data through a single data mediator, using standardized data APIs. DataGrid fills this need. The Data Grid is a data integration solution for consolidating and managing data stored in various network elements. It will enable three key activities: At runtime, the DataGrid can process an incoming query in a standard protocol (initially LDAP, eventually ODBC, XPath, ...) by converting it into one or more queries to the underlying data sources (e.g., network elements, databases) in their respective protocols, reformating the results, and returning the results to the requesting application. Data Grid will also support updates against the unified schema. Using DataGrid can significantly speed up building new applications by making it easier to find the relevant data. For instance, an application like Selective Reach Me which collects user preferences about how to reach users in different circumstances and uses current network status (e.g., in- call status, location) to determine where to route incoming calls, currently needs to poll a variety of data sources (e.g., HLR, 5ESS, MSC, preferences database). With DataGrid, queries can be sent in a single protocol to a single data mediator. Furthermore, caching and materialized views within the DataGrid can be used to speed up the access. In general, most data-rich, next-generation services will benefit from the existence of the DataGrid. DataGrid can also be used as a building block for building larger, cross-carrier data sharing (e.g., GUP data profiles). DataGrid is a very helpful building block for managing subscriber profile data, policies, and preferences.
GUPster
The GUPster project is looking at the problem of user profile management for converged networks. Inspired by the work by 3GPP (telecom standard body), it proposes a framework for the privacy conscious data integration of user profiles where user profiles are represented as XML documents against an agreed upon global schema. At the core of GUPster is XSquirrel, a new XML query language with a sub-document semantics. Using XSquirrel, we can describe queries, mappings (between profile components, i.e. XML sub-documents, and data sources) and access control rules (boolean condition over profile components, i.e. XML sub-documents). Using language rewritings, queries are rewritten against access control rules and transformed into sub-queries that are sent to the corresponding data sources. The sub-document semantics makes it possible to merge the partial answers into the final result. Selected publications: CIDR 2003, Sigmod 2004 (demo), VLDB 2004 (demo).
Access to Web Data
The departments efforts here include work on developing tools for making it easier to analyze and/or access Web-based data. One body of work is centered around the "WebVCR" and its applications. A second area is centered around OfficeTunnel, which involves the use of Web technology and a new approach to cross enterprise firewalls in order to make one's office environment available on essentially any remote device. A third body of work is focused on supporting new capabilities on the Web such as collaborative browsing tools.
Map routing
The Maps-On-Us project developed a system for computing routes from point A to point B, including the generation of maps and turn-by-turn directions. This system was supported by Bell Labs for 2 years, and in 1998 sold to Switchboard (which provides mapping for the US Postal Service web site, among others).
Lightweight Trigger Access Process (LTAP)
LDAP directories are passive repositories that do not provide active monitoring and alerting facilities, i.e., triggers. LTAP adds active facilities in a portable manner to LDAP servers. Active facilities are necessary for LDAP servers to monitor accesses to directories and then take appropriate actions. Such functionality is crucial for making Directory Enabled Networks real.
XML
XML is key in supporting services and will be used in building and operating many Lucent products in the near future. The department is very active in the development of the next generation of XML infrastructure. This new infrastructure will provide XML users with new tools, easier development capabilities, increased performance and reliability. The activity is strategically centered around newly emerging W3C standards: XML Schema 1.0, XPath 2.0, XQuery 1.0, and SOAP that will soon be used to replace the old generation of XML tools. Research work includes XML Schema modeling, query languages (Galax), type systems for XML, distributed XML processing, XML storage (LegoDB), and constraints for XML data.

Some of the department research projects have been and are conducted jointly with members in other parts of Bell Labs, and several projects involve active collaboration with Lucent product units. And there are collaborations with outside institutions, including AT&T Labs, Avaya Labs, FORTH (Crete), the Indian Institute of Science, the Indian Institute of Technology, the University of Pennsylvania, the University of Toronto, the University of California, and the University of Manchester.

Department members:

Recent visitors and department affiliates include:

Department alumni include:

Administrative Support

Internal Home Page

Lucent employees will find more information in the internal department home page.

----------

Questions or Comments
Last modified: 26 October 2004.

Copyright © Lucent Technologies. All rights reserved.