The Telecommunications Software Engineering Research Group (TSERG) consists of Professors G. v. Bochmann, L. Logrippo, R.L. Probert (Coordinator), H. Ural of the School of Information Technology and Engineering (SITE) at the University of Ottawa, and Professor J. P. Corriveau of the School of Computer Science at Carleton University. Collaborators include Professors T. Lethbridge, J. Raymond, G. White of SITE at the University of Ottawa, and Professors R. Buhr and M. Woodside of the Department of Systems and Computer Engineering at Carleton University, as well as various individuals from local industry. Approximately six full-time Ph.D. and twenty full-time Master students are being supervised by TSERG professors. There are three full-time and two part-time staff members in the group.
The research interests of the group focus on the specification, design, testing, and verification of communications protocols and software, and incorporate many other mainstream research topics in Computer Science. These include the design and development of CASE-based methodologies, especially object-oriented design of real-time distributed systems, design of open software architectures including Open Distributed Processing (ODP), as well as theory and practice of formal methods for software design and testing. Many of these methods are related to well-known distributed software engineering models, such as Communicating Extended Finite State Machines, Process Algebras, and Message Sequence Charts. TSERG has received funding support for collaborative industry-university software research involving federal and provincial research agencies such as the Natural Sciences and Engineering Research Council (NSERC) of Canada, the National Research Council (NRC), and the Telecommunications Research Institute of Ontario (TRIO), and some of the corporations in "Silicon Valley North" (Ottawa, Kanata, Nepean) such as Nortel (formerly Bell-Northern Research), ObjecTime, Mitel Electronics, Newbridge Networks, Stentor (formerly Bell Canada), IBM, and also other large corporations such as Motorola.
The mandate of the group is to provide leadership in telecommunications software engineering research and development. To this end, TSERG members participate in conceiving and developing international standards for telecommunications systems architecture, protocol specification and protocol testing. TSERG members serve as Canadian representatives at standards meetings, and advise and assist Canadian industry and government in communications software R&D. As well, TSERG members are regularly chosen for the program committees of many of the most significant international conferences in the areas of Telecommunications Software Engineering and Protocol Engineering. These include IWTCS (International Workshop on Testing of Communicating Systems), ICNP (International Conference on Network Protocols), FORTE/PSTV (International Conference on Formal Techniques and on Protocol Specification, Testing, and Verification), AMAST (Algebraic Methods Applied to Software Technology), and ICODP (International Conference on Open Distributed Processing). Finally, TSERG professors and students have an outstanding record of research collaboration and technology transfer with many of the leading industries in Telecommunications Software Engineering and Protocol Engineering. This surfeit of opportunities to work directly with the best industrial software engineers on targeted, realistic, industrial-scale research problems within the university setting may be one reason why many post-graduate students are attracted to TSERG.
Technology transfer (we prefer the term "Technology Sharing", since industry input is required for successful interchange of techniques, problems and useful ideas) takes place in TSERG via many paradigms. TSERG professors are very active in providing graduate-level courses on site to industrial computer and communications software engineers. For example, in the new ConGESE (Consortium for Graduate Education in Software Engineering, described elsewhere) Master's program in Software Engineering, important links are forged between TSERG professors and future industrial research collaborators.
In the area of technology personnel transfer, TSERG has had great success in placing its graduate students and employees in excellent positions in industry. It is common for students in one of our many joint university industry projects to obtain an insider's view of the collaborating corporation, and thus be well informed to make career decisions.
TSERG is also a model of inter-university research collaboration, particularly in targeted research where industry-supported research contracts supported the following collaborations across universities:
i) L. Logrippo (UO) and R. Buhr (Carleton University) in "Lotos Formal Semantics for Use-Case Maps", a project in which a highly effective design model (use-case maps) is given a formal algebraic semantics (LOTOS). As a result, LOTOS prototyping and validation tools can be used to check the correctness and consistency of designs before they are coded.
ii) R. Probert (UO), H. Ural (UO), J.-P. Corriveau (Carleton University) and T. Shepard (RMC, Queen's) in "Formal Testing Framework for Object-Oriented Design of Real-Time Systems". In this project, the ObjecTime industrial-strength design environment which supports a real-time, object-oriented design methodology is augmented by cost-effective testing components based on sound principles for testing distributed systems, including the use of the international standard test framework and language (TTCN - the Tree and Tabular Combined Notation) where appropriate (practical and cost-effective). As a result, ObjecTime acquires world-leading testing capability and TSERG professors and students see their work achieve global distribution imbedded in an industrial-strength toolkit and applied to real industrial problems.
iii) L. Logrippo (UO), G.v. Bochmann (U. of Montral), G.Sifakis (VERIMAG, Grenoble), and A. Danthine (U. de Lige) in the "Comprehensive International LOTOS Toolkit" project are developing practical tools to make it easier for communications software engineers to take advantage of the characteristics of the formal language LOTOS. In addition, real-life applications for LOTOS are being studied and the language is being extended to make it more generally applicable. Several industries and government organizations in Canada and Europe are supporting this project.
iv) H. Ural (UO), K. Inan (METU)
This collaboration is motivated by the European Community COST 247 research project and is focused on formal methods in communications software validation with a particular emphasis on communication systems that are specified in formal description techniques, namely SDL, ESTELLE, and LOTOS.
v) R.L. Probert (UO), and Kassem Saleh (Kuwait University) in "Reverse Engineering: Recovery of Communications Protocol Design from Execution Traces". One of the major concerns of the Telecommunications and Computer Industries is the re-engineering of existing software to upgrade performance, portability, reliability, and to add new features in as seamless a manner as possible. Changing existing code, however, is highly error-prone and counter-productive. Thus, methods for recovering the original design rationale are much sought after. This project was inspired by issues observed and suggested through collaboration with Nortel, Mitel, IBM, and Stentor, and supported by TRIO and NSERC through industrial partnership grants. This project has established the underlying theory and is currently studying issues of industrial-scale application.
vi) R. Probert (UO), L. White (Case Western Reserve University) in the "Large System Interface Testing Strategies" project in which a mathematical modelling technique based on experimental design principles is adapted to minimize the cost of testing combinations of system configurations to ensure robustness and reliability of large-scale distributed systems before they are released to the customer. The project draws on techniques from finite fields and orthogonal Latin squares, as well as approximation algorithms. The reduction in testing cost by using these techniques can be several orders of magnitude. Corresponding software tools are under development in Smalltalk and C++, and preliminary versions of these tools have been demonstrated to several interested corporations.
One of our major projects is in the framework of TRIO, the Telecommunications Research Institute of Ontario, which is one of a number of centres of excellence funded by the province of Ontario.
The title of our largest project is Validation-Directed Software Engineering. The Principal Investigators are R.L. Probert and L. Logrippo, and the co-Investigators are J.P. Corriveau and H. Ural.
These four TSERG professors are pooling their efforts to arrive at an integrated and sound approach to the life-cycle of communications software development. The project is in collaboration with several telecommunications corporations, and with other TRIO researchers located at Carleton University in Ottawa, and at McMaster University in Hamilton.
This research involves "pragmatic" theoretical foundations relating a variety of formalisms such as LOTOS, Message Sequence Charts, and TTCN. An integrated protocol engineering workbench of software design methodology tools is being developed and applied to protocols of current interest such as ISDN, GSM, and ATM.
The common theme in our approach is VALIDATION. Given a description or an implementation of a protocol or communications software, does it really do what is desired? The validation environment will provide validation-oriented tools to assist the software designer to develop reliable and useful specifications, designs, and test suites. Since validation is conditional upon proper design, design models and architectural models are being investigated as well. We follow the standards organizations ISO and ITU's efforts towards the definition of an 'Open Distributed Processing' (ODP) reference model.
TSERG professors and students tend to be pioneers in new programs and achievements. A. Williams, a TSERG Ph.D. student, was one of the first recipients of an NSERC/Nortel post-graduate industrial scholarship, for example. A previous TSERG Ph.D. graduate received the first TRIO/IBM post-doctoral fellowship. H. Ural has been developing formal test construction methods and design verification techniques as well as establishing lower bounds on the size of the complete test suites. His work has been used by the Corporation for Open Systems (COS) in Virginia and by Nortel Ltd. for developing tools for automatically generating high-coverage test suites. L. Logrippo was a founding participant in the development of the International Standard formal description language LOTOS and has developed a comprehensive toolkit for it. R. Probert was the first non-employee to receive the Nortel President's Award of Excellence for Telecommunications Protocols research and development especially in the development of the international standard testing language TTCN. J. Raymond has developed new techniques and software tools for multi-media long-distance learning and pioneering the implementation of these methods on low-cost phone networks and on the internet. G. White has developed constraint programming methods which allow a novel approach to designing some types of large information systems, such as scheduling systems. J.P. Corriveau is developing entirely new paradigms for Object-Oriented traceability.
We are active in a new TRIO initiative, the "Industrially Specified Research Program" which promotes the concept of students and professor(s) on site in the industrial lab to maximize communication with the relevant industrial software engineers. In a new TRIO research internship agreement, a company which develops a very successful object-oriented real-time software design environment called ObjecTime has agreed to sponsor this type of collaboration with Probert, Corriveau and Ural. Several graduate students are involved, although only two are directly supported by the research contract. ObjecTime infuses cash and in-kind support (personnel, software licenses, access to design documentation) which is matched by TRIO. This project, entitled "An Industrial Strength Quality Engineering Methodology and Framework based on TTCN/MSC Notations and Tools", is intended to investigate and propose techniques for validating models of object-oriented real-time system analysis and design. The project marries the TSERG expertise in TTCN, validation frameworks, and object-oriented design and test to the ObjecTime expertise in real-time design and to the highly successful ObjecTime product line.
NSERC has funded two three-year strategic programs of research by Ural and Probert studying formal methods in design and validation of communications protocols and distributed systems.
Nortel and the Government of Ontario (URIF) funded a project, involving Logrippo and students, which dealt with the use of LOTOS methodology in the development of mobile telephony protocols. A larger project, involving also the development of ODP-based architectural models for mobile telephony, is being funded by Motorola and NSERC.
Many TSERG projects are in progress, often leading to exciting, cutting-edge collaborative research with industry. Research prospects for the future in protocols and object-oriented communications software engineering are very promising.
Contact: School of Information Technology and Engineering
Copyright © 2000 University of Ottawa