Deterministic Network Calculus

Conclusion

Network calculus is a theory designed to compute upper bounds on delays and memory usage in networks, and one aim of this book is to set a solid foundation for future works.

From the pioneering works of [CRU 91a, CHA 94, PAR 93, CRU 96, LEB 96, CHA 97] in the 1990s up to this book¹, thousands of research articles on network calculus have been written². While the first works were mainly motivated by the support of quality of service in the Internet, the current industrial applications seem to relate to the domain of critical real-time systems like avionic and factory systems, where a deadline miss may result in catastrophic system failure, including human injury, and where other methods like simulations are not sufficient to get enough confidence.

Currently, there are about 12 tools based on network calculus, of which about half are related to industry, which are listed in section C.1.

Thus, network calculus looks like a mature research domain with great successes.

As a (partial) conclusion, we would like to highlight in section C.2 the contributions, benefits and drawbacks of network calculus before giving our point of view on the remaining challenges in section C.3.

C.1. Tools

Several tools have been developed in the field of network calculus, including many academic prototypes designed to support theoretical experiments and several industrial products, mainly devoted to avionic or factory systems. Each tool needs to perform more or less complex operations on functions. ...