Multiagent-based fault tolerance management for robustness

Abstract

Despite the use of software engineering best practices and tools, it would be very risky to assume that the software that is developed today is fault-free. Moreover, we have to consider the fact that the software could face unexpected situations not considered during its design. Robustness is a highly desirable and sometimes indispensable software requirement, especially for critical systems, where the consequences of a system failure can be catastrophic. This chapter outlines existing fault tolerance techniques, followed by a discussion of the potential that multiagent systems have to enhance the design of robust, fault-tolerant systems, thereby improving large-scale, critical, and complex system reliability. © 2008 Springer-Verlag London.