AREA: an Automatic Runtime Evolutionary Adaptation mechanism for Creating Self-Adaptation Algorithms in Wireless Networks

The application requirements and the spatial environments of wireless networks continue to become more and more complex and changeable. Most existing algorithms for wireless sensor networks are designed with a specific type of environment in mind. While such algorithms work well in the environment they have been designed for, once the environment changes beyond the domain in which the design can adapt, the algorithms can hardly work properly. This paper proposes a novel design mechanism called the automatic runtime evolutionary adaptation (AREA) mechanism. It has been designed to automatically adapt, during runtime, to the variation of environments in wireless networks. This adaption is realized by self-creating and selfevolving algorithms: AREA allows the created algorithm not only to be adaptive but also to evolve to other adaptive abilities according to the variation of the application requirement and the spatial environment. The AREA mechanism is validated by applying it to a data aggregation example in wireless networks. This shows that the mechanism can adapt to the changing environments and outperform the other strategies.