Hardware design issues of fuzzy neural networks

It has been well established that fuzzy neurons and fuzzy neural networks (FNN) are highly adaptive to changing conditions, possessing robust learning capabilities and inherent transparency for supporting a high level of knowledge interpretability. Consequently, they have the potential to provide exceptional mechanisms for building intelligent systems that must operate in dynamic and rapidly changing environments. However, to fully exploit the potential of FNN structures and their parallel nature, efficient hardware implementation techniques need to be developed. Here we are concerned with their realization using "standard" digital hardware so they may appeal to a wide range of applications, and our objective in this study is to investigate this avenue and identify various critical design issues.