This article argues that full-parallel algorithms of traditional sigmoidal neural networks based on average fire rate which have excluded useful time operation seem improper for the invariance.Thus. we present a spiking network that intreduces a space search mechanism into neural computing. The search can transform space coordinate into time coordinate. Consequently

the operations based relative positions are converted into those based relative time which are very easy for neural realization through delay connections. Accordingly

we develop a world-centered model （WCM） that consists of a space searcher and a feature transfer vector （FTV） memory. WCM is a pure neural network that represents the neural principle of invariant recognition.