Since the advanced work done by Elowitz et al., the study of stochasticity in gene expression has inspired many experimental and theoretical studies.  Stochasticity is generally regarded as a destructive factor because fluctuations in protein level can inevitably obscure the transduction signals and negatively affect the cellular regulation.  Most previous works therefore focused on the design of networks which could reduce or resist fluctuations.  In the present talk we explore another possibility in which biochemical systems can instead exploit the inherent noise.  We focus on the p53 regulatory network, the core of a cell's modulator for switching between survival and apoptosis.  We found that the fluctuation, originating from stochastic expression of p53-responsive genes, introduces marked advantages for the system to sense external stimuli and sustain the function of switches.  The underlying mechanism further broadens the range of parameters exhibiting the survival-apoptosis switches functionally.