Aiming at distributed H∞ filtering problem of measurement data loss in case of uncertainty in wireless sensor networks（ WSNs）

a distributed H-infinity filtering algorithm based on sensor uncertainty prediction of packet loss rate is proposed. By designing a linear full-order filter

so that filtering error can converge in mean square asymptotically to zero

while suppressing H ∞ filtering attenuation caused by external interference.According to Bernoulli distribution filtering parameters consistent with the principles of random distribution

by searching the optimal filtering parameters to ensure random stability in the case of packet loss rate uncertainty of true value prediction

using the Lyapunov function. Simulation results show that this filtering algorithm can play an effective filtering effect if there is uncertainty observational data loss.