In this paper, the problem of designing state feedback decentralized control schemes for a class of networked large-scale systems with sensor network failures is developed. Because the sensor networks may be disturbed by perturbations and time-delays, and also be served by loss of transmitting effectiveness, the performance of networked systems will be degraded by using the faulted signals. Under the unknown assumption for the failures, adaptation laws are proposed to estimate the transmitting factors and the upper and lower bounds of the state error, which composed by the effects of perturbations and time-delays on-line, respectively. Then based on the information of adaptive schemes, a class of decentralized state feedback controllers is designed for guaranteeing the stability of the closed-loop large-scale adaptive control system in the presence of the uncertain faults. The proposed adaptive design technique is finally evaluated in the light of a simulation example.