We present, in this paper, ongoing work that investigates a new error detection policy aiming at enhancing communication integrity in the presence of permanent errors (single and multiple). We consider critical embedded systems which are based on complex networks including active interstage nodes. This property increases the occurrence probability of permanent errors. The novelty of the proposed policy lies in the fact that unlike classical policies using a single error detection function, it is based rather on a set of different error detection functions. The different used functions must be complementary in terms of detection capability in order to increase the resultant error detection capability. Our reference application to illustrate the proposed concepts is the Flight Control System (FCS). However, our objective is also to apply the proposed approach to other application domains sharing similar features and characteristics.