Robustness improvement in binary cellular non-linear network architectures

This paper introduces a systematic approach to enlarge the robustness in binary cellular nonlinear networks (CNN). In particular, the work is devoted to positive range CNN models with high gain nonlinearity and 1-bit of programmability. The robustness is increased by appropriately modifying the bias/threshold term. The CNN cell model and the robustness improvement method are presented within the framework of threshold logic gate (TLG) design, where they prove to be valuable approaches.