Low Power SoC for Mil-Std 1553 Cable Fault Detection

The safety of aircraft operation has always been the major need across globe. Faults on the aircraft system wiring have led to disasters, causing loss to human life and nation's economy. To avoid such incidents, early detection of faults like Opens or Shorts on cables play a very important role in preventing catastrophic failure and ensuring safety of flight. Any fault like Open or Short occurring on the Mil-Std 1553 cable used in aircraft systems is addressed by introducing redundancy having 1:1 or 1:2 channels instead of only one. A fault detection solution comprises of a Test Signal generation method and a processing method. From the literature survey, the author infers that Spread Spectrum Time Domain Reflectometry (SSTDR) is the Test Signal generation method used for avoiding interference between Test Signal and the actual Data Signal on Mil-Std 1553 cable. Continuous Wavelet Transform (CWT) can be used for statistical feature extraction and Probabilistic Neural Network (PNN) can be used for fault type localization and classification. So CWT-PNN algorithm is the processing method. In this research, the author has proposed and implemented a fault detection solution which is the combination of SSTDR, CWT and PNN. The test set-up inclusive of this combination has been simulated for three different scenarios, namely No Fault, Open Circuit Fault and Short Circuit Fault in MATLAB environment with the modeling of Mil-Std 1553 cable in SIMULINK. A prototype of the Mil-Std 1553 cable fault detection system is implemented on Artix 7 Nexys DDR Field Programmable Gate Array (FPGA) as a System on Chip (SoC) design. It is observed that for both open circuit and short circuit cases, the simulated faults in MATLAB-SIMULINK environment are conforming to the simulated faults from the FPGA prototype with an accuracy of less than 1mm cable length. The FPGA prototype power consumption is 16mW.