A VLSI Design of Real-Time Sensor Fusion Unit using Kalman Filters

As the functions of mobile smart devices are diversified with the increase of sensor types, the importance of sensor hubs for sensor data processing is also increasing. One possible application of the mobile sensor hub is to calculate the orientation of the sensor. In this paper, we design the sensor fusion structure as VLSI circuit that can be implemented as a co-processing unit which identifying device orientation using 9-axis IMU sensor and Kalman filters. The simulation reported an average error of about 2.27 degrees. The proposed design is fully verified and evaluated through simulation tools for Artix-7 FPGA and 32 nm CMOS standard logic cell library. The average power dissipations of FPGA and 32 nm standard cell technologies were estimated as 7 mW and 0.89 mW, respectively. Also, the gate counts for standard cells are reported as similar to 45,000 in total. The sophisticated design is applicable to the digital IP core for mobile sensor hub architecture.