An FPGA-based controller for a 77 GHz MEMS tri-mode automotive radar

This paper presents a Xilinx Virtex 5 FPGA platform based signal processing algorithm that was designed, implemented and experimentally verified for use in a MEMS based tri-mode 77 GHz FMCW automotive radar to determine range and velocity of targets in the vicinity of a host vehicle. It provides short (SRR), medium (MRR), and long-range radar (LRR) coverage using a single FPGA. The MEMS radar comprises of MEMS SP3T RF switches, microfabricated Rotman lens and a microstrip antenna embedded with MEMS SPST switches, in addition to other microelectronic components. A CA-CFAR module has been used to eliminate false targets in a multi-target clutter affected scenario. The developed algorithm enables the MEMS radar to detect 6 targets in a time span of 6.1 ms between a distance of 20-170 m with range resolutions of 0.07, 0.11 and 0.19 m respectively for SRR, MRR and LRR. A maximum relative velocity of 300 km/h can be determined with a velocity resolution of 6.84 km/h. The refresh rate is 2.048 ms for each mode of radar which is nearly 40 times lower than the commercially available BOSCH LRR3 radar. The developed FPGA based radar signal processing algorithm can be implemented in an ASIC which can be batch fabricated to lower the production cost for high chip volumes. This will enable automotive radars to become a standard item for all the vehicles on the road.