The Kalman filter is a set of mathematical equations that provides an efficient computational (recursive) mean to estimate the state of a process, in a way that minimizes the mean of the squared error. This filter is very powerful in several aspects: it provides estimations of past, present, and future states, and it can do so when the precise nature of the modeled system is unknown, and even with the presence of measurement and process noise. Moreover, Kalman filter for linear estimate is the most complex and precise algorithm used for target tracking. However, using Kalman filter algorithms in software for multitarget tracking (MTT) radar system would result in a very long computational time which may not be suitable for today’s warfare constraints, or real-time processing. Consequently, a hardware alternative has to be developed which may result in big area overhead which is not suitable for today’s area constraints such as sensor nodes in a sensor network. In this paper, we break the arrays into their scalar forms, and develop fully-pipelined hardware architecture for the radar tracking Kalman filter, with time division multiplex blocks to decrease the silicon area.. The proposed architecture contains 6 multipliers, 2 dividers, 9 adders, 5 subtractors, one control unit, and some registers and multiplexers for pipeline and control. Simulation results show that the loss in accuracy between the exact track and the estimated is found to be only 4.9%.
展开▼
