Signal-to-noise losses associated with hardware-efficient designs of moving target indicators

In many radar systems the signal processor contains a moving target indicator which is implemented by either an infinite or finite impulse response digital filter. The finite impulse response configuration is most often used, and there are two major implementation alternatives: the moving window (or sliding window) and the fixed window (or block window) methods. The hardware requirements and the signal-to-noise characteristics for various implementations of moving and fixed window moving target indicators are determined and compared. It is demonstrated that the moving window configuration produces correlation of noise samples which is not present in fixed window systems. Equations are given that relate the effective number of outputs from these systems when the same number of input samples are used. It is shown that a square-law detector, or a linear detector, does not alter the relative performance of the moving and fixed window moving target indicator systems. A tabulated result is given that compares hardware (i.e., memories, adders, multiplexers, multipliers, multiplication speed) for the moving and fixed window systems. One significance of this effort is that it allows a tradeoff to be made between the better signal-to-noise improvement for the moving window system and the less complex hardware requirements of the fixed window system.