An Optimized Method for Nonlinear Function Approximation Based on Multiplierless Piecewise Linear Approximation

In this paper, we propose an optimized method for nonlinear function approximation based on multiplierless piecewise linear computation (ML-PLAC), which call OML-PLAC. OML-PLAC finds the minimum number of segments with predefined fractional bit width input/output, maximum shift-and-add operations, user-defined widths intermediate data, and absolute error (MAE). addition, minimizes actual MAE as much possible by iterating. As a result, under condition satisfying segments, can be minimized. Tree-cascaded 2-input 3-input multiplexers are used to replace multi-input in hardware architecture well, reducing depth critical path. The is applied logarithmic, antilogarithmic, hyperbolic tangent, sigmoid softsign functions. results implementation prove that has better performance than current state-of-the-art method.