Weight-convergence analysis of adaptive antenna arrays based on SMI algorithm

Weight-convergence analysis of adaptive antenna arrays based on the sample matrix inversion (SMI) algorithm is presented. First, we derive a required number of samples for weight computation to achieve an output signal-to-interference-plus-noise ratio -10 log₁₀ r dB (r ≤ 1) below an optimum value. Next, we evaluate weight convergence of code-division multiple-access adaptive antenna arrays based on the SMI algorithm, where two types of weight computational methods are compared between "chip-level computation" and "symbol-level computation". Analytical and simulation results reveal that "chip-level computation", before despreading process, enables faster convergence than "symbol-level computation", after the despreading process.