Abstract: This paper presents a measurement-based study on channel modeling and simulation for low-antenna-height Mobile Ad Hoc Networks (MANET) in complex 5.8 GHz urban street environments. It addresses the challenges of modeling non-Gaussian tap amplitudes while simultaneously preserving Doppler correlation structures. First, a high-precision broadband system was developed to capture time-varying channel impulse responses (CIR). Using the Space-Alternating Generalized Expectation-maximization (SAGE) algorithm and Akaike Information Criterion (AIC), an accurate Tapped Delay Line (TDL) model was established. Results indicate significant statistical differences between line-of-sight (LOS) and non-line-of-sight (NLOS) conditions. Dominated by dense scattering and diffraction, NLOS exhibits a root mean square delay spread (RMS-DS) of 409 ns compared to 197 ns for LOS, while the coherence bandwidth decreases from 1.149MHz to 0.508MHz. Furthermore, a Copula-based statistically consistent simulation method is proposed. By employing a complex Gaussian reference process to carry the target Doppler Power Spectral Density (DPSD), arbitrary amplitude distributions are strictly embedded via Probability Integral Transform (PIT) and inverse distribution mapping. This approach utilizes a phase inheritance mechanism to maintain temporal structures, effectively decoupling amplitude distribution from time correlation. The generated channel sequences accurately approximate both target amplitude and DPSD statistics, achieving a mean K-S distance of 0.011 for amplitude consistency and a mean DPSD NMSE of -18.8 dB. This method reliably reproduces the statistical characteristics of complex urban channels, providing a foundation for communication system design and simulation.