The current 5 th Generation-New Radio (5G-NR) systems are designed to meet the demands of various applications, offering high data rates, low latency, and high reliability. In the current 5G-NR systems, the Discrete Fourier Transform-spread-Orthogonal Frequency Division Multiplexing (DFT-s-OFDM) waveform is the most commonly employed waveform to transmit user data in uplink transmissions, specifically in coveragelimited scenarios. To enable coherent data demodulation, pilot signals, commonly referred to as Demodulation Reference Signals (DMRS), are transmitted along with data-carrying DFT-s-OFDM symbols. However, in the current DFT-s-OFDM architecture, the DMRS and data are transmitted on distinct OFDM symbols. This time separation necessitates the demodulation process to commence only after receiving the initial DMRS symbol, resulting in significant processing delays. Furthermore, the existing architecture consumes more DMRS resources and necessitates complex time interpolation techniques to support users with high mobility. To address these challenges, we propose an improved DFT-s-OFDM architecture that enables instantaneous data demodulation, leading to a substantial reduction in processing delays. Furthermore, despite not utilizing any time interpolation techniques, the proposed method effectively caters to high-speed users, thereby conserving computational resources and hence contributing to minimizing the system complexity and latency. We thoroughly investigate the proposed architecture and evaluate its performance in different simulation settings. The results demonstrate that the proposed architecture significantly improves packet error performance, particularly in high Doppler scenarios, while using 16% lesser DMRS overhead. This reduction in DMRS overhead frees up additional resources for data transmission, ultimately enhancing data rates.