SNR of Raman distributed fiber optic sensing system gradually decreases with the increase of sensing distance, the power of a single pulse will be limited by the nonlinear scattering threshold, which can't be infinitely improved. Although the incident optical power can be increased by increasing the pulse width, the spatial resolution of the system will be weakened, which makes it difficult to realize the high spatial resolution temperature sensing at longer distances to satisfy the engineering requirements. In this paper, a strong autocorrelation pulse coding technique is proposed. The Golay complementary sequence with autocorrelation characteristics is introduced into the Raman distributed fiber sensing technology, and anti-Stokes scattering signal is analyzed and reconstructed by Fourier transform to restore the autocorrelation of the Golay complementary sequence destroyed by transient effect of EDFA. This technology can significantly increase the optical power on the premise of no stimulated Raman scattering, which effectively improves SNR of the system and realizes the long-distance Raman distributed fiber optic sensing technology.In this paper, Raman distributed fiber optic sensing system based on the strong autocorrelation pulse coding technique (SAC-codes) is established (Fig.1), and the correctness of the technique is verified by numerical simulation experiments, which achieves the accurate identification of temperature demodulation and localization of FUTs (Fig.4). The dominances of the technique are verified by analyzing the SNR after decoding. The scheme of temperature demodulation is utilized to obtain the temperature information in FUT, which verifies the performance enhancement of the technique in terms of temperature resolution (Fig.5), and obtains the final spatial resolution of the system (Fig.6).The correctness of the SAC-codes scheme is verified by numerical simulation experiments, and the accurate identification of temperature demodulation and localization of the FUT is achieved. The performance of the single-pulse scheme, the Golay-codes scheme affected by transient effects, and the SAC-codes scheme are quantitatively analyzed, numerical simulation experiments show that the SNR at the fiber tail is improved from 9.14 dB to 10.29 dB (Fig.3), which verifies the superiority of the SAC-codes scheme. Meanwhile, the performance enhancement of the SAC-codes scheme is experimentally explored, which dramatically improves the temperature resolution of the system, which improves from 22.69 ℃ to 4.12 ℃ (Fig.5), eventually, a spatial resolution of 1 m was achieved based on a 100 km sensing distance (Fig.6).In this paper, a Raman distributed fiber optic sensing technique based on strong autocorrelation pulse coding is proposed to realize the need of long-distance temperature monitoring. The feasibility of this technique is verified by numerical simulation, and finally, it is proved experimentally that this technique obtains a large improvement in performance indexes such as SNR and temperature resolution compared with the traditional single-pulse scheme, which provides a new solution for the research of long-distance Raman distributed fiber sensing system.