At present, Resonant Optical Gyroscope (ROG) is developed towards integration, miniaturization, low power, and high resolution. However, the rotation signal of ROG is weak, and easily affected by reciprocal and nonreciprocal noises within the gyro system. With the growth of technical studies in ROG in recent years, the performance of ROG has been significantly improved. The improvement of noise suppression techniques and the combinations of multiple techniques in singe ROG system played important role in the measurement of a high signal-to-noise ratio and high precision rotation signal. With the aim of integration or miniaturization ROGs, noise suppression techniques also need to be further developed.So far, more studies on noise suppression focused on the suppression of nonreciprocal noises, which have different or inverse influences in the two counter propagating lights in the ROG system. The main part of nonreciprocal noises is backscattering noise. For the suppression of this kind of noise, a review of current studies is performed. More studies using the improved modulation techniques for noise suppression, which separate the sidebands in frequency domain in different ways. The separated frequencies cause lower interferences of the backscattering light or the influence can be filtered in the subsequent signal processing steps. However, the residual intensity modulation induced error appears as the modulation difference of the two loops. The error must be eliminated by improved demodulation designs. Another approach for backscattering noise suppression is using of multiple light sources. In this technique, Optical Phase-Locked Loop (OPLL) is used to lock the lasers in stable frequency difference, resulting in low coherence and interference.The second strength part of nonreciprocal noises is Kerr noise. The techniques of Kerr noise suppression are also reviewed. The main idea of these techniques is laser intensity monitoring. Realization of laser intensity monitoring used in ROGs is usually based on the secondary harmonic demodulation of the detected phase modulated laser signals. Theoretical and experimental studies have been carried out on the linear relation of the signal and laser intensity. As for intensity stabilization, there must be intensity modulators in ROGs to control the laser intensity according to the feedback of intensity related demodulation signals, which introduce more optical devices. Therefore, the studies for simplified intensity feedback loop are also performed, which use fewer modulators or directly compensate the induced error in the gyro output.A less effective part of the nonreciprocal noises is polarization fluctuation noise. The previous studies of this kind of noise are also presented. At present, the widely used polarization maintain fiber and polarizers in ROGs make the suppression of polarization fluctuation noise easier and more effective. However, for the use of WRR in the integrated ROGs, the resonators usually do not have the polarization maintain property inside, which remains a better solution requirement.As for the reciprocal noises in ROGs, with the intended reciprocal design of ROG and the nonreciprocal property of Sagnac effect, more important noise source lies inside the ROG system, namely the frequency and intensity noises of the light source. For the frequency noise suppression, more advanced laser frequency locking techniques can be performed. The presented reports introduce more laser frequency locking loops to improve the locking precision, which also improved the zero-bias stability of ROGs. Another approach for frequency stabilization is the use of OPLL, which works well in multi-laser ROG system. The self-injection locking technique in the laser side is also presented as an approach in frequency stabilization. The technique features a whole optical loop in the phase locking loop but precision improvement is still needed. The other noise of laser intensity here is mainly for the output power fluctuation in frequency domain, which cause the relations in the ROG for rotation sensing not as perfect as theoretical designs. The suppression of this type of noise is more considered in the laser design.Finally, tendency of integration and miniaturization of ROGs is refocused. For this reason, the ROG design must be reexamined. 1) Miniaturization of suppression systems. Miniaturization of optical devices makes lower signal-to-noise ratio as less robust against noise interferences. Besides, the adding of optical devices for noise suppression is also limited, which limit the realization of some noise suppression techniques. The more advanced optical designs and new techniques for multiply noise suppression in simple design are needed in the future. For example, the use of integrated optics design and the realization of integrated resonant cavities. 2) Fast response of noise suppression processing. More suppression techniques and feedback loops in use requires more complex signal processing and control algorithms. Powerful and simplification of design such as efficiently use of FPGA is also important in the future for fast processing. 3) Intelligent noise suppression technology. With the complexity of gyro systems, it is difficult to comprehensively analyze the ROG error analysis mechanism and construct error models, Intelligent algorithms are needed to help achieve this goal, such as the nonlinear suppression method based on cubature Kalman Filter-Phase Space Reconstruction (CKF-PSR) and the improved Variational Mode Decomposition (VMD) method. 4) Novelization of noise suppression mechanisms. New mechanisms and novel sensing structures should be presented, such as the use of broadband light sources, so that the coherence length of laser as well as backscattered light is shortened and good suppression can be achieved without adding additional optics. On the other hand, with the development of new types of optical gyro, new types of noises, such as optical power noise due to resonance tuning and gain fluctuations, and quantum noise, are constantly being derived, in spite of the consequent increase in the sensitivity and signal-to-noise ratio. Therefore, new mechanisms need to be developed to suppress these noises.