In order to achieve more comfortable and convenient physiological information monitoring, this paper proposes a grid structure-based physiological information monitoring sensor (Grid Optical Fiber Sensor, GOFS), which can be used to monitor the breathing and heartbeat of patients in resting and walking states. The sensor uses the phenomenon of chest expansion during respiratory movement and heartbeat movement of the human body, and applies pressure to the optical fiber through the grid to form microbending loss, thereby obtaining breathing signals and heartbeat signals. At the same time, based on the loss models of microbending caused by pure bending and that by coupling, a microbending loss model based on grid structure is established. In terms of data processing, EMD-ICA algorithm is used for noise cancellation, and it shows high consistency in the final comparison test with commercial detection equipment. The overall respiratory data and heartbeat data were analyzed by Bland-Altman plot, and the results showed that within the 95% confidence interval (±1.96SD), the respiratory measurement error of the sensor was $0.08\pm 0.83$ times/min, and the correlation with the standard value was 0.9974; the heart rate measurement error was $-0.4\pm 4.2$ times/min, and the correlation with the standard value was 0.967. Calculating the MAPE value for each set of data, the maximum error between the respiratory measurement results and the measurement results of commercial testing equipment is 2.298%, and the maximum error of the heart rate measurement is 2.05%. The tests show that the physiological monitoring sensor based on grid structure designed in this paper has a wide application prospect in the field of medical rehabilitation monitoring.