抄録
This paper proposes a new high resolution radiometer used as an airborne or a spaceborne microwave imaging sensor. In these applications, the resolution of radiometer is restricted by large ambient temperature variation and the short observation time available for each pixel which the antenna scans succesively.
The proposed system has two-different-temperature reference noise sources at the input, which is periodically switched amang the antenna and the two references to detect true gain variation3), 4), and a digital signal processor at the backend which calculates signal and gain variation using stored data. Since reference noise measurements are not limited by antenna scan, and hence, large number of reference data taken in different period can be used for smoothing the rms fluctuation, the use of digital processing of stored data is advantageous to decrease the statistical uncertainty in the reference measurements.
This paper presents a complete analysis of the radiometer performance, signal processing algorithm which eliminates cross-talk between signal and AGC data, and total rms fluctuation obtainable. The estimated total rms fluctuation attains as low as that of usual Dicke type as increasing number of reference data used, and asymptotically, close to that of Total Power type. The results of experimental test of the radiometer are also described.
