In the Taiji program, laser interferometry is utilized to detect the tiny displacement produced by the gravitational wave signals. Due to the large-scale unequal arm, the laser frequency noise is the largest noise budget in the space interferometer system. To reduce the influence of laser frequency noise, a technology called the Time Delay Interferometry (TDI) is utilized to deal with it. The TDI is a kind of data post-processing method, which forms the new data stream by the method of the time delay to initial data. But the premise of TDI needs to obtain accurate absolute arm length between satellites. Thus, for that requirement, we discuss the ranging system scheme and implement a ground electronics verification experiment. The ranging system is based on Direct Sequence Spread Spectrum (DS/SS) modulation, and it mainly includes three parts, which are the signal structure, a Delay Locked Loop (DLL), and a data processing algorithm. In DS/SS modulation, types of pseudo-random code can make a difference to the quality of correlation and the ranging accuracy. Therefore, to design the optimal pseudo-random code, we compare the correlation and flexibility in choosing lengths of the m sequence, gold sequence, and Weil code. Weil code that has a shift-cutoff combination with the best autocorrelation is chosen as the ranging code. The ground electronics verification experiment is set up for simulating the physical process of signal transmission and verifying system performance. The main device of the experiment is a FPGA card based on the K7 chip from Xilinx, which is used to simulate the function of communication and ranging between satellites. Meanwhile, we change the length of the Radio Frequency (RF) coaxial cable to correspond to different ranges. The experimental process can be summarized as follows. Firstly, 16-bit data at 24.4 kbps and 1024-bit Weil code at 1.5625 Mbps are modulated with Binary Phase Shift Keying (BPSK) in the 50 MHz sampling frequency. Then the signal is transmitted through RF coaxial cables of 10 to 60 m in length. In receiving end, the signal is consolidated by DLL and the ranging information is collected. To measure the range accurately, we use a centroid method to optimize the collected data. The results show that the ranging accuracy is better than 1.6 m within 60 m. In conclusion, this experiment proves the principle of the scheme and its feasibility, laying a technical foundation for optical system verification in the future.