For the image processing part, our model fits all the requirements. Mobilenet allows us to install the model into the Raspberry Pi4, at the same time, it could keep the high accuracy. Moreover, because it is lightweight, the speed of analyzing is fast (0.019s per image). Our model could save the image to the specific file which could benefit the wireless communication part.

For the wireless communication part of our project, we almost finished all objectives we previously set. First, we built two functional sequences based on the USRP transmission on the GNU Radio software, one is supposed to be connected to the transmitter side, and the other one is supposed to be connected to the receiver side, and they have been tested to make a successful transmission. After the spring break, we moved to the dormitory so we began to build a simulated sequence. After we realized the simple transmitting and receiving function, we figured out there existed some error when doing the transmission. Thus, we analyzed the whole sequence and added several functional blocks to improve the reliability as well as the efficiency of the system. Finally, we finished with a functional sequence which can realize a more accurate image transmission.