Hi!
I'm working on a long-distance laser communication project. The principle is simple: a first raspberry pi pico receives text, converts it into binary and flashes a laser, every T seconds. On reception, phototransistor values are recorded, with measurements taken every T/3 seconds. Once the signal has been received (the recording is switched off manually), a KNN algo takes care of detecting the 0s and 1s among the measurements. The result is a string of 0s and 1s, with 3 times as many measurements as bits sent. The aim is to reconstitute the initial chain (knowing that there are some measurement errors and “offsets”), and that's the problem. For now, I just divide the string by blocks of 3 bits, average the 3 to find out whether the bit was actually a 0 or a 1, and reconstruct the binary string. The problem is that there are “offsets” in the measurements, due to the delays of the two raspberries. On short messages at low frequencies, these delays are not perceptible and the message received is correct, but this is not the case at higher frequencies and for longer messages. If you have any ideas for improving these programs, I'd love to hear from you!
I'm working on a long-distance laser communication project. The principle is simple: a first raspberry pi pico receives text, converts it into binary and flashes a laser, every T seconds. On reception, phototransistor values are recorded, with measurements taken every T/3 seconds. Once the signal has been received (the recording is switched off manually), a KNN algo takes care of detecting the 0s and 1s among the measurements. The result is a string of 0s and 1s, with 3 times as many measurements as bits sent. The aim is to reconstitute the initial chain (knowing that there are some measurement errors and “offsets”), and that's the problem. For now, I just divide the string by blocks of 3 bits, average the 3 to find out whether the bit was actually a 0 or a 1, and reconstruct the binary string. The problem is that there are “offsets” in the measurements, due to the delays of the two raspberries. On short messages at low frequencies, these delays are not perceptible and the message received is correct, but this is not the case at higher frequencies and for longer messages. If you have any ideas for improving these programs, I'd love to hear from you!
Statistics: Posted by erhelito — Thu Feb 20, 2025 5:58 pm