UDP packet loss whilst moving around

[Mutski]

Hi all

I have an application where my ESP32 is receiving a fairly constant stream of UDP packets whilst it is physically moving around. I find that when I get about 20 meters (60 feet) away from the access point, that there starts to be a lot of packets dropped (maybe like 10-20%) if moving around, but rock solid once stationary. I don't believe it's a weak signal as I can move right out to about 200 meters (600 feet) from the access point before it starts to drop the majority of packets.

I thought maybe it was multipath issues so I have also tried out in an open field across a football oval where there is nill competing WiFi and no buildings etc to generate multipath issues. Always direct line of sight to the access point, great RSSI, but same results.

Does anybody have experience with this and any kind of solution I can try?

Thanks - Matt

[ESP_Sprite]

Could be Doppler from your device accelerating/decelerating inconsistently making the signal harder to 'lock on' to, but I'm just guessing here, and I wouldn't really know what to do about it.

[MicroController]

Current ESP32 chips don't have integrated propulsion systems so I wonder what makes your chip "move around".
Some kind of electric motor by any chance?

[Mutski]

Current ESP32 chips don't have integrated propulsion systems so I wonder what makes your chip "move around".
Some kind of electric motor by any chance?

Yeah I've attached a 1HP motor (human power) to it. . The human has legs and arms.

[MicroController]

Yeah I've attached a 1HP motor (human power) to it. . The human has legs and arms.

Interesting approach. I may try that too in a future project (Starting with one arm and one leg for the prototype, of course.)
What I was actually meaning to get at is that I don't think that motion by itself, esp. at walking speeds, would interfere noticeably with WiFi reception. EMI from a running motor might, but I'm confident your human's emissions in the WiFi band are 1) pretty low and 2) invariant w.r.t. moving vs. standing still.
Apart from velocity, is there anything else in the system which would be different when in motion? Varying antenna orientation maybe? (Or something on the software side (accelerometer,...)?)

[Mutski]

No EMI from motors or anything special when it's in moving around. I have tried with the antenna in all orientations. It happens if you hold your arm out and wave it around - always direct line of sight to the access point. Also happens if you just walk around lazily, again always keeping it direct line of sight to the access point. The Doppler effect does seem the most likely culprit, but how to solve it may be difficult.