All of us know the limitations of FPV, especially when it comes to range and reliability. So is there an alternative? Some people in the hobby are starting to experiment with Cellular towers and data in order to send video signals as well as control signals to FPV planes and in some cases Drones.
How it Works
In layman’s terms the technology works by your UAV sending a signal back to your tablet, laptop or phone rather like normal FPV and you controlling it by sending a signal on the phone network which is then relayed to the SIM card in connected to your Pixhawk. It is basically like normal FPV and radio control apart from the fact that everything is done via the nearest Cell tower. The benefit of using Cell towers is a huge increase in range, especially near urban areas where cell coverage is often highly consistent.
In traditional FPV Cell towers represent a large amount of noise and harmonics spewing out over your video signal. However, using GSM technology to stream FPV you can work with your local cell tower rather than against it. Using this system you can conceivably fly further than even the best UAF systems.
The signal sent is digital which increases latency so it would not be appropriate for Miniquads, only larger AP Drones and incredibly long range FPV flights. Currently the main commercial option for this kind of technology is SkyDrone, which offers various products which allow for video to be sent across cell towers. The big advantage with using the Sky Drone option is that they have managed to optimize their System to limit latency.
There are two main issues which this kind of video and (in some cases) control technology:
- Reliability, many of the DIY options for this kind of technology use non optimised codecs for links with Cellular towers, causing transmission errors. Sky Drone have managed to solve this problem by creating auto-bandwidth software which changes frame rate and resolution in order to keep video as smooth as possible
- Second issue is fire walls. The majority of net carriers often only provide heavily firewalled 3G/4G connections which causes high latency on the digital signal. A way in which this has been overcome by Sky Drone is by “punching a hole” through the fire wall and having a direct connection between the Drone and the ground station. So in essence it is piggy backing on the net work rather than trying go through it.
Of course this is a highly specialist area and does not represent a massive leap forward in FPV, but rather the utilization of existing technology in order to service a specific need. As far as we are concern this kind of technology would be brilliant for very long range FPV in planes or as a way to reliably control AP Drones from a ground stations. The main issues of size, expense, latency and unreliability mean that this kind of technology is only really useful for Large UAVs with advanced Flight controllers with autopilot capabilities so that if there was an issue, you had a back up as well.
It is also worth noting that you will likely need two 4G LTE SIM cards in order to use this technology. With good HD quality video at 30 FPS you will be using around 5GB per hour of footage, meaning you will need an incredibly expensive plan.
Specs of the SkyDrone
|Weight:||165 gr (incl. lens)|
|Dimensions:||83.9 x 56.4 x 56 mm|
|Autopilot Connection:||MAVLink compatible autopilot; optimized for APM and PixHawk from 3D Robotics|
|Video Resolution:||Full HD (up to 1080p 30fps)|
|Lens:||Interchangeable, default: 112° FOV|
|Antenna:||Internal antenna + external antenna socket (optional)|
|Typical End-to-End Latency:||<150ms|
|Network Compatibility:||Option 1 (Global): LTE Bands: 1,2,3,4,5,7,8,13,17,18,19,20 3G: 1,2,4,5,8 (Modem Carrier Cert: AT&T, Verizon, Vodafone, Telefonica, Orange)
Option 2 (Japan): LTE Bands: 1,3,8,9,11,18,19,21,26 3G: 1,6,8,9,11 (Modem Carrier Cert: KDDI, NTT DoCoMo)
|Gimbal Compatibility:||Gimbal connected to an MAVLink compatible flight controller can be controlled via Oculus Rift Head-Tracker|
Though this is unlikely to apply to the majority of you guys, it is still an interesting concept. It will be interesting to see whether the technology moves from just DIY work by Engineers to a more consumer and hobby based buyer. There are definitely many applications for people looking to longer range FPV the decider will be whether there is enough interest to make it worth while improving the technology.