What Is Cellular Bonding?
While many of us are excited by the possibilities of 5G today, those possibilities are mostly unrealized at present. In terms of streaming, the higher data rates promised with 5G are not as important as the new bands/frequencies that enable a given tower to serve more active connections, as opposed to being completely blocked up with hundreds of other cellular users trying to push or pull data at the same time.
In the absence of widespread high-bandwidth 5G availability, one technology that has been with us for several years remains the most reliable solution for getting your streams to the cloud when Wi-Fi and wired connections are unavailable: cellular bonding.
How Bonding Works
For those unfamiliar, cellular bonding is a process whereby multiple connections to the internet—e.g., multiple cellular carriers—are all used, simultaneously, to send and receive data. This is very different than having multiple Wide Area Networks (WANs), or a Multi WAN router that can use multiple internet connections at the same time, and even switch between them and offer some failover. Sending office VoIP traffic over one WAN connection and customer internet browsing over a second WAN connection is not bonding. Having one primary connection and then having everything drop to a second connection if the first one fails is also not bonding.
Bonded video service is a process that involves two distinct halves of a whole. Most often, it is a paid service, and there is a key reason why true bonding has a price—the “de-bonding” service, as I call it.
Regardless of who manufactures or brands your bonding router, each of these devices is designed to connect to a server and service at the other end. Let me explain why.
When your video come in to the device, it is compressed into data and that data is sliced into packets (1, 2, 3, 4, 5, 6, 7, 8…). Each of these packets is distributed, or should be distributed, like playing cards at the poker table—a packet to each connected service. Each of those packets is numbered. Each of those packets flies through the air or a wire, or bounces around the city or the country, and gets to the receiving server. Whether it be Teradek Core, LiveU server, Kiloview Server, or Peplink’s SpeedFusion, there’s a very specific destination these packets go to. Because each route they take is different, they will never, ever, arrive in order.
You cannot “bond” your video directly to YouTube, or Facebook, or Vimeo, or to any delivery network.
These de-bonding servers exist to receive the packets, in whatever order—e.g., 3, 1, 4, 2, 6, 5, 7, 8...—as they arrive by postal carrier, UPS, carrier pigeon, tied to a rock, etc. The server “de-bonds” the packets and puts them back in order—1, 2, 3, 4, 5, 6, 7, 8—and then, with a single reliable fast connection to the internet, it then sends this single stream of data to your intended destination. This de-bonding service requires a computer to dedicate itself to processing your video, and renting someone else’s computer for this task has a cost.
Sometimes, these servers also serve to transcode the stream. For instance, if the bonding device sends HEVC/H.265 to be very efficient with data, but not all Content Delivery Networks (CDNs) accept HEVC directly, so then the de-bonding server takes the received HEVC and transcodes it to MPEG-4/H.264 to send along to the destinations. This transcoding process takes server horsepower, and thus often adds cost to the service.
Another common feature of the de-bonding server is multicasting: You send one stream from your bonding device to the server, at which point the server splits it and casts it to multiple destinations. This again saves on data being sent because you’re only sending one stream, but it can end up going to 2, 5, 10, or more different destinations. Again, this process of splitting, managing connections, and sending out all these streams is, in almost all cases, part of a paid service.
So if you have a bonding device, and you’re not paying for any (de)bonding service, it may behoove you to look more closely into whether you are truly getting bonding service, or whether there’s something else going on.
You’ll note a recurring theme: using as little data as possible to send the stream. Whether it be leveraging HEVC or cloud multicasting, there is a key reason why these tools strive for the lowest-bitrate data stream possible. That reason is reliability. Consider sending one 3 Mbps HEVC stream that is multicast to 10 destinations, versus directly sending ten individual outbound 6 Mbps H.264 streams. This would require 60 Mbps up just for the streams, and typically, with data fluctuations, you want double what the stream will be, so you’d really need at least 120 Mbps upload bandwidth to handle that first mile reliably.
For businesses, or broadcast, there’s often the money to make sure that big bandwidth is there, and probably a lot more to spare. For the vast majority of independent producers and freelancers, however, that big bandwidth is often not there. Plus, when using cellular connections—especially when moving—you are at the mercy of what the towers can provide to you (and everyone connected to the same towers) at that second. So, the lower you can keep your demand, the less you will need to push through, and the more likely it is that all your packets will make it through in a timely manner.
With more and more content being streamed every day, everyone from broadcast producers to individual users has a vast assortment of technologies available to them to ensure that their program can reliably reach viewers around the world. They key comes in understanding what bonding truly is, how it is implemented, and what advantages it offers. Then you can compare services more thoroughly and understand what it is you need, and how the services you are comparing satisfy your needs. You can also assess whether new technologies like SRT or buffering can provide the reliability you want, without the need for true bonding services.
Any way you look at it, thanks to bonding and some of the other technologies discussed here, the days of sending glitchy, buffering video to viewers should be behind us. There are many options available, at a wide variety of price points, to ensure that streaming professionals can deliver the best video possible.
In this interview from NAB 2019, LiveU clears up some of the hype around 5G (it won't make an impact until next year and there's no special health risk) and talks about the testing it's already doing with 5G modems.
LiveU's George Klippel describes the cellular bonding workflow for live production from capture to distribution using LiveU's technology in this clip from Streaming Media East 2018.
ReelSolver's Tim Siglin and BlueFrame's Chris Knowlton discuss the challenges of streaming via cellular bonding in congested cell-traffic areas, and strategies for boosting performance, in this panel discussion from Streaming Media West 2017.
If you take the steps and follow the guidelines provided within this article, you should be able to configure, acquire, and use a functional cellular bonding system that reliably gets your signal up to the cloud, and to multiple output points from there.
Cellmux technology is now an essential part of many video operations, but connections aren't always reliable. Here's how to ensure strong uptime and throughput.
Companies and Suppliers Mentioned