Here’s a statistic to blow you away: video will account for 82% of all Internet traffic by 2022, up from 75% in 2017. Considering Internet traffic is expected to reach 4.8 zettabyte, that’s just under 4 zettabyte for video alone. One zettabyte is a thousand exabyte, which is a thousand petabyte, which is a thousand terabyte. Let that sink in.
However, this doesn’t mean that the Internet is mostly video. There are millions and millions of text websites, pictures, and documents floating around online. We’re talking about Internet traffic here. Most static websites, pics, and docs require very little bandwidth to function properly. Video, instead, accounts for such a large portion of Internet traffic because it requires a significant amount of bandwidth.
This trend will only increase as the quality of the videos that we upload improves. 4K, 8K, and 360° videos are only a few of the technologies that are now becoming commonplace and that are in large part responsible for the percentage growth of video in all Internet traffic.
As part of video, Virtual Reality and Augmented Reality, while currently still in their early stages, are expected to keep on asking for more bandwidth too. Cisco predicts that both VR and AR to grow at a compounded rate of 65 percent every year until 2022.
Live video streaming is another technology that’s growing rapidly. Popular video game streaming platform Twitch often sees live streams with millions of people watching, and businesses are increasingly using Instagram and Facebook Live to connect with their audience.
Add this to YouTube, Netflix, Vimeo, Hulu, and, let’s not be shy about it, all the popular porn websites, and it should be somewhat less as a surprise that video traffic will account for 82% of all Internet traffic by 2022.
Problems With Video Traffic
I live in a relatively rural area of a densely populated European country. My Internet access on all fronts (mobile data, wifi, cable) is not very good, to say the least. The highest quality of video I can watch without the video buffering every other second is 720p, and uploading videos online takes a long time and cripples the entire network.
This isn’t just my problem. People around the world can’t keep up with the high demands of the latest video technologies, and suffer from low video quality, video stutters, or frequent video buffering. Anyone who’s ever experienced this will know what an incredibly frustrating (albeit first-world) problem this can be.
The reason why many people struggle with these problems has much to do with the backbone infrastructure required to deliver these videos: Content Delivery Networks (CDN). CDNs are systems of distributed servers responsible for delivering web content from its origin to the end user. They’re expected to carry 72% of all Internet traffic by 2022, because they’re an efficient way for content creators and publishing platforms to distribute their content worldwide without having to build expensive databases themselves.
CDNs have data centers around the so-called Points of Presence (POPs) scattered around the world. The problem is that these POPs often aren’t near enough to users. This is called the “last mile problem”, and it’s the reason why my download speed is 15 Mbps while it could technically reach 100 Mbps. The last mile problem is the label for the speed bottleneck in most communication networks.
Solving this problem the traditional way would require CDNs or Internet Service Providers (ISPs) to build many more POPs around the world, as well as upgrade their existing POPs. This is unlikely to happen, because it would be prohibitively expensive.
Poor video quality, video stutters, and frequent buffering are problems for the viewers of video content. But these aren’t the only problems. In today’s environment, content creators are at a disadvantage too. Here’s how the process works: you film your cat derping around and decide to upload it to YouTube, the video publisher. YouTube hosts your video and uses a combination of the Google and Akamai CDN to deliver your cat video to viewers around the world.
As such, content creators have both the video publisher and the CDN between themselves and their viewers, and they’re at the behest of whatever rules either of both parties implements. It’s a centralized and inefficient ecosystem that ultimately results in less revenue for the content creator and less innovation too.
Introducing the Theta Network
The Theta Network allows for decentralized video streaming, powered by blockchain technology. It implements a peer-to-peer, decentralized CDN that can deliver video streams to users around the world, who can earn Theta tokens for sharing their bandwidth to then reward those tokens to content creators. It also reduces centralized CDN costs for video publishers.
Theta was founded by Mitch Liu and Jieyi Long. They wanted a better way for people to watch videos in high quality without any problems, while also being able to interact directly with content creators. On their advisory council, Theta has Steve Chen, co-founder of YouTube, and Justin Kan, co-founder of Twitch. They’re sponsored by Samsung and Sony, among others. So suffice to say their idea has grabbed the attention of high-profile people and companies.
Liu and Long are also the co-founders of SLIVER.tv, a live esports streaming platform that was launched in 2016 and raised $17 million in venture financing from Silicon Valley VCs. It had six million unique visits in July 2018, and is known for its 360° spherical video streams.
SLIVER.tv is the parent company of Theta, and the Theta blockchain was integrated into SLIVER.tv in December 2017. Theta successfully launched their mainnet on the 15th of March 2019, which is when all ERC-20 Theta tokens, then on the Ethereum blockchain, were exchanged 1:1 for native Theta tokens on the Theta blockchain.
How Does Theta Work?
The fundamental premise of Theta is that viewers’ devices are geographically much closer to each other than they are to CDN POPs. As such, if those devices can serve as cache nodes for other devices to pull video content from, we have a network that can improve video stream quality and address the last mile problem of CDNs, while also offloading some of the CDN costs that video platforms pay for.
Here’s how this works technically: Theta has a tracker server that organizes nodes in a spatial database. Whenever a device joins the Theta network, this server records its IP address, latitude and longitude, and a variety of performance variables. This allows viewers to quickly and efficiently retrieve a list of geographically close nodes it can pull video from.
But a device joining the network isn’t the only thing that the tracker server should do. Any device on the network needs to be continuously contributing its bandwidth in order to be useful. This is why each device will have a socket connection with the tracker server to send “heartbeat” signals. If the tracker server doesn’t receive heartbeats from a device for a certain period of time, it will remove the device and update its database.
The device of the viewer will maintain a local cache to buffer downloaded stream data. But if the duration of the cached stream data is less than a certain threshold, it will check Theta’s database to find the closest contributing node and see if they have the desired video stream segment. If not, the device will resort to the centralized CDN.
The Theta Blockchain
Streaming video with blockchain technology is different and more complex than distributing static content on a blockchain. Whereas the transactions on most blockchains are single, defined units, such as one financial transaction or one action in a game, a blockchain transaction for video streaming is a data packet, of which there are many in every single video.
This poses a few challenges. First of all, watching a live video stream with a few thousand concurrent viewers will require the blockchain to process thousands of transactions per second. Consensus on these nodes needs to be lightning-fast. Secondly, all these transactions need to be added to the blockchain, which can quickly take up a lot of storage space on the nodes.
The Theta runs on a modified version of the Proof of Stake (PoS) consensus mechanism. A modified version, because most blockchains on a PoS consensus mechanism only have a selected number of validators. EOS, for example, only has 21 validator nodes. Not only is this very centralized, but Theta also needs as many nodes as possible, because the more nodes, the higher the chance of a node being close to a viewer and thus faster than a CDN.
As such, Theta has a validator committee of ten to twenty validators that can quickly come to a consensus on transactions. But they also have a second pool of validators, which they call “guardians”, who finalize blocks of transactions at regular checkpoint blocks. It’s a multi-level BFT consensus mechanism. Multi-level, because of the separate validators and guardian pools, and BFT, because it relies on the Byzantine Fault Tolerance protocol, which ensures safety as long as two-thirds of the entire validator set acts honestly.
This is only one of the innovations that the Theta team has implemented into its blockchain. They also propose novel techniques to reduce the space that transactions take up in a block, all of which is detailed in the whitepaper on their website.
Theta on SLIVER.tv
Considering the Theta blockchain is already implemented into SLIVER.tv, I had to try it out. Not that there’s much trying out, given how naturally it’s integrated on the video streaming platform. You can choose between any of today’s popular streaming games, from CS:GO to Apex Legends, and watch anyone who happens to be live streaming at that time.
Once you’ve selected a stream, you’ll immediately be contributing your bandwidth and computer resources to the Theta network. You become a node, basically. In return, you receive coins, which you can use to participate in giveaways, win prizes, and earn TFuel, Theta’s cryptocurrency. You can decide to donate TFuel, one of Theta's two cryptocurrencies (the other being THETA, for staking and becoming a validator) to your favorite content creators or sell it at whatever rate it’s going for ($0.01 per token today).
First of all, let’s start by saying that it’s quite fun and addictive in itself to watch other people play a game online. But the Theta blockchain adds a level of interaction that I’ve not seen on platforms such as Twitch. There’s a chat box, questions that allow you to earn coins and tickets, and frequent prizes to be won. Streamers frequently interact with their viewers and it feels like you’re continuously earning something.
Watching some of the streams and participating in some of the questions that streamers frequently ask their audience, I was able to gather 52,000 coins and buy around 90,000 tickets, which I used to participate in two contests: one with an AMD Ryzen GPU giveaway, and another with a Nintendo Switch giveaway. All this in the span of maybe an hour watching someone play Apex Legends.
Don’t go to SLIVER.tv if you want to earn money or win lots of prizes. That’s not the point. The point is to make spectating less of a passive activity; to increase interactivity with content creators, all on a blockchain that’s designed to reduce the problems that many people face with traditional video broadcasting.
I think it’s a great project that will only grow as video streaming and watching video streams becomes a more popular activity. While Theta is currently only implemented on SLIVER.tv, I can see it being implemented on larger video platforms, who will use it to increase interactivity and distinguish themselves from video publishing platforms that are more centralized.
The future is video, but our current infrastructure doesn’t allow viewers to take full advantage of the latest video technologies. Everything is too centralized, and content creators lose out as a result. Theta provides a decentralized video streaming platform that has the ability to solve these problems for both viewers and content creators.