Play2Live shows how to use P2P CDN for superior streaming results
Online videos, video on demand (VOD), Over the Top (OTT) services, video streaming — these are the new buzzwords as we have shifted to a “video first” world. Cisco predicts that by 2021 a million minutes of video content will cross the network every second! It’s said that more video content is uploaded in 30 days than the total content created by US television networks over the past 30 years.
And gaming and eSports are very much part of this worldwide trend. It is estimated that 9 billion hours will be dedicated to watching professional gaming by 2021.
Play2Live (P2L) is a start-up that aims to be a big player in the gaming and eSports market. It offers a decentralized streaming platform for streamers, gamers and eSports fans. It has added blockchain technology and monetization options that are unique in the streaming world and has moved further than anyone else to provide a great user experience.
To better understand just how unique this P2L offering is, it may be useful to look at the broader context of entertainment in the digital era, at some of the technical challenges that must be addressed, and at what the competition is doing.
The phenomenon of “fan-centricity”
A recent PwC report, “Global entertainment & media outlook 2017–2021” provides some useful insights. A major theme is that any player in the media or entertainment space (and that includes gaming and eSports), “…. simply cannot thrive without the economic, social, and emotional power of fans”. It is not enough to cater for “casual eyeballs” or infrequent users. Clickbait doesn’t work anymore. Brands must develop fan loyalty.
Gaming is not short of fans. The PwC report noted that this highly tech-savvy, mostly-millennial audience tended to spend up to 90% of their free time on their hobby. Seth Ladetsky, senior vice president of Turner Ad Sales, describes them as “a highly-engaged, dedicated millennial fan base”. Rob Lynch of Arby’s, which is a sponsor of an eSports league, notes that “eSports is built on passion and loyalty from fans and athletes”. (Read here to better understand this fan profile.)
Reaching this fan base has become the focus of a growing number of organizations, including Sony, Microsoft, Nintendo and Amazon. Even Google has recently announced that it is working on a game streaming service.
Play2Live may be a small start-up compared to the giants in the game, but it can also compete, provided it can leverage technology to create a great user experience.
The first technical challenge is to meet the demand for speed and immediate gratification. A web page that’s slow, a game that won’t load or a video stream that buffers will be abandoned, as users move on to one that works. In fact, a survey found that nearly 20% of viewers will wait for less than 5 seconds, and another 30% will leave within 10 seconds. Incredible speed is the new normal.
The next technical challenge is to address scale, quality and cost in video streaming. And part of this is about how to address the “last mile” of getting content directly to users.
Video streaming and the last mile — not always good partners
It doesn’t matter how good your content is if you can’t guarantee global delivery, and particularly delivery over the “last mile” to users.
Content producers generally use Content Delivery Network (CDN) providers to deliver video, large files, and other web content to content users. CDNs provide geographic reach through networks of HTTP servers and data centers around the world. They also provide very high bandwidths (“pipes”), that allow for large amounts of data to pass through simultaneously. They use algorithms and increasingly sophisticated software solutions to select the best server to be used for each request for content.
A user experiences fast internet speed when there is high bandwidth (a wide pipe) and low latency (little delay).
Of course, CDN services come at a price. For example, if you want to send a 1Mb file to 10,000 users, you would use 10Gb of bandwidth and have to pay for it.
But even CDNs have limits. And this is especially true for the growing market of video streaming and video-on-demand. In fact, according to the PwC Outlook, 78,7% of all online data will be video by 2021. Back in 2011, VC Marc Andreessen wrote, “Software is eating the internet”. Today it is probably true to say that “Video is eating the internet”.
It’s one thing for a CDN to serve and download a data file. It’s a completely different challenge with a video stream that must be delivered within seconds of it being produced. One of the difficulties with streaming is that you are not downloading a single large video file. Instead, you are downloading segments, each a few seconds long. Your web browser has to request each of these segments. The CDN must direct each request to the nearest node, or PoP, and send the segment to you. The segments go into a storage or buffer system and your video machine then “stitches” them together into a smooth stream. If there is a delay in receiving the segments, we get that dreaded circling symbol to indicate “buffering”.
CDNs can generally cope with these requests, even for thousands of different video streams, provided there are only a few viewers per stream. However, when millions of people want to watch video streaming of a single event, such as the Super Bowl or an eSports world championship, CDNs struggle. Everybody is trying to download the same segments at the same time. This eats up bandwidth, creates buffering issues and leads to dropped connections. Some people have referred to this as “internet gridlock”.
CDNs can extend their infrastructure and hire more bandwidth, but this demands large capital expenditure. Alternatively, a number of CDNs can pool their resources to increase geographic coverage, bandwidth capacity and reliability. However, this places pressure on the telecommunications service providers (Telco’s or TSNs) who transmit the video content over their networks.
So, many Telco’s have also got in on the act and set up their own content delivery networks. They own the network “backbone”, they own the “last mile” to users, and they can better manage their network resources if they are also in control of the content that moves over it. This gives them significant advantage over traditional CDNs, and has added a significant new value-added service income stream to their businesses.
There has been a move also for Telco’s to get together to interconnect their services, in direct competition to the larger CDNs such as Amazon CloudFront, Level3 and Akamai. The so-called Operator Carrier Exchange (OCX) was set up in 2011 for this purpose.
Another solution is decentralized distribution of content through peer-to-peer CDNs (P2P CDNs). The traditional CDN is a client-server model. Your computer or smart phone “client” receives data from a central server. In P2P, only one client receives data from the server, and then shares it with a network of other clients in the same vicinity, who pool their own bandwidth resources. This is particularly powerful for video streaming. Instead of “eating up bandwidth”, the more viewers there are, the more bandwidth there is. Individual clients can hold different segments of the video and share them with each other.
The different models can be depicted like this:
Does P2P CDN actually work?
P2P CDN is a relatively new technology, and there are debates about whether it is a viable solution. But there are some examples of where it is being used.
An important player is BitTorrent, a CDN company founded in 2004. It developed a peer-to-peer protocol, or set of rules, for uploading and downloading files. For example, the owner of a game loads it onto his/her own computer in a file called a torrent. This computer, with the full game on it, is called a seed. The file is cut up into a number of pieces. People wanting a copy request it from the seed, using a program called BitTorrent client. Each person will be sent just a piece of the file and will get the remaining pieces from others who have also requested the game. So, at any given moment, each computer is downloading parts of the file from others, and uploading different parts to them. These computers, working together, are called a swarm. The more computers there are in the swarm, the faster and more efficient it becomes. A centralized computer acts as a tracker, to keep track of where all the pieces are. A decentralized version is also available, where the P2P network keeps track of its own pieces.
There are rules of good behavior. It is considered very bad form to simply download what you want and then leave the network without allowing others time to use what you have downloaded. Equally, you are expected to obey the rules of not downloading copyright-protected materials. Your own ISP may also have rules about the amount of data you can download, especially during peak time.
There are a variety of “torrent” products on the market, many of them free.
· The basic BitTorrent is free, but very advertisement-heavy, unless you upgrade to the Pro version. Market share was about 18% of all internet traffic in Europe in 2013, but this has dropped as YouTube, Netflix and Facebook have taken over much of the market.
· In 2016, BitTorrent acquired uTorrent, which was at the time a very popular open source programme used by game, music and movie fans around the world. BitTorrent turned it into a closed-source program, and added advertising to it. This made it very unpopular with fans.
· qBittorrent has a simple interface, with no ads
· Taxati has encryption and security features, lightning speed, plus a social media chat feature
· Transmission is available for Mac, Linux and Windows
· Deluge is available for most platforms, is very customizable and has the facility for a variety of plugins
In 2007, Internap (now known as INAP) launched P2P alongside its CDN service. It worked together with a video booster developed by Pando Networks. Pando Networks, described as “a peer-to-peer media distribution company” went on to develop specialized applications for game and software publishers, and was used by the NBC television network in the USA for their HD TV streaming. In 2013, Pando was bought by Microsoft. It is believed that this was to use the technology for the streaming features of Xbox. It has also been used to speed up the delivery of Windows updates across local networks. (Symantec has a similar P2P system for its software updates.)
According to a Reddit post, the original software was hijacked by malicious authors, and used to deliver viruses into various games. Riot, producers of the game League of Legends, recommended that all users remove the Pando software and promised to develop their own P2P technology, but it seems that this did not happen.
Some other vendors are offering a P2P layer for last mile delivery.
· Peer5 claims that it solves the two biggest problems of streaming: scalability and geographic coverage, and “turns viewers into a P2P network that offloads up to 98% of CDN bandwidth”. The goal of the company is to develop the world’s largest CDN, without deploying a single HTTP server.
· Swarmify grew out of the founder’s frustration with buffering and stall-outs in videos. It is pitched as the replacement for YouTube, offering users a clean, Netflix-like quality streaming experience, without the advertisements and distractions that pull viewers away from your site to others. It promises “crowdsourcing” for content, and is said to be 8x as reliable as the largest CDNs.
· Streamroot video acceleration technology has been used by a number of companies, including Eurosport for live streaming of sports events, Russia Today for its television channel and Dailymotion for its 300-million video-user platform. Using the P2P model means that Eurosport, for example, does not have to download all videos from its own servers — saving bandwidth and easily handling big spikes in demand. Dailymotion has saved huge infrastructure costs, being able to offload of up to 93% of the traffic, and using 20% less server capacity and bandwidth. As importantly, there has been 20% increase in average play time and user engagement.
Streamroot powers over 400 million video sessions per month and supports MPEG-DASH, HLS and Smooth Streaming. It is a hybrid model, dynamically switching between peers and the central CDN, depending on availability and urgency. According to Rodi, one of the Founders, “The Streamroot tracker uses GeoIP data, ISP and some network topology elements to connect the best candidates together.”
· BemTV is an open-source P2P streaming project, based on WebRTC browsing technology. Streamroot acquired BemTV in 2016, and has retained the open-source format, thus providing a broader all-in-one solution for its users.
· Viblast is another company founded on frustration — this time on poor video streaming of the soccer world cup. It targets broadcasters, promising 2x better video quality and 70% more efficiency for both web and mobile audiences. An image from their website shows the saving in bandwidth, and, therefore in costs, for broadcasters:
Play2Live, P2P and “fan-centricity”
If the requirement today is to leverage technology to create a great user experience, and to develop fan loyalty, then Play2Live (P2L) has done this and so much more.
Play2Live combines its own technology with strategic partnerships and a generous profit-sharing model to deliver a top class “fan-centric” streaming platform, that is unique in the marketplace.
Play2Live will implement a P2P CDN system to address speed, quality, scale and cost of the gaming experience. The P2P system is supported by a partnership with G-Core Labs, a CDN whose infrastructure has been specifically set up for gaming (hence the G in G-Core). It has records for the highest numbers of players online simultaneously as well as recognition for delivering the fastest CDN service to 12 countries in Europe, Russia and Asia.
Beyond this, however, what makes Play2Live different is that it is a decentralized streaming platform, based on blockchain technology.
· Most importantly, blockchain allows for monetization of all aspects of gaming and for all participants in the ecosystem to earn money. This includes users/viewers. There are 15 different ways to generate revenue for the platform — and 11 of these are shared. The internal token, the LUC, is the unit of payment for all services on the platform.
· This technology allows for a full ecosystem for gaming, where streamers, fans, tournament organizers, suppliers of games and skins, bookmakers and betting agencies can interact seamlessly and directly with each other, without leaving the platform. The whole platform is, in effect, a peer-to-peer system, offering interactivity not possible on traditional streaming platforms.
· All parties can control exactly what they want to see or participate in. They can influence tournament formats and issue and accept challenges.
· Advertisements are switched off by default, and users are rewarded if they choose to watch them.
· There is a crowdfunding system for tournaments, prize pools or broadcasting costs for streamers, and an escrow system for guaranteed pay-outs.
An absolutely unique feature of the Play2Live platform is that it will reward those who become part of the P2P network and share their computer resources with others. They will earn LUC tokens, which they can use to interact on the platform, or they can exchange for crypto- or fiat currency. This is illustrated below:
Play2Live really is a step ahead of the competitors, no matter how big they are. P2P CDN has not simply been clever technology to ensure a high-quality streaming experience. Instead, it has become part of the whole Play2Live philosophy of decentralization and profit sharing. Play2Live has used it as yet another way to attract and incentivize a very loyal fan base.