bitpro가 원문에서 글자만 모은 내용이에요. 표·그림·수식은 빠지거나 깨질 수 있어요. 전체 5쪽 중 1쪽
A Decentralized Video Delivery and Streaming Network
Powered by a New Blockchain
Last Updated: Nov 11, 2018
Version 2.0
Abstract
This whitepaper introduces the Theta Network, a new blockchain and token as the incentive
mechanism for a decentralized video streaming and delivery network.
The Theta Network and protocol solves various challenges the video streaming industry faces
today. First, tokens on the Theta blockchain are used as an incentive to encourage individual
users to share their redundant computing and bandwidth resources as caching or relay nodes
for video streams. This improves the quality of stream delivery and solves the “last-mile”
delivery problem, the main bottleneck for traditional content delivery pipelines, especially for
high resolution high bitrate 4k, 8k and next generation streams. Second, with sufficient
network density the majority of viewers will pull streams from peering caching nodes, allowing
video platforms to significantly reduce content delivery network (CDN) costs. More importantly,
by introducing tokens as an end-user incentive mechanism the Theta Network allows video
platforms to deepen viewer engagement, drive incremental revenues, and differentiate their
content and viewing experience from their competitors.
The Theta blockchain introduces three main novel concepts:
● Multi-Level BFT: A modified BFT consensus mechanism which allows thousands of
nodes to participate in the consensus process, while still supporting very high
transaction throughput (1,000+ TPS). The core idea is to have a small set of nodes,
which form the validator committee, produce a chain of blocks as fast as possible
using a PBFT-like process. Then, the thousands of consensus participants, called
guardians, finalize the chain generated by the validator committee at regular
checkpoint blocks. The name multi-level BFT consensus mechanism reflects the fact
that the validator/guardian division provides multiple levels of security guarantee. The
validator committee provides the first level of consensus — with 10 to 20 validators,
the committee can come to consensus quickly. The guardian pool forms the second
line of defense. With thousands of nodes, it is substantially more difficult for attackers
to compromise the integrity of the network, and thus provides a much higher level of
security. We believe this mechanism achieves a good balance among transaction
throughput, consistency, and level of decentralization, the three pillars of the so-called
“impossible triangle”.
● Aggregated Signature Gossip Scheme: A basic all-to-all broadcasting of the
checkpoint block hash could work between guardian nodes, but it yields quadratic
communication overhead, and therefore cannot scale to 1,000+ nodes. Instead, we
propose an Aggregated Signature Gossip Scheme which significantly reduces
messaging complexity. Each guardian node keeps combining the partially aggregated
signatures from all its neighbors, and then gossips out the aggregated signature. This
way the signature share of each node can reach other nodes at an exponential rate,
2
leveraging the gossip protocol. In addition, the signature aggregation keeps the size of
the node-to-node messages small, and thus further reduces the communication
overhead.
● Resource Oriented Micropayment Pool: An off-chain “Resource Oriented
Micropayment Pool” that is purpose-built for video streaming. It allows a user to
create an off-chain micropayment pool that any other user can withdraw from using
off-chain transactions, and is double-spend resistant. It is much more flexible
compared to off-chain payment channels.
This white paper will describe these concepts and the Theta blockchain in detail. The Theta
Network launched with ERC20-compliant tokens and were integrated into the SLIVER.tv
platform in December 2017. The Theta blockchain mainnet code has been released, and the
first live mainnet implementation is planned to launch on March 15, 2019, at which time each
ERC20 Theta token will be exchanged 1:1 for native Theta tokens.
3
T A B L E O F C O N T E N T S
Vision 5
Introduction 5
Video Streaming Market 5
Video Streaming Challenges 6
Background 7
Opportunity 9
Theta Mesh Delivery Network 11
Geo-Optimized Tracker Server 12
Intelligent Player Client 13
Theta Blockchain Ledger 14
The Consensus Mechanism 15
Multi-Level BFT 15
System Model 17
The Block Settlement Process 17
Block Proposal 18
Block Consensus Among Validators 19
Analysis 20
The Block Finalization Process 21
Scaling to Thousands of Guardians 22
Analysis 24
Reward and Penalty for Validators and Guardians 25
Turing-Complete Smart Contract Support 26
Off-Chain Micropayment Support 27
Resource Oriented Micropayment Pool 27
Double Spending Detection and Penalty Analysis 30
Ledger Storage System 32
Storage Microservice Architecture 32
History Pruning 32
State Synchronization 34
A Dual Currency System and Token Mechanics 34
Future Work 36
Founding & Advisory Team 37
4
Vision
Introduction
Video Streaming Market
Live video streaming accounts for over two-thirds of all internet traffic today, and it is expected
to jump to 82% by 2020, according to Cisco’s June 2016 Visual Networking Index report. In the 1
US, millennials between the ages of 18 and 34 are driving the growth of video streaming, and
are heavy users of services like Netflix, Youtube, and HBO. Streaming video among this group
has jumped 256% from an average of 1.6 hours per week to 5.7 hours per week according to a
SSRS Media and Technology survey, and mobile devices are leading the charge in video
consumption growing 44% in 2015 and 35% in 2016. The top five video streaming players in 2
the US are Facebook, Google/Youtube, Twitter and related properties, Live.ly and Twitch.
Figure 1 . Global IP video traffic growth
At the same time, global virtual reality (VR) traffic including 360° video streaming content is
estimated to grow 61-fold by 2020, at a staggering 127% CAGR according to the same Cisco
report.
1
https://www.cisco.com/c/dam/global/ko_kr/assets/pdf/2016-VNI-Complete-Forecast-PT.pdf
2
http://www.zenithmedia.com/mobile-drive-19-8-increase-online-video-consumption-2016/
5
Figure 2 . Global virtual reality traffic growth 3
Video Streaming Challenges
Content Delivery Networks (CDN) play an important role in the video streaming ecosystem. It
provides the backbone infrastructure to deliver the video streams to end viewers. One major
limitation of today’s CDN networks is the so-called “last-mile” delivery problem . Typically, CDN
providers build data centers called Point-of-Presence (POPs) in many locations around the
world, with the expectation that these POPs are geographically close to the viewers. However,
the number of POPs are limited, hence are not near enough to the majority of viewers,
especially in less developed regions. This “last-mile” link is usually the bottleneck of today’s
streaming delivery pipeline and often leads to less optimal user experience including choppy
streams, bad picture quality, and frequent rebuffering.
To streaming sites and content platforms, another major concern is the CDN bandwidth cost.
For popular sites, the CDN bandwidth cost can easily reach tens of millions of dollars per year.
Even if platforms own proprietary CDNs, maintenance costs are often high.
These issues are becoming even more prominent with the coming era of 4K, 8k, 360° VR
streaming, and upcoming technologies such as light field streaming. Table 1 compares the
bandwidth requirements of today’s mainstream 720p/HD streams vs 4K, 360° VR and future
lightfield streams, quickly jumping by orders of magnitude.
3 https://www.cisco.com/c/dam/global/ko_kr/assets/pdf/2016-VNI-Complete-Forecast-PT.pdf
6
Standard Resolution Bandwidth / Mbps Magnitude
720p HD 1080x720 5 to 7.5 1x
1080p HD 1920x1080 8 to 12 1.6x
4K UHD 3920x2160 32 to 48 6.4x
8K 360° VR 7840x4320 128 to 192 25x
16K 360° VR 15680x8640 512 to 768 100x
Lightfield --- 5000+ 1000x
Table 1 . Bandwidth comparison: today’s 720p/1080p video vs 4K and 360° VR streaming,
vs future volumetric/lightfield streaming
To solve the VR and light field video delivery problem, the industry has started to explore
“ foveated streaming ” technology. Instead of streaming the entire video in full resolution, this
technology reduces the image quality in the peripheral vision (outside of the zone gazed by the
fovea) in order to reduce bandwidth requirements. As the viewer turns his or her head to look at
a different direction, the system adapts the spatial video resolution accordingly by fetching the
high resolution video packets for the viewing direction from the server. For the foveated
streaming technology to work well in practice, the round-trip time between the server and the
viewer has to be small enough. For viewers that are geographically further from the CDN POPs,
their VR stream viewing experience is compromised even with foveated streaming technology.
Background
SLIVER.tv (the “company”) has been at the forefront of developing next-generation video
streaming technologies for VR and spherical 360° video streams since 2015, and is the parent
company to Theta Labs, Inc.. SLIVER.tv has raised over $17 Million in venture financing from
notable Silicon Valley VCs including Danhua Capital, DCM, Sierra ventures, leading Hollywood
media investors including Creative Artists Agency, BDMI, Advancit Capital, Greycroft Gaming
Track Fund, and marquee corporate investors including GREE, Colopl, Samsung Next and Sony
Innovation funds. Additionally, the company has strong Chinese investors and partners
including Heuristic Capital Partners, ZP Capital, Green Pine Capital Partners, and Sparkland.
In a technology derived from “foveated streaming” SLIVER.tv’s most recent technology patent
granted #9,998,664 , “METHODS AND SYSTEMS FOR NON-CONCENTRIC SPHERICAL
PROJECTION FOR MULTI-RESOLUTION VIEW” , specifically addresses the problem of 4
generating highly efficient spherical videos for virtual reality (VR) streaming, highlight, and
replay. The technology performs non-concentric spherical projection to derive high resolution
displays of selected important game actions concurrently with lower resolution displays of
static game environments, thus optimizing tradeoff between visual fidelity and data transfer
load.
SLIVER.tv today is the leading next-generation live esports streaming platform with over six
million unique visits in July 2018, with a vision to transform the esports engagement
experience. As video gaming has grown in popularity to become a $40+ billion market, bigger
than Hollywood and Bollywood combined, the rise of multiplayer competitive gaming as a
4
https://patents.google.com/patent/US9998664B1
7
spectator sport has become a major new industry, dubbed esports . Esports is a global
phenomenon with major tournaments and major pockets of fans and competitive teams in
Europe, Asia and North America. The online gaming and esports ecosystems have exploded
over the past five years.
A recent 2017 SuperData research put the combined audience for gaming video content on 5
YouTube and Twitch at 665 million, more than twice the US population. This surpasses the
viewership of 227 million for HBO and Netflix combined. Today, esports and gaming video
content account for a significant portion of all video content streamed over the Internet.
SLIVER.tv additional core patents and technology focus on various applications of cutting edge
live streaming to esports content. The company’s US Patent #9,573,062 “ METHODS AND
SYSTEMS FOR VIRTUAL REALITY STREAMING AND REPLAY OF COMPUTER VIDEO GAMES ” , US 6
Patent #9,473,758 “ METHODS AND SYSTEMS FOR GAME VIDEO RECORDING AND VIRTUAL
REALITY REPLAY ” and US Patent #9,782,678 “ METHODS AND SYSTEMS FOR COMPUTER VIDEO 7
GAME STREAMING, HIGHLIGHT, AND REPLAY ” pioneer the capture and live rendering of popular 8
PC esports games including League of Legends, Dota2 and Counter-Strike: Global Offensive in a
fully immersive 360° VR spherical video stream, effectively placing the viewer and audience
inside the 3D game through a live video stream, and rendering 360° highlights, replays and
special effects in real-time.
Since launching in 2016, SLIVER.tv has broadcast numerous global esports tournaments in
360º VR in partnership with premier brands including ESL One, DreamHack and Intel Extreme
Masters. At key events in the US and Europe, SLIVER.tv has live streamed top esports games to
millions of fans of Counter-Strike: Global Offensive (CS:GO) and League of Legends (LoL) . 9
SLIVER.tv launched its Watch & Win esports platform in July 2017 and the first virtual token
designed around esports content streaming and fan engagement. Since launch, the company
has attracted millions of esports fans circulating over 1 Billion virtual tokens by actively
participating and engaging with live esports matches. These users viewed over 50 million
minutes of live esports streaming, nearly 100 years worth of content in the first few weeks of
launch. This positions the company as one of the largest esports streaming sites built around a
virtual community today. 10
The SLIVER.tv platform is continuing to expand quickly driven by word-of-mouth, referral and
social channels.
5
https://www.superdataresearch.com/market-data/gaming-video-content/
6
https://www.google.com/patents/US9573062
7
https://www.google.com/patents/US9473758
8
https://patents.google.com/patent/US9782678B2
9
https://venturebeat.com/2017/03/16/polands-intel-extreme-masters-was-biggest-event-in-esports-history/
10
https://www.sliver.tv/press
8
Figure 3 . Total visits on desktop and mobile web, in the last 6 months
Opportunity
The company’s mission is to leverage blockchain technology to create the first Decentralized
Video Streaming and Delivery Network whereby video viewers are incentivized to share
redundant computing and bandwidth resources to address today’s video streaming challenges.
Using the Ethereum EVM “World Computer” metaphor, the Theta Network can be viewed as
the “ World Cache ” formed by the memory and bandwidth resources contributed by viewers.
Specifically, viewers around the globe can contribute their devices as “ caching nodes ” whereby
they form a video delivery mesh network that is responsible for delivering any given video
stream to viewers anywhere around the world optimized for local. The Theta Network can
effectively address the technical challenges discussed in the previous section. First, viewers’
devices are geographically much closer to each other than to the CDN POPs. This reduces
packet round-trip time and improves the stream delivery quality, and thus addresses the
“last-mile” delivery issue. Second, with a sufficient amount of caching nodes, most viewers will
receive the stream from caching nodes, which will help streaming sites reduce their CDN
bandwidth cost. Third, caching nodes also reduce round-trip time making foveated and next
generation streaming technology practical.
To encourage viewers to contribute their computing and bandwidth resources, we introduce the
Theta token as an incentive mechanism. Caching nodes can earn tokens as they relay video
streams to other viewers. Not only does the Theta Token motivate viewers to join the network
as caching nodes, it also greatly improves the streaming market efficiency by streamlining the video
delivery process . We will discuss later in the paper, but within the Theta Network, advertisers
can also directly target viewers at a lower cost, viewers earn Theta Tokens for their attention
and engagement with their favorite content, and influencers earn Theta Token as gifts directly
from viewers. More interestingly, streaming and content platforms can open up new and
incremental revenue opportunities with Theta.
The full launch of the Theta protocol introduces a new blockchain and a native token structure
where:
9
● Caching nodes earn tokens for caching and relaying video streams to other viewers
● Viewers optionally earn tokens from advertisers as engagement rewards, and can in
turn gift to favorite influencers and content creators
● Streaming sites and platforms can drive incremental new revenues through sales of
premium goods and services, and deepen user engagement through Theta
● Advertisers fund advertising campaigns with tokens to support influencers, streaming
sites and viewers
● Streaming sites and platforms can offload up to 80% of CDN costs
The Theta protocol builds upon the following novel concepts:
● Multi-Level BFT: A modified BFT consensus mechanism which allows thousands of
nodes to participate in the consensus process, while still supporting very high
transaction throughput (1,000+ TPS). The core idea is to have a small set of nodes,
which forms the validator committee, to produce a chain of blocks as fast as possible
using a PBFT-like process. Then, the thousands of consensus participants, called the
guardians, can finalize the chain generated by the validator committee at regular
checkpoint blocks. The name multi-level BFT consensus mechanism reflects the fact
that the validator/guardian division provides multiple levels of security guarantee. The
validator committee provides the first level of protection — with 10 to 20 validators,
the committee can come to consensus quickly. The guardian pool forms the second
line of defense. With thousands of nodes, it is substantially more difficult for attackers
to compromise, and thus provides a much higher level of security. We believe this
mechanism achieves a good balance among transaction throughput, consistency, and
level of decentralization, the three pillars of the so-called “impossible triangle”.
● Aggregated Signature Gossip Scheme: A naive all-to-all broadcasting of the
checkpoint block hash could work between guardian nodes, but it yields quadratic
communication overhead, and so cannot scale to 1,000+ nodes. Instead we propose
an Aggregated Signature Gossip Scheme which could significantly reduce messaging
complexity. Each guardian node keeps combining the partially aggregated signatures
from all its neighbors, and then gossips out the aggregated signature. This way the
signature share of each node can reach other nodes at exponential speed thanks to
the gossip protocol. In addition, the signature aggregation keeps the size of the
node-to-node messages small, and thus further reduces the communication
overhead.
● Resource Oriented Micropayment Pool: An off-chain “Resource Oriented
Micropayment Pool” that is purpose-built for video streaming. It allows a user to
create an off-chain micropayment pool that any other user can withdraw from using
off-chain transactions, and is double-spend resistant. It is much more flexible
compared to off-chain payment channels. In particular, for the video streaming use
case, it allows a viewer to pay for video content pulled from multiple caching nodes
without on-chain transactions. By replacing on-chain transactions with off-chain
payments, the built-in “Resource Oriented Micropayment Pool” significantly improves
the scalability of the blockchain.
10
Theta Mesh Delivery Network
Peer-to-peer streaming focuses on timely delivery of audio and video content under strict,
near real-time parameters. Peer-to-peer livestream delivery works best when many people
tune in for the same stream at the same time. High concurrent user count means more
peering resources are available, and thus the peer nodes can pull the stream from each other
more effectively. The whole system capacity increases as more peer nodes become available.
Moreover, robustness of the system is increased in a peer-to-peer network, as nodes do not
need to rely on a centralized server to retrieve content. This is espec프로젝트로 돌아가기