As major cryptocurrencies continue with their attempt to enter the mainstream, one of the core limitations of public blockchains remains unresolved: every transaction requires to be processed by every single node in the network. Unfortunately, in 2018 the inability to handle high-frequence transactions in a safe and economically feasible manner has been among the most ravaging threats for scaling cryptocurrencies.
As of 2018 Bitcoin SegWit (BTC) could still only handle 7 transactions-per-second. With a throughput of only 1MB block size it is often leads to substantial transaction delays and, especially at times of network overloads, high costs per transaction. The issue is even more frustrating, especially when we talk about smaller-size (micro) transactions.
Ethereum faces similar issues: the higher the workload on the network, the more expensive the cost of mining becomes because the amount of energy needed to complete the process increases in total and per transaction. The issue was brought to the light through crypto-kitties ERC20 Token based digital product, who’s popularity slowed-down the network for millions of users.
VISA, for example, handles 4,000 per second, so the potential for scalability is dependent on sufficient speed and security. Major projects, such as BTC, Ethereum (ETH) or Ripple (XRP) or altcoins, such as Railblocks/Nano (XRB), IOTA (MIOTA) need to offer a competing functionality can be massively adopted as a payment medium.
We will discuss the most popular on and off chain scaling-solutions that are currently being developed by the community to tackle this issue. For the purpose of this article we refer to scaling in the context of challenges that must be overcome to make one or another cryptocurrency useful to a global user base of billions of people.
The term “on-chain scaling” is frequently used to exclusively refer to increasing the blockchain capacity by means of bigger blocks.However, in the literal sense of the term, it should refer to any sort of protocol change that improves the network’s capacity at the blockchain layer. These approaches tend to provide at most a linear capacity increase, although some are also scalability improvements.
- Blocksize/block weight increase
- The witness discount of segregated witness
- Smaller size of Schnorr signatures
- Bellare-Neven signature aggregation
- Key aggregation
The term “off-chain scaling” refers to approaches that increase the utility of the network without touching the blockchain, or by making superlinearly efficient use of the blockchain load they produce.
- Batching multiple payments into one transaction
- Virtual payments within the system of a custodian (Tipbots, Coinbase,…)
- Payment Channels/Tumblebit/Lightning Network
- Colored Coins
Generally, off-chain solutions scalability improvements go hand-in-hand with a different trust model and different trade-offs. They often require additional software and complexity over on-chain scaling approaches.
For example, Lightning Network (LN) will require its users to be online when receiving payments while on-chain transactions allow passive receipt. On the other hand, it will easily scale the count of payments between participants, but only scale the number of participants in a limited fashion as becoming a LN user requires on-chain transactions.
On-Chain/Layer 1 Scaling Solutions
Up to 2018 two main methods for scaling have been offered, also called Type 1 and Type 2 solutions. The former attempts to resolve the issue on the main network (on-chain) whereas, the later is dwelling into offloading a certain amount of transaction on a complementary network (off-chain). We examine both methods for Bitcoin & Ethereum below :
On-Chain/Layer 1 – Divide Network into a Bicameral Entity
What if, instead, the network was divided into two sections, which could operate semi-independently?Section A could process one batch of transactions, while Section B processed another batch. This would effectively double the transaction throughput of a blockchain, since our limit is now what can be processed by two nodes at the same time. If we can split a blockchain into many different sections, then we can increase the throughput of a blockchain by many multiples.
On-chain scaling ethereum
Source: Onur Deler via medium.com
This is the insight behind “sharding”, a scaling solution offered by Vitalik’s Ethereum Research group and others. This will break shred main chain into two, both busy with different tasks simultaneously. If you want to learn more about sharding, we recommend this extensive FAQ and this blog post.
On-chain scaling Bitcoin
Bitcoin community has also attempted to make its chain more efficient, which had led to the hard fork in the August of 2017 and creation of Bitcoin Cash (BCH) and Bitcoin with SegWit (BTC).
Bitcoin Cash On-Chain Scaling
Not to get too much into the details, but when Bitcoin Cash forked it offered an 8MN block size, following the belief that the only way of scaling is through on-chain capacity increase.
Results of scaling #bitcoin the right way…
— Armin van Bitcoin ⚡ (@ArminVanBitcoin) March 26, 2018
According to Bitcoin’s core developers, this will protect the integrity and independence of mining process, while preserving security. To the credit of the developers, the transaction cost and speed of BCH remain superb to Bitcoin Segwit ever after the split.
Bitcoin Cash Scalability Limitations
The main criticism lies in the fact that block increase may not solely solve the issue with scalability, as it will require for a block to grow as the popularity increases.
Furthermore, with the block growth, smaller miners might lose chances of competing with large mining facilities that are directly interested in the success of Bitcoin Cash and proof of work concept. Having miners dictate the way technology develops is risky as till now the decisions are made by all stakeholders.
Bitcoin On-Chain Segwit Scaling
Bitcoin Segwit team, on the other hand, is looking to fix things by enabling transactions to be written in the block more effectively, and enabling layer-2 (off-chain). This, therefore, reduces the wait time and capacity of the block without directly increasing its block sizes and enables the main network to interact with off-chain technology.
Bitcoin Segwit On-Chain Scaling Limitations
Ever since the hard fork initiated by bitcoin core (BCH developers) Segwit had experience relatively slow adoption yielding around 30% of transaction made this way, as of March 2018.
Bitcoin Segwit On-Chain Scaling Limitations
Still, a 1MB block size has proven to be insufficient at the times of high workload on the bitcoin blockchain. Low adoption rate, that has been pivoting towards 35% since early 2018 may also lead to Segwit only having a mediocre impact on the transaction speed and costs.
Finally, some worry that a switch to off-chain solutions (that segwit is a first step to) will lead to a concentration of power within Nodes with more economic powers, which consequently contradicts Satoshi Nakamoto initial plans for the decentralization.
Off-Chain/Layer 2 – Complimentary Solutions with Sufficient Crypto-Economic Consensus
The core principle behind off-chain crypto scaling solutions is that on-chain (Layer 1) can serve as a beacon of safety, a judge – that can be called upon anytime involved parties want to either settle the off-chain transactions/contract or dispute them. In this case Layer 1 should see if the logic programmed in Layer 2 has been complied with. In case of wrongdoing, Layer 1 should be able to automatically discover that wrongdoing and punish the malicious user, while compensating and rewarding the victim.
This being said, Layer 2 solutions might not prove as safe as Layer 1 (why we love blockchain so much) but if the development and economic climate support it, off-chain scalability can offer sufficient protection to be relied upon.
Albeit the name, Stated Channels promise to not fully centralize and regulate the off-chain transactions, but offer a help in taking a big chunk of the workload from the main blockchain via offering enough security and economic incentives.
State channels and plasma will always be needed; they have side benefits beyond scaling (eg. low latency)
— Vitalik “Not giving away ETH” Buterin (@VitalikButerin) November 2, 2017
The idea of state channels is not new, and can be useful not just for payments but also for numerous other matters we already do online. We will cover the most promising DAG, Ethereum & Bitcoin projects below.
See how the state channel works by the following example:
Imagine that Jake and Jill both want to play a game of tic tac toe, where the winner receives 1 ethereum each. The simple way to do this would be to use a smart contract and write down each time the player makes a move and each time a transaction is sent to the chain. Depending on the set of predetermined rules, the chain assigns the reward to the winner.
Albeit, working players make the entire network process their puzzle, which therefore drives the cost up for the entire network.
Instead, wouldn’t it be more effective to let Jake & Jill finish the game and only inform the main network once the final result is known?
State-Channel system could help Jake & Jill on a premise that in case Jake or Jill can’t agree on something they can always call up a main network to ask for a resolution based on predefined rules.
First, we create a smart contract “Moderator” on the main-chain that and make sures it understands the rules and reward levels.
Then, Jake and Jill start the game by each making their turn, signing the transaction and both keeping a saved copy for each transaction. While the game continues, zero transactions are actually processed by the Layer 1 network, as Jake and Jill just send the transactions to each other. It’s like writing a cheque, where no money actually changes hands.
Finally, when both are ready to finish the game, the final results are sent to the “Moderator” paying only one on-chain (Layer 1) transaction fee. The moderator makes sure that the final results are legit (a certain period is granted after the first check) and in case of fraud, they take additional time to serve justice.
What happens in case of a fraud attempt?
In case of one party trying to lie (Jake) to a moderator (just a dump contract) that can’t know which “final results” are actually correct, it needs to have time for another party (Jill) to verify that the version of the end result which Jake shared is actually final. In case of fraud is visible, the “Moderator” would penalize the fraudulent side.
Features and limitations
Although state-channels offer many practical and ideological advantages, state-channels also lead to certain risks over using on-chain and therefore skepticism among blockchain developers. Basically, we have two or three camps, who all argue for their solution to solve the scaling problems best. The following limitations are popular as of 2018:
- Availability issue. If Jake’s internet goes down during Layer 2 transaction dispute (for instance, maybe Jill damages it to win the dispute), then Jake might not able to provide proof against Jill, which could lead to Jake being seen as a fraud. Layer 2 enthusiasts argue that in such case Jake can still ask a third party to keep all the info about the dispute and maintain the dispute on Jakes behalf.
- State Channels cost at least two on-chain transactions (to be opened and closed), however when the channel/judge is being setup, both payments and application actions can be performed at minimum costs.
- Layer 2 channels work best when a predetermined number of parties is assigned to each channel, as adding new participants would lead to re-programming the channel/judge.
- Only opening and closing of the channel is public. The rest happens solely between members of the channel. To create a channel, governments might try to regulate the hosts of state channels.
- Instant finality. Ensure that as soon as the action is done in layer-2 its final. The only way as of now to reverse the change is to call On-Chain (Layer 1).
Bitcoin Of-Chain (Layer 2) Solutions
The Bitcoin community started experiencing practical issues with the technology approximately the same time when Hotishi’s kid was suddenly embraced by the vast number of people.
Even before the scalability issues directly damaged the adoption of the the crypto giant, Bitcoin Core and Blockstream, a company founded in 2014 by Adam Back who currently employs most bitcoin core developers, started working on the most talked-about Crypto State Channel (Bitcoin and Litecoin) – Lightning Network.
Lightning Network Discussion
The bitcoin off-chain project run by Adam Back offers to solve the cost and usability issues without compromising any of the aspects that make cryptocurrencies like Bitcoin valuable for societe.
The community has had polarised views, ranging from enthusiasts, who argue that Blocksteam’s vision of decentralised internet via state channels is the future, to skeptics, who argue that Blockstream is planning to undermine decentralisation by consolidating state channels around the wealthy investors.
The Bitcoin core developers have refused to scale the bitcoin network for several years now, which has resulted into skyrocketing fees on the bitcoin network. This has worked in favor of the lightning network, which should offer a better scaling solution. Most transactions will occur outside the blockchain, allowing for lower fees and instant transactions.
There’s a part of the community (and early adopters) who doesn’t agree with the current scaling plan of Bitcoin, which has resulted into a hard fork (split) coin called Bitcoin Cash. Bitcoin Cash plans to scale the bitcoin network by raising the block size and allowing everyone to continue using the blockchain. Blockstream investor emails have leaked, which shows Blockstream motivation to steal transaction fees from miners and pay to Liquid sidechain customers.
Elite CurrenSee is following the matter closely and will keep you updated regarding the situation. As of the beginning of 2018 we believe that it is too early to say weather Lightning Network will be released in a form that will be adopted by the community.
What we can argue now is that, despite its unclear motives, we have faith that relying on Layer 1 will protect the community from future attacks and malicious developers.
Look closely at the presumed dispute between miners and bitcoin core/blockstream developers because this could lead to indications of price changes.
From the Blockstream website we were able to find the following info:
Liquid is a commercial sidechain that combines Confidential Transactions and several other Elements to provide high-speed transactions between Bitcoin Exchanges.
Instead of proof of work (like in Bitcoin), Liquid depends on a distributed group of “signatories”, or functionaries, to create blocks. These functionaries are hardened black boxes that enforce the rules of the network as long as they are connected to the Internet. We have distributed these functionary boxes across the world to a number of participants, known as “the functionary grid”.
Since it went into beta during 2017, the main intention of liquid was to move large chunks of money instantly and without costs related to on-chain transactions.
We can argue that most of the traits of the Bitcoin liquid channels show up in many altcoins and some of the DAG Chains that don’t rely on proof of work. The time will show if the world will work with many cryptos or whether a consolidation will leave us only with the universally-accepted cryptocurrency forms.
Ethereum Of-Chain Solutions
Unlike Bitcoin off-chain channels who are mainly busy with getting microtransactions cheap and comfortable, Ethereum’s game plan as of 2018 is to also offer a platform for developers and enthusiasts where they would be able to program all sorts of operations. For example, you could program a “judge”, who will be making sure all channel members give back the money to the one who brought you all the drinks last night.
Raiden, similarly to Lightning Network, should allow users to open a channel with larger entities that would mitigate the need to create an individual channel each time a user wants to transact with a friend. Hence, promising an interconnectedness is required for the project to make economic sense.
On August 11 2017, Vitalik Buterin and Joseph Poon released a paper titled Plasma: Autonomous Smart Contracts. The paper went on how Ethereum can reach higher number of transactions.
When it comes to security and reliability, Plasma offers the same idea as state channels that revolves around call out to main chain (Layer 1).
But Plasma is set to improve on the idea and moving it in a new direction. This is done by allowing “child-blockchains” to be created on top of the Ethereum main chain. These child-blockchains can, subsequently, create their own child-chains, which can further create their own child-chains, and so on.
The results come to many operations done on side-chains without overloading the main chain too much, as well as keeping the cost of such operations to a minimum.
A Plasma chain can be much faster and charge significantly less for its operations on it, as transactions are not being replicated across the whole chain. See more on Plasma here.
In addition to Counterfactual and Raiden, there are several application-specific channel implementations on Ethereum. For instance, Funfair has built state channels (which they call “Fate channels”) for their decentralized gambling platform and Spankchain has built one-way payment channels for adult performers (they also used a state channel for their ICO), and Horizon Games is using state channels in their first Ethereum-based game.
Directed Acyclic Graph Chains (DAG)
Even before BTC took off, other versions of cryptocurrencies existed but the most beloved characteristic that bitcoin offered was decentralisation. Hence the argument blockchain vs dag with both sides packing enough arguments to rest their case. The tamper-proof and immutable ledger that bitcoin offered was hard to beat and the public embraced it.
Despite losing at the start, DAG chains started gaining traction in 2017, as the scalability of bitcoin-like blockchain started to see the first major bumps on the road.
DAG’s main feature lies in the fact that it doesn’t need the whole chain to process one transaction. The lack of miners in this system gets rid of the blockchains threat of when/if the entity owns over 50 percent of the network and can double spend.
According to Dr. Popov:
“The tangle naturally succeeds the blockchain as its next evolutionary step, and offers features that are required to establish a machine-to-machine micropayment system.”
As far as DAG’s comparison to Ethereum goes, it is not quite as robust as far as smart contracts go, however it does offer them, and they are available in a simple, readable form for the user instead of the complexity you see in Ethereum.
RaiBlocks is set to solve the issues that Bitcoin has been struggling with: high computational latency (164 min per transaction), high transaction fees, scalability, and the energy consumption (260KWh per transaction) for each bitcoin transaction with proof of work consensus mechanism.
Like XRB based on DAG blockchain technology, IOTA aims to provide an easy, fast and safe way of exchange to IOTA participants and to enable them to buy and trade datasets with ease.
Currently, most of the data is not getting collected and whatever data is available is not being used efficiently. The more devices and companies get connected to the IOTA, the more data will become available on demand.
A short summary
On-chain scalability issue has been known to the community for a while but the issue was escalated by events like crypto-kitties network slow-down, bitcoin humongous slowdown, and increased fees in the end of 2017.
However, as the industry is still in it’s incumbent state, we might need to wait another couple of years to see the best approach (on-chain, off-chain, DAG) and technology to steer away from the dark.
Traders, Investors and Crypto-enthusiasts should keep their eyes peeled, engage in research, and go for the future that has the best odds of working/winning. If you have related questions, feedback on our research, or just a basic question – don’t hesitate to reach out to us.