Javier Castro - Fotolia
When Satoshi Nakamoto invented bitcoin in 2009, it was recognized as a revolutionary alternative to central banks. But it didn't take long for people to realize that the underlying technology -- blockchain -- was capable of replacing intermediaries in other industries, too: healthcare, voting, real estate, legal documents, security and personal identity protection, rentals, charities and education.
With the popularization of blockchain projects, some of the underlying problems with blockchain also came into clearer view, suggesting the platform, in fact, was not ready for mass adoption. For instance, a single crypto-game slowed down the entire Ethereum network, while the transaction latency and low fees of bitcoin surged during its peak use.
Bitcoin -- as the first and biggest blockchain network -- faced criticism for not being environmentally friendly, wasting CPU hours and requiring sophisticated hardware to be profitable. The need for select hardware particularly riled blockchain enthusiasts, as the requirement ran counter to the protocol's democratic nature, because it in effect centralized power in the hands of large mining pools, critics said.
Because of these problems with blockchain, the hype pendulum has swung to the other extreme, with some now claiming the technology has no real use, citing as evidence the fact that 10 years after the release of bitcoin's white paper there are no other mainstream blockchain apps.
To be sure, there are many projects working to solve blockhain's problems, each taking different routes. But simultaneously, efforts such as Facebook's plan to launch its own cryptocurrency threaten what some say is blockchain's real value. Interviews with blockchain experts suggest perhaps the most salient reason for blockchain's slow adoption: The technology is inherently complex.
Problems with blockchain: Addressing the 'trilemma'
At the core of any blockchain protocol is the chosen consensus algorithm: the code that defines who the nodes talk to, how information is shared and who should be trusted. This algorithm needs to address -- and balance -- three topics simultaneously: security, scalability and decentralization. The blockchain community calls this "the trilemma."
"For example, bitcoin uses proof of work as its consensus mechanism which, as the network continues to live on, requires a lot of computational efforts to mine blocks," said Michaele James, executive vice president of engineering, at 9th Gear Technologies. "To increase the scalability of a blockchain, you are almost always hindering either network security (for example, using proof of stake), or its decentralization (for example, using proof of authority)."
According to James, the proof of stake (PoS) consensus method allows more nodes to participate in moving the blockchain forward at quicker speeds than proof of work (PoW), but at the same time is more prone to exploits as there is a lower barrier-to-entry for attackers.
Similarly, the proof of authority (PoA) consensus model allows the network to be faster because a few authorized nodes are selected to mint new blocks, but now the forward movement of the blockchain is centralized around the few authorized nodes.
For this reason, the best balance is achieved by tailoring the project to its specific use case. One way to do this is by limiting the bad actors.
How private permissioned chains solve problems with blockchain
"At 9th Gear, we are using a private permissioned blockchain, where your everyday person cannot access our blockchain network. Instead, participants of our network go through a screening process to verify their intentions and validate their stake in the proper operation of our network," James explained.
The permissioned network solves a couple of problems with blockchain for 9th Gear: It provides an element of trust not found -- and often purposely avoided -- in a public blockchain network, James said. In addition, it gives 9th Gear the ability to use a more trustful consensus method such as IBFT (the Istanbul Byzantine Fault Tolerant consensus mechanism) that, according to James, "allows for both higher throughput and faster transaction settlements than a standard public blockchain network."
While permissioned chains offer much higher efficiency and faster network, they do so at the cost of the distributed, public and democratic nature of blockchain. This might seem like defeating the purpose of blockchain's existence, but it turns out such chains are actually well-suited for enterprises.
Ian Kane, COO and founder of Ternio, and a proponent of permissioned blockchains, argued that permissioned does not mean centralized. "Our protocol is fully decentralized, but not fully democratized like the bitcoin protocol."
He said enterprise organizations eschew public blockchains because they can become a competitive intelligence tool. "Companies want to know who is getting access to the data on chain, so everyone can garner transparency without getting access to information they should not have."
Ternio, which is approved as an Amazon Advanced Technology Partner to provide blockchain as a service to any AWS client, is capable of delivering over 1 million transactions per second. As Kane explained, this model is best suited for supply chain transparency, data immutability and stable coins (i.e., cryptocurrencies not vulnerable to the typical price instability of digital currencies).
Proof of authority
PoA is a method used by Microsoft on Azure. The algorithm is more efficient as it removes the need for mining while retaining Byzantine Fault Tolerance. The requirement, though, is that all consensus participants are known and reputable.
Mark Moore, COO and co-founder of Inxeption, has used this concept in commerce. "There aren't many (or any) areas of commerce in the U.S. which aren't governed by law. Therefore, there is trust and no need to burn up CPUs trying to prove or authenticate a transaction is real."
Blockchain provides Inxeption a distributed database with an inherent crypto capability for protecting, and covering, data or transactions. This allows different parties to share data outside of their protected zones and still have assurances of data protection and consistency.
It also inherently keeps a trusted audit trail so all parties can be certain the database has not been tampered with. According to Moore, using a centralized database for the same purpose would require a lot of fundamental additions and would be even more difficult to implement.
Precomputation targets problems with blockchain's speed
Projects aimed for public use do not have the luxury of deploying permissioned chains. For those, ingenious algorithms are necessary to reach high performance without sacrificing decentralization or security. Elixxir.io, a decentralized messaging and payments platform, has found a way to increase the speed of the blockchain without making participation private.
"Elixxir deploys a breakthrough in mix networks, called cMix, which leverages precomputation to perform the vast majority of processing before transactions even occur," said Founder David Chaum. A mix network is a sequence of nodes and servers that relay messages or transactions in a way that preserves the confidentiality of the sender, receiver and content, by sending them separately.
According to Chaum, precomputation can bring transaction speeds on par with mass-market consumer apps. "The result is a service that performs 1,000 times faster than systems without precomputation."
Another method is suggested by Richard Dennis, founder of the cryptocurrency temtum. His solution is based on removing the competition between blockchain nodes. "Rather than PoW or PoS consensus algorithms, the temtum algorithm works by randomly choosing a 'leader node' to process transactions over a 60-second period," Dennis said. This equates to five blocks before the beacon generates a new unique value to move to a new "leader node."
This way, temtum picks the next node to confirm all transactions for a set period with a single calculation. This means the billions of calculations currently being done via a PoW are no longer needed, while still ensuring that every node on the network comes to the same conclusion without requiring more information from other nodes.
"This process is highly secure and extremely efficient, removing competition between nodes and preventing the need to spam all nodes with every transaction," explained Dennis. Temtum, which has been five years in development, has just recently launched its platform. The technology, which has been independently verified and pen tested by Crest accredited agency BSI, is also considering future threats.
"The selection [procedure] is quantum secure, whereby a quantum source of randomness (light photons) is used, meaning there's no way of predicting the node that will process transactions," Dennis said. "A unique Performance Integrity Protocol backed by AI ensures that any leader node is also trustworthy by monitoring performance, bandwidth and uptime over long periods of time."
Blockchain for IoT: Proof of honesty
IoT deployments have their own problems with blockchain. IoT sensors and devices have low battery and storage. Thus, blockchain IoT developers must deal with feasibility issues, particularly if they want to store a high volume of otherwise low-value records (such as IoT telemetry) on the blockchain.
Stephanie So, co-founder and chief development officer at Geeq, a company that offers public blockchain infrastructure as a service, explained that PoW systems are typically unsuitable for this, as they need to propagate the messages throughout the entire network.
Thus, Geeq's approach is a proprietary validation system, dubbed "Proof of Honesty," which relies on a small network (around 10 to 25 nodes) for each device without requiring propagating data to a larger network.
The system is designed to ensure that the data logged on the blockchain is not corruptible, by monitoring the sensor input and filtering out invalid readings (as defined by the prespecified blockchain's conditions) before putting it in the blockchain. "Our protocol also guarantees that an honest chain will exist that is immutable," So said.
What you should do
In this piece, we've touched on a few of the problems with blockchain and the many creative ways to implement blockchain technology. Developing for blockchain is different from the traditional client-server approach taken in central solutions. We rarely care about how our MySQL or Apache server is coded and only focus on our own database or website's codes. With blockchain, how the chain is implemented matters, and companies need to find the perfect balance between scalability, security and decentralization.
The most important takeaway is that there is no silver bullet; CIOs must look at the problem they are trying to solve, the parties involved and their restrictions to find if blockchain -- and what version of it -- fits their environment.