CRYPTO 101 recently had the privilege of sitting down with Dr. Xinshu Dong for a conversation regarding his company, Zilliqa and their efforts to deliver a solution that addresses the scaling challenge for blockchain technology. What is this scaling challenge, exactly? The use of blockchain networks has grown exponentially in recent months and the resulting high transaction volumes have exceeded the expected throughput capacity of those networks’ nodes. This has caused delays in the networks’ speeds, affecting their performance and end-user experience. But before we dive further into the issue, let’s find out a little more about Dr. Dong and how Zilliqa came to be.
Background
Xinshu earned his PhD in Computer Science from the National University of Singapore with a focus on methods and techniques for improving the security for a wide spectrum of web systems. Building upon this foundation, he then joined research efforts at the Univeristy of Illionois’ Advanced Digital Sciences Center (ADSC) located at the Fusionopolis research facility in Singapore. As a part of the ADSC team, Xinshu worked on how to make the more vulnerable software components of Cyber-Physical Systems (CPS) more secure. You can probably see the trend here, so it is not surprising to learn that he began to study blockchain technology in his…free time?
Recognizing an opportunity for the application of his academic work and personal studies towards solutions for blockchain scalability and security, Dr. Dong founded Zilliqa last summer. The non-profit’s intent is to provide a platform that delivers higher throughput (transactions per second) without sacrificing the security and decentralization advantages that are a fundamental part of blockchain protocols. They considered starting with enhancement of existing chains, but soon realized that this would disrupt their compatibility with existing applications. So, they started from scratch with development of their own.
Public or Private Blockchain?
The Zilliqa team began its development work with a private blockchain in mind, but soon identified some unique aspects of privatized ecosystems that threatened full manifestation of their technology:
- Private blockchain endeavors are typically aiming to maximize performance and privacy.
- Security is not a high priority for these use cases, since private blockchains are utilized by a single organization or group of entities that have already made off-chain security arrangements to preserve the integrity of their network.
- Since security is not within the private blockchain’s scope, a small amount of nodes (30 to 50 on average) are implemented in order to save network overhead costs. Nodes generally ensure the security or immutability of the chain’s data through their consensus algorithms.
The node reduction conflicted with Zilliqa’s strategy because their technology improves the speed of a blockchain (via a higher throughput) that relies upon a higher number of nodes; nodes that have scaled proportionally with the growth of the network. Basically, the more popular a blockchain-based application becomes, the more users will join the incentivized network, naturally boosting the node population. With more nodes on the network, the blockchain can now function at a higher throughput. Recognizing this, Zilliqa shifted course towards development of a public blockchain. But public blockchains come with their own kinds of concerns.
In a public ecosystem, developers must assume that any and every node could act maliciously to violate the security of the blockchain. Zilliqa has accepted this challenge, believing in the strength of numbers their technology will help to ensure. But if their strategy is based on growth, how will they address the same throughput challenges faced by blockchains today? Won’t more users and more nodes bombard the networks with more transactions, too? To deal with this, Zilliqa is implementing a node management technique known as sharding.
Sharding
Your typical blockchain is designed to require that the entire network of nodes “talk” among themselves to determine the rejection or acceptance of each transaction prior to recording their consensus into the applicable block. Having the entire network participate in this process is highly inefficient. Sharding divides the entire network of nodes into smaller groups and assigns a subset of transactions to each group. That smaller group vets its batch of transactions and produces its output for eventual aggregation with the output from the other groups to form the final block.
This parallel processing improves efficiency despite a larger node population overall. Zilliqa is streamlining even further by incorporating mechanisms to have a substantial number of nodes collectively decide nodal group assignments on a revolving basis, for both newcomers and existing participants. This reallocation maintains a healthy balance for group sizes but also strengthens the network’s security because network attackers can exploit a static node as a weakness.
The Future
Dr. Xinshu Dong and Zilliqa are well on their way to delivering a solution to one of the crypto world’s most pressing issues. As you can see above (at time of writing), their latest testnet trial run showed performance was just shy of 2,500 transaction per second (leveraging six shards and 3,600 individual nodes). For comparison, the holy grail benchmark for throughput is often seen as the estimated 8,000 transactions per second average of VISA and MasterCard networks. While the numbers and technical results are impressive, Dr. Dong believes that Zilliqa’s success will depend on their commitment to realizing the full potential of blockchain technology – thus enabling a more efficient and more secure global community.
For the latest information on Zilliqa and their work, you can find them on Medium, Twitter, Reddit, Slack or YouTube
—————————————————————————————————————————————–
About the author: Ryan LaMonica is a management consultant and blockchain enthusiast with a background in engineering, energy, marketing and risk management. The views reflected in this article are his own and do not reflect those of his employer. He currently resides outside of Atlanta, Georgia where he and his wife manage the energy and risk of their four amazing children.