I am not invested in any cryptocurrencies. I am all in on crypto. These statements are not contradictory. While decentralized finance (DeFi) has decidedly garnered much attention, both laudatory and critical, the algorithms and architectures underpinning many DeFi protocols have many useful properties that can be applied to solve a multitude of problems unrelated to money and finance.
Despite all the ups and downs, scams, and hacks of the cryptocurrency world, the underlying decentralized networks have generally chugged along smoothly, processing transactions and maintaining their ledger state. This is unsurprising, as these networks are Byzantine fault-tolerant consensus networks, and, from an adversarial perspective, provably nigh-impossible to attack, even if the attacker can infiltrate the network. In a world where new vulnerabilities, exploits, and hacks are accelerating, Byzantine fault-tolerant consensus networks represent a bastion of security.
Of course, extreme robustness and resilience have a cost. Most applications require greater responsiveness than can be practically delivered within Byzantine fault-tolerant systems, and such systems require redundancy that is excessive or even counter-productive for most applications. It is likely undesirable for your instant messenger conversations to be replicated across all the nodes of a network that expended considerable message and logical complexity to achieve that consensus in the assumed presence of adversaries.
However, a Byzantine fault-tolerant network need not be an entire system. Given their performance characteristics, they should be employed sparingly. They are well suited to be the source of trust, reliability, and security for a broader system, effectively marrying the robustness of the Byzantine fault-tolerant network to the performance of other architectures and designs. One architecture that has particular synergy with a Byzantine fault-tolerant network is the purely peer-to-peer (P2P) application. Consider the advantages of the idealized P2P design:
- No single point of failure – A purely P2P app needs only that the device running it be online to function. There is no cloud or other backend infrastructure that can go down and disrupt service for all users.
- Substantially reduced attack surface – a P2P app's attack surface is largely restricted to what it exposes to connected peers. No component of the design needs to promiscuously accept incoming connections.
- Arbitrary and organic scalability – as each device running the application is effectively its own backend, as usage grows, so does the available hardware.
- Stronger privacy and data sovereignty – ownership and restricted access to self-hosted data is the default, and any sharing must be intentionally designed into the application.
Of course, P2P apps are uncommon on the modern internet, as a P2P app requires some mechanism for peers to discover, connect, and authenticate each other. In other words: find who you want to talk to, establish a connection over the network with them (often through firewalls and network address translation (NAT)), and authenticate that they are who you think they are. Historically, these mechanisms were the weak point of P2P apps, as exemplified by the fact that the tracker servers and indexes were historically the vulnerability of torrent protocols.
Now let us consider an architecture where a Byzantine fault-tolerant network is used for peer discovery, connection, and authentication. The security, reliability and redundancy of the Byzantine fault-tolerant network will ensure peers are always able to find, connect, and authenticate each other. Upon the establishment of an authenticated connection between peers, the more performant P2P architecture can take over. In short, this design would leverage a Byzantine fault-tolerant network for the public infrastructure required to establish a private P2P connection. In effect, it would allow the P2P architecture to be not just viable, but more robust and scalable than the more typical client-server architecture by filling the one historic hole in P2P designs with the extreme redundancy and security of a Byzantine fault-tolerant consensus network.
At NetcoreNetwork, we have built this proposed architecture. We began by designing and building a Byzantine fault-tolerant network that is optimized not for currency or financial transacting, but for facilitating P2P connectivity. We next moved on to building Causeway, which is peer-side tooling to leverage the new Byzantine fault-tolerant network to establish secure, authenticated connections. And with this in place, we have finally begun to build apps. While it is still early (we are a seed-stage startup with a lean team), the initial results have been promising: performance similar to or exceeding the classic client-server model in use cases ranging from remote collaboration to IoT devices, and with dramatically reduced attack surface and the promise of greater scalability. We are beginning to share these applications with the world as our Byzantine fault-tolerant network approaches a production-ready state, and look forward to seeing what others will build with this new, private, secure, and performant architecture.
Website: https://www.netcore.network/
Twitter: https://x.com/netcorenetwork
LinkedIn: https://www.linkedin.com/company/netcorenetwork/
About the Author: Greg Tseng is a co-founder and lead architect of NetcoreNetwork. A Thiel Fellowship Alumni with a passion for cryptography and P2P networking, Greg believes the true potential of crypto extends far beyond digital currencies. His work focuses on leveraging Byzantine fault-tolerant networks to build more secure, private, and resilient internet infrastructure. At NetcoreNetwork, he is dedicated to making P2P connectivity practical, scalable, and accessible to developers worldwide.
Greg Tseng — Co-Founder & Architect at NetcoreNetwork https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgPye7WlPp6MDkY4Njz8yzPPVy2pOgs6-wJXsRDKI96X7iCEJaBURFmRgrwiymJ28UxAta9rYCIVzB76-RCW7cD10GjBUxKX3Qx_PKB7R9dQuww2hVgLBEAz-2b0fLlp_LHD0rgMrdhheUJ5OIQ2YqQo4Khpz_fRsSbT9E3tlt5iEfPN0kafS5mm5m8tmU/s1700-e365/g.png





