Nick Chong · 32 mins ago · 3 min read
Since its humble origins among academic circles and DARPA researchers, the Internet evolved from a limited, fragmented set of networks to the ubiquitous global behemoth that it has become today. At the core of the Internet’s ability to empower unprecedented levels of communication is the standardization of a fundamental set of protocols, including the application, transport, internet, and link layers of the OSI model.
While these standardized protocols, especially exterior gateway protocols like BGP, are critical to the Internet’s functionality, their origins decades ago reveal ample room for improvement to keep pace with the contemporary demands of connectivity.
Blockchain, despite its volatile, and often circumspect, hype cycle concerning applications outside of bitcoin, has revealed that it may prove a unique component in the next iteration of the Internet’s landscape.
NOIA Network CEO, Domas Povilauskus details:
“With the emergence of Distributed Ledger technologies, the possibility to have one central source of truth, but keeping it both architecturally and politically decentralized became reality.”
The notion of a censorship-resistant medium for reconciling truth is powerful and has some promising applications as the thread that connects emerging technologies striving to improve the Internet’s infrastructure.
The Media Layer and Net Neutrality
An amplified discussion recently surrounds the concept of net neutrality, mainly because legal regulations preserving it were abrogated at the end of 2017. On a practical level, net neutrality refers to the preclusion and discrimination of content by Internet Service Providers (ISPs), which are the primary controllers of gateway access points to the Internet. The idea is predicated on the notion of a more flat, equal, and open Internet.
The problem of removing protections on net neutrality range from ISPs capable of deliberately slowing Internet speeds or censoring certain types of content, whether from political or other outside influence. The core issue stems from the centralization of ISPs, who are (for the most part) confined to giant corporations like Verizon, AT&T, and Comcast.
The reason that these firms can control and censor the Internet is because of their command of a pivotal component of the Internet’s infrastructure—the lower layers of the OSI model, known as the “media” layers.
The media layers (i.e., physical, data, network) of the OSI model contain notable technologies, like Ethernet, which serve as the basis for the connection between nodes and beaming data packets across the globe—the “gateway” technologies of the Internet. Obviously, central control of such technologies inevitably is prone to outside influence, creating situations where instances of protocol, IP, and peering discrimination are all feasible.
Fortunately, regulatory pressure will continue to mount on the government as activists promoting an “open Internet” have an impressive, and sizeable, backing. However, there are also some interesting technical solutions at work behind the scenes that seek to flatten the Internet’s infrastructure.
A common theme among many of these projects? Blockchain
For example, the company Marconi is building a blockchain-based technology at the Ethernet layer that enables a more secure and decentralized Ethernet layer (i.e., layer 2 of OSI), circumventing the centralized control of ISPs. Their solution works by encrypting Ethernet layer traffic, making it more private and secure, while also incentivizing the construction of mesh networks where routing, switching, and packet processing are furnished by a community of users.
The overarching theme that such projects have is that they leverage blockchains, decentralized ledgers of trust reconciliation, as the glue that replaces the role of centralized entities. And that idea even extends higher up the Internet’s infrastructure—to the application layer.
The Application Layer and Programmable Routing
The application layer of the Internet is the one that functions closest to the end-user of the modern web. Technologies incorporated in this layer, which is the top layer of the OSI model, include HTTP and BGP.
While not as centrally controlled as some of the lower layers (e.g., Ethernet), the current model deployed by HTTP and BGP are outdated. Let’s start with discussing HTTP.
HTTP stands for “Hypertext Transfer Protocol,” and was famously conceived by Tim-Berners Lee at CERN in 1989. It is the foundation of content routing on the web, and is what works on the back-end to distribute and serve data packets on the web whenever you click on something. However, despite its undisputed successes, HTTP falls short in a few areas.
Firstly, HTTP downloads media from a single computer (i.e., server) at a time. This leads to expensive content distribution via higher bandwidth usage. Secondly, networking with HTTP is not always reliable, as it is prone to intermittent connections and does not serve content that is removed from servers, precluding old content from being accessible on the web.
Initiatives like the open-source IPFS project are tackling the shortcomings of HTTP head-on. The concept relies on P2P routing of hypermedia across the web, using an indexing and routing protocol with a decentralized naming service. Content and connections become more resilient, and delivery costs are vastly reduced.
Although not specifically based on blockchains, IPFS is wholly congruent with blockchains. Blockchains can serve as an effective, secure medium for indexing and referencing content that is hashed using IPFS. The technologies are also mutually beneficial, as IPFS enables blockchains to inscribe larger data sets using IPFS.
Concerning BGP, which stands for “Border Gateway Protocol,” the problem is more the fragmentation of the Internet’s early growth and the resulting inefficiencies that it has led to. For instance, BGP is prone to daily hijacks, which can cause connectivity disruptions and privacy breaches.
“BGP Hijack occurs when a bogus BGP router, most often an ISP, announces ownership of certain IP address prefixes,” says Povilauskus of NOIA Network. “Due to the decentralized nature of the Internet and BGP, these addresses are instantly propagated, often around the world. Until it is caught and resolved, traffic is being routed through the attacker, or packets are dropped, or people are directed to an impersonated website.”
The salient proposal for improving BGP is called IPv6, which is the most recent version of the Internet Protocol that expands packet sizes and allows for customizable header information. For their part, NOIA Network is blending IPv6 with segment routing (SR), a unique protocol routing method that helps routers understand and make decisions on subsets of information—an initiative to create a “programmable Internet.”
The adhesive holding together the confluence of technologies? Blockchain.
“BGP was partly influenced by not having any single point where all internet topology is held. Theoretically, such a database could be created, but it would be owned by someone. Either by a company or an organization. Therefore BGP design was completely local-centric, to operate without global knowledge of the network.”
Blockchain consequently functions as the medium for empowering a decentralized, public network that retains the same performance capabilities as if it were centralized.
“Distributed Ledger can hold topology of participating SR routers giving you the ability to program data with a global view of available network, on public internet,” details Povilauskus.
Overall, there are pockets of the Internet that remain decentralized and other critical areas that are concerningly centralized. The problem is the friction between the two, and the resulting inefficiencies that result from a lack of congruency between layers of the Internet’s infrastructure.
Net neutrality should remain a prominent topic of discussion in public discourse in the coming years, and while regulatory pressure should prove useful, some of the most potent arbiters of change are taking the form of innovative technology—fueled by blockchains.