- Atlas
- Computer Science
- Network Architecture
Network Architecture
9 frameworks, from 1890 to today · Networks · Computer Science
- Circuit Switching 1890-1980
- Packet Switching 1965-Present
- Vendor Network Architectures 1974-1990
- Layered Reference Models 1978-Present
- End-to-end principle 1981-Present
- Internetworking and Federation 1983-Present
- Middlebox and Overlay Architecture 1994-Present
- Clean-Slate Architecture 2005-Present
- Software-Defined Networking 2008-Present
Overview
Network architecture is the set of large-scale decisions about how a communication network is organized: who decides where traffic goes, how much intelligence sits in the network versus at its edges, how different networks and technologies are joined, and how the whole system is divided into parts that can be built and changed independently. It is the difference between laying cable and designing a system—between asking "how do we move this bit?" and asking "what kind of network should exist at all?" The stakes are enormous because architecture choices outlive hardware: a decision about where to put control logic or how to split responsibilities into layers can shape what is easy to build, secure, and upgrade for decades. The history of network architecture is largely a history of arguments about that division of labor.
Circuit switching and the telephone monopoly
The first global communication network was the telephone system, and its architecture was hierarchical and circuit-switched. A circuit-switched network reserves a dedicated path—a physical circuit—between two callers for the entire duration of a call1. In the telephone system this meant a strict hierarchy: local exchanges fed into toll centers, which fed into regional and national switches, with each level aggregating traffic from the level below. The architecture mirrored the organizational structure of the monopoly telephone administrations that built it. It was extremely good at what it was designed for—reliable, predictable voice service—but its control was centralized, its hierarchy rigid, and setting up a call meant tying up resources end to end whether or not anyone was actually speaking.
Packet switching on both sides of the Atlantic
The decisive break came from packet switching, an idea developed in the 1960s in several places at once. Rather than reserving a circuit, a packet-switched network chops data into small packets, each carrying a destination address, and lets them share links and be forwarded independently2. The intellectual case was made independently by Paul Baran at RAND in the early 1960s (motivated by survivable communications) and Donald Davies at the UK's National Physical Laboratory, who coined the term "packet" around 1965–662. The practical demonstration came with the ARPANET, whose first packet switch was installed at UCLA in 1969. Packet switching changed the architectural question fundamentally: if links are shared and packets are independent, then where does the intelligence go? The ARPANET's answer was to put addressing and forwarding into the network's switches, the Interface Message Processors.
Vendor architectures and the layered idea
As computer networking commercialized in the 1970s, vendors built proprietary architectures to lock customers in. IBM's Systems Network Architecture (SNA), announced in 1974, was the most influential: a complete, hierarchical design for connecting terminals and mainframes, with a master control point and a strict tree of subordinate nodes. Digital Equipment Corporation's DECnet and others offered rival schemes3. These systems worked well within one vendor's world but made it hard to connect different manufacturers' equipment—the "heterogeneous network" problem.
The architectural response was layering: dividing communication into a stack of levels, each providing services to the one above and relying on the one below, so that a layer can be redesigned without disturbing its neighbors. The most influential formalization was the Open Systems Interconnection (OSI) reference model, developed within the International Organization for Standardization from the late 1970s and published in 19843. OSI's seven layers—physical, data link, network, transport, session, presentation, application—became a common vocabulary. On the research side, the ARPANET community developed its own layered model around the Transmission Control Protocol and Internet Protocol (TCP/IP), split in 1978 into a connection-oriented transport layer (TCP) and a connectionless internetwork layer (IP). Where OSI imagined a rich, standardized stack with the network providing many services, TCP/IP imagined a minimal network that simply moved datagrams—independent packets—and pushed complexity to the endpoints.
The end-to-end argument
In 1981, Jerome Saltzer, David Reed, and David Clark published an argument that became the philosophical core of the Internet's architecture4. The end-to-end argument holds that a function should be implemented at the endpoints of communication rather than inside the network whenever the endpoints are the only ones who can implement it correctly and completely. Reliable delivery, for example, must ultimately be checked by the receiving application, since the network cannot know whether data arrived intact at the far end; therefore building reliability into every router is redundant, and the network is better kept simple. This did not claim that in-network functions are never useful—it argued for placing them carefully. The argument convinced a generation of designers because it explained why the Internet could run over any underlying technology and support any application: the network stayed dumb and general, and intelligence lived at the edges.
Federating separate networks
The Internet's second architectural principle was federation. Rather than one network, the Internet is a network of networks: independently administered domains that agree to carry each other's traffic. Early work on internetworking protocols in the 1970s established how packets could cross from one network technology to another. The Border Gateway Protocol (BGP), first specified in 1989 and refined over the following years, became the mechanism by which autonomous systems—separately governed networks—exchange routing information and apply their own policies about whom to carry traffic for. This made global interconnection possible without a central authority, at the cost of a routing system governed by local policy rather than global optimization. The architecture of the Internet thus embedded a political stance: interconnection by voluntary agreement among peers.
Middleboxes, overlays, and the erosion of the clean model
By the 1990s the tidy end-to-end picture was under strain. Firewalls, network address translators (NATs), and proxies—collectively called middleboxes—sat in the middle of paths and modified or blocked traffic. NATs, which let many private machines share one public address, became ubiquitous as IPv4 addresses ran short5. Meanwhile overlay networks built virtual topologies on top of the Internet: peer-to-peer systems, content delivery networks, and later virtual private networks and overlay routing services. These developments showed that the "dumb network" was in practice full of intelligent devices, and that architecture was increasingly negotiated between what the protocol designers intended and what operators deployed.
Clean slate and the case for redesign
Frustration with the Internet's ossification—its difficulty in adopting new core protocols—led in the 2000s to the clean-slate movement, most visibly the US National Science Foundation's FIND and GENI programs and the European FIRE initiative. Clean-slate researchers argued that incremental patching could not solve deep problems of security, mobility, and addressing, and that the field needed to design new architectures from scratch, unconstrained by backward compatibility. Projects explored named-data networking (where you request content by name rather than address), new naming and identity schemes, and network virtualization as a way to run many experimental architectures at once. The movement produced important ideas and testbeds, though no wholesale replacement of the Internet.
Software-defined networking
The most consequential architectural shift of the 2000s and 2010s came from a different direction: separating the control plane (the logic that decides where traffic goes) from the data plane (the hardware that forwards packets). The idea grew out of earlier work on programmable networks and active networks in the 1990s, and was crystallized by the OpenFlow protocol, introduced around 2008 by researchers at Stanford and Berkeley. OpenFlow let a centralized controller program the forwarding tables of many switches through a standard interface6. This made network behavior programmable from software, enabled network-wide policies and experiments, and led to commercial controllers and to the wider adoption of the control/data-plane split in data centers and enterprise networks. Programmable data planes, such as the P4 language introduced in 2014, extended the idea to the switches' own packet-processing logic. The architectural question—how much centralized control is safe and scalable—remains one of the field's live debates.
Where the field stands
Today's network architecture descends from these layered strata. The Internet's end-to-end and federation principles still structure the public network, though middleboxes, overlays, and cloud provider networks complicate the picture. Software-defined control and programmable hardware are now standard in data centers and increasingly elsewhere. Open questions remain sharp: how to secure routing and naming without a central authority, how to accommodate mobility and massive numbers of devices, how much intelligence should return to the network for performance and policy, and whether new architectures can ever displace the installed base. The answers will be argued, as they always have been, over the division of labor between the center and the edge.
References
- Circuit switching, Wikipedia
- Packet switching, Wikipedia
- OSI model (History, introduction), Wikipedia
- End-to-end principle, Wikipedia
- Network address translation (introduction, History), Wikipedia
- OpenFlow (introduction, History), Wikipedia
The frameworks of Network Architecture
Each framework is a way of working the field has used: when it held, and the question it answers.
- 1890-1980
- 1965-Present
- 1974-1990
- 1978-Present
Layered Reference Models
How do we divide communication into independent parts?
- 1981-Present
- 1983-Present
Internetworking and Federation
How do separate networks join without a ruler?
- 1994-Present
Middlebox and Overlay Architecture
What happens when devices sit inside the path?
- 2005-Present
- 2008-Present
How the ideas connect
Each arrow runs from an earlier idea to the later one it shaped. The list below says what passed between them.
Packet Switching reacted against Circuit Switching
Packet switching was defined against the circuit-switched telephone model: Baran and Davies argued that reserving a dedicated path end to end was wasteful and fragile, and proposed instead sharing links among independently forwarded packets. The ARPANET's first switch in 1969 demonstrated the alternative in practice.
Carried by Paul Baran; Donald Davies.
Internetworking and Federation grew out of Packet Switching
Internetworking grew out of packet switching as the problem of joining separate packet networks: the early internetworking protocols of the 1970s let packets cross between different network technologies, and the adoption of TCP/IP in 1983 made the network of networks operational.
Carried by Vint Cerf; Robert Kahn.
Layered Reference Models reacted against Vendor Network Architectures
Layering was formulated as the answer to the heterogeneous-network problem created by incompatible vendor architectures: if communication is split into levels with standard interfaces, equipment from different manufacturers can interoperate. OSI and the TCP/IP layering both defined themselves against the closed vendor stacks.
Carried by Open Systems Interconnection reference model; Transmission Control Protocol.
Middlebox and Overlay Architecture reacted against End-to-end principle
Middleboxes and overlays arose in practice as a departure from the end-to-end ideal: firewalls, NATs and proxies put intelligence back inside the path, and overlays built virtual topologies above the network. The tension with the end-to-end principle is what made these developments architecturally contentious.
Vendor Network Architectures influenced by Packet Switching
Vendor architectures such as IBM's SNA were built on packet-switched and store-and-forward ideas for connecting terminals to mainframes, but wrapped them in proprietary hierarchical control. The packet concept supplied the transport mechanism; the vendors supplied the closed architecture.
Carried by IBM Systems Network Architecture (1974).
Layered Reference Models influenced by Packet Switching
The TCP/IP layered model grew directly out of the ARPANET packet-switching work, splitting the earlier monolithic protocol into a connectionless internetwork layer and a transport layer in 1978. The datagram idea from packet switching is what the IP layer was built to carry.
Carried by Vint Cerf; Robert Kahn.
End-to-end principle influenced by Layered Reference Models
The end-to-end argument was framed within the layered model, asking at which layer a function such as reliable delivery belongs. Saltzer, Reed and Clark argued that functions should be pushed to the highest layer that can implement them correctly, which presupposes the layer stack.
Carried by Jerome Saltzer; David Reed; David Clark.
Internetworking and Federation influenced by Layered Reference Models
Federation was implemented through the layered TCP/IP model: IP provided the common internetwork layer across heterogeneous underlying networks, while BGP later handled routing between autonomous systems. The layering made it possible to interconnect networks without standardizing what lay below.
Carried by Border Gateway Protocol.
Clean-Slate Architecture reacted against Layered Reference Models
Clean-slate architecture defined itself against incremental patching of the existing layered stack, arguing that backward compatibility with the established layers prevented solving security, mobility and addressing problems. The movement proposed designing new architectures from scratch rather than adding layers.
Clean-Slate Architecture influenced by End-to-end principle
Clean-slate researchers took the end-to-end argument as a starting design principle and argued that the Internet's ossification had blocked its further application. Their case for redesign rested on the same division of labor between network and endpoints.
Software-Defined Networking influenced by End-to-end principle
Software-defined networking kept the end-to-end principle's separation of simple forwarding from edge intelligence, but moved control logic into a centralized controller rather than the endpoints. OpenFlow's designers framed the controller as restoring programmability to a network whose core had been kept deliberately simple.
Carried by OpenFlow.
Software-Defined Networking influenced by Middlebox and Overlay Architecture
SDN drew on the programmable-network and active-network work of the 1990s, which had sought to put custom logic into the network path. OpenFlow generalized that impulse into a standard control interface for switches.
Carried by OpenFlow.
Software-Defined Networking influenced by Clean-Slate Architecture
Clean-slate work on network virtualization as a way to run many experimental architectures at once fed into SDN's use of virtual networks and centralized control for experimentation. The two movements shared people and testbed infrastructure in the late 2000s.
Beyond the field
Where Network Architecture's ideas came from in other fields, and what they led to there: the strongest 12 of 52 links.
Circuit Switching provoked Packet Switching in Networks
Packet switching was formulated as a deliberate alternative to circuit switching, whose reservation of an end-to-end path for the duration of a call was the established telephony model. Baran, Davies and others defined packet switching by rejecting that dedicated-circuit assumption in favour of statistical sharing.
Carried by Paul Baran; Donald Davies.
Circuit Switching provoked Packet Switching in Computer Systems
Packet switching was formulated as a deliberate alternative to circuit switching, whose dedicated-path allocation was seen as wasteful for bursty data. The ARPANET designers defined their approach by rejecting circuit switching's reservation model.
Carried by Paul Baran; Donald Davies.
Circuit Switching provoked Packet Switching in Routing And Internetworking
Packet switching was formulated by Baran and Davies as a deliberate alternative to circuit switching, whose reservation of a dedicated path per call was the model being rejected for bursty data traffic.
Carried by Paul Baran; Donald Davies.
Circuit Switching provoked Packet Switching and Layered Network Architecture in Communication Systems
Packet switching was defined against circuit switching: Baran and Davies argued that message blocks routed independently would survive failures that circuit-switched networks could not, and the layered architecture was built on that rejection.
Carried by Paul Baran; Donald Davies.
Packet Switching led to Internet Protocol in Routing And Internetworking
The Internet Protocol was designed by Cerf and Kahn as a way to interconnect dissimilar packet-switched networks, so packet switching is the substrate the protocol generalises.
Carried by Vint Cerf; Robert Kahn; A Protocol for Packet Network Intercommunication (1974).
Packet Switching influenced Local Area Network Standards in Networks
Packet switching as a network architecture is the basis on which LAN standards were built, with local networks being packet-switched subnetworks within larger internetworks. The standards define the link layer of a packet-switched architecture.
Layered Reference Models influenced Request/response application protocols in Network Protocols
Request/response application protocols such as HTTP were designed against the layered reference model of the OSI and TCP/IP stacks, with HTTP explicitly sitting above a reliable transport. The layered model supplied the architectural frame in which request/response protocols were placed.
Carried by Tim Berners-Lee; RFC 1945 (1996).
Internetworking and Federation led to Interdomain Policy Routing in Routing And Internetworking
Interdomain policy routing grew directly out of internetworking and federation work, as independently administered networks needed rules for exchanging routes across administrative boundaries.
Middlebox and Overlay Architecture influenced by Layered Network Architecture in Networks
Middlebox and overlay designs were conceived against the fixed layered stack of OSI and TCP/IP, placing functionality at layers the architecture did not anticipate. The end-to-end and layering debates framed where middleboxes could legitimately sit.
Carried by Jerome Saltzer; David Clark.
Middlebox and Overlay Architecture influenced by Internetworking and the End-to-End Argument in Networks
The end-to-end argument and internetworking model are the doctrine middleboxes and overlays violate and argue against, making the later work intelligible only as a response to them.
Carried by Jerome Saltzer; David Clark.
Middlebox and Overlay Architecture influenced by Packet Switching and Layered Network Architecture in Communication Systems
Packet switching and its layered architecture are the substrate that middleboxes intercept and overlays tunnel over; the design assumptions of that stack define what a middlebox or overlay is reacting to.
Carried by Louis Pouzin; Vint Cerf.
Clean-Slate Architecture reacted against Layered Network Architecture in Networks
Clean-slate architecture defined itself by rejecting the ossified layered stack, arguing that incremental patching of TCP/IP could not solve security and mobility. The layered architecture is the target the clean-slate programme attacks.

Network Architecture is a planet in Nousfarer
Its city has a tower for each of these 9 frameworks. Land, walk in, and climb one concept at a time.
Land on Network ArchitectureRelated fields
Fields its ideas are linked to
- Routing And Internetworking 13 links
- Networks 9 links
- Network Protocols 7 links
- Software Defined Networking 6 links
- Communication Systems 5 links
- Wireless Networks 3 links
- Computer Systems 2 links
- Systems Architecture 2 links
- Technology And Internet Law 1 link
- Digital Privacy 1 link
- Systems Security 1 link
- Cloud Computing 1 link
More in Networks
- Congestion Control
- Network Protocols
- Networks
- Peer To Peer And Content Delivery
- Routing And Internetworking
- Software Defined Networking
- Wireless Networks
See where Network Architecture sits on the Map of Human Knowledge