Wired LAN design is grounded in Ethernet, the dominant LAN technology specified by IEEE $802.3$.2 In the context of Network Theory and Data Communication Components, this topic has two pillars: first, understanding the evolution of Ethernet from traditional $10$ Mbps systems to Fast Ethernet and Gigabit Ethernet; second, understanding the interconnecting devices that extend, segment, or route traffic across a network.2

Traditional Ethernet emerged from shared-media operation and originally relied on CSMA/CD to handle collisions on half-duplex links.2 As Ethernet evolved to switched, point-to-point links, full-duplex communication became standard and collision handling largely disappeared from modern wired LAN practice.2 IEEE milestones commonly referenced in introductory networking are:

• Traditional Ethernet: $10$ Mbps, including $10\text{BASE-T}$ over twisted pair.2
• Fast Ethernet: $100$ Mbps, standardized by IEEE $802.3u$, including $100\text{BASE-TX}$.2
• Gigabit Ethernet: $1000$ Mbps, including fiber variants under IEEE $802.3z$ and copper $1000\text{BASE-T}$ under IEEE $802.3ab$.2

A second core learning goal is mapping devices to the OSI model. Repeaters and hubs operate at Layer $1$; bridges and Layer-$2$ switches operate at Layer $2$; routers operate at Layer $3$; and gateways perform protocol translation and may function across multiple layers depending on the type of conversion being performed.2 This layered view explains why a hub creates one shared collision domain, while a switch creates a separate collision domain per port.2

Footnotes

1. Overview And Guide To The IEEE 802 LMSC - IEEE overview of the 802 LAN/MAN standards family and its historical context. ↩

2. IEEE 802.3 Ethernet Overview - IEEE overview summarizing Ethernet history, speeds, and key standards including $802.3u$, $802.3z$, and $802.3ab$. ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷

3. Collision and Broadcast Domain - Educational explanation of collision domains, broadcast domains, and how hubs, bridges, switches, and routers affect them. ↩ ↩² ↩³ ↩⁴

4. Recognize the purpose & functions of various network devices:Router, Switch, Hub & Bridge. - Cisco-oriented explanation of hubs, switches, bridges, half-duplex behavior, and collisions. ↩

5. What Is an Ethernet Switch? - Cisco explanation of Ethernet switching, dedicated links, and the difference between switches and hubs. ↩ ↩²

6. Ethernet over twisted pair - Technical summary of twisted-pair Ethernet variants such as $10\text{BASE-T}$, $100\text{BASE-TX}$, and $1000\text{BASE-T}$. ↩ ↩² ↩³

7. What is a network gateway? - Cisco description of gateways, protocol translation, and their relationship to routers and OSI-layer functions. ↩

Ethernet standards in wired LANs

Ethernet naming follows a useful pattern: the leading number indicates data rate in Mbps, BASE denotes baseband signaling, and the suffix indicates the medium or variant, such as T for twisted pair. Thus, $10\text{BASE-T}$ means $10$ Mbps baseband over twisted-pair copper, while $1000\text{BASE-T}$ means $1000$ Mbps over twisted pair.

1. Traditional Ethernet

Traditional Ethernet is usually associated with $10$ Mbps operation. Early implementations often used shared media, where multiple stations contended for the same communication channel and therefore required CSMA/CD in half-duplex mode.2 In such an environment, collisions are not anomalies but expected events that the protocol detects and recovers from.

2. Fast Ethernet

Fast Ethernet increased nominal throughput by a factor of $10$, to $100$ Mbps, through IEEE $802.3u$. Common physical implementations include $100\text{BASE-TX}$ over copper and $100\text{BASE-FX}$ over fiber. It preserved Ethernet framing and upper-layer compatibility, which made migration relatively smooth for organizations expanding LAN performance without changing the entire protocol stack.

3. Gigabit Ethernet

Gigabit Ethernet increased throughput again to $1000$ Mbps. IEEE $802.3z$ introduced gigabit operation over fiber, while IEEE $802.3ab$ standardized $1000\text{BASE-T}$ over twisted pair up to $100$ meters.2 A notable engineering difference is that $1000\text{BASE-T}$ uses all four wire pairs bi-directionally, unlike lower-speed twisted-pair Ethernet variants that used fewer active pairs.

For practical LAN analysis, the throughput growth can be summarized mathematically as:

$$100\ \text{Mbps} = 10 \times 10\ \text{Mbps}, \qquad 1000\ \text{Mbps} = 10 \times 100\ \text{Mbps}$$

This scaling preserved the Ethernet ecosystem while greatly improving aggregate LAN capacity.

Footnotes

1. Ethernet over twisted pair - Technical summary of twisted-pair Ethernet variants such as $10\text{BASE-T}$, $100\text{BASE-TX}$, and $1000\text{BASE-T}$. ↩ ↩² ↩³ ↩⁴ ↩⁵

2. IEEE 802.3 Ethernet Overview - IEEE overview summarizing Ethernet history, speeds, and key standards including $802.3u$, $802.3z$, and $802.3ab$. ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷

3. Recognize the purpose & functions of various network devices:Router, Switch, Hub & Bridge. - Cisco-oriented explanation of hubs, switches, bridges, half-duplex behavior, and collisions. ↩ ↩²

Connecting devices and their OSI layers

A wired LAN is not just cabling and hosts; it is built from devices that regenerate signals, forward frames, or route packets based on progressively richer addressing information.2 The essential progression is from bit forwarding at Layer $1$, to MAC-based forwarding at Layer $2$, to IP-based path selection at Layer $3$.2

Repeater

A repeater operates at the physical layer and regenerates incoming signals so that they can travel farther without excessive degradation. It does not inspect addresses, identify destinations, or segment traffic logically.

Hub

A hub is effectively a multiport repeater and also operates at Layer $1$.2 It repeats bits out all ports, meaning all attached devices share the same communication medium behavior. Consequently, a hub forms a single collision domain and commonly uses half-duplex Ethernet semantics.2

Bridge

A bridge operates at the data-link layer.2 It examines MAC addresses, learns which stations are reachable on each side, and forwards frames selectively. A bridge divides collision domains, reducing unnecessary frame propagation, but does not inherently split the broadcast domain.

Layer-2 Switch

A Layer-2 switch is essentially a high-speed multiport bridge operating primarily at Layer $2$.2 It learns MAC addresses per port and forwards frames only where needed. Each switch port forms its own collision domain, which is why switches scale much better than hubs.2

Router

A router operates at the network layer.2 Instead of forwarding based on MAC addresses inside one LAN, it forwards packets between different networks using logical addressing and routing decisions. Routers also separate broadcast domains.

Gateway

A gateway connects dissimilar systems and performs protocol or data-format translation. In practice, the exact OSI-layer association depends on the translation function. Some gateways are described as operating at the network layer, while others may span several layers or even all layers conceptually because translation can involve application semantics as well as addressing.

Footnotes

1. Collision and Broadcast Domain - Educational explanation of collision domains, broadcast domains, and how hubs, bridges, switches, and routers affect them. ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹ ↩¹² ↩¹³ ↩¹⁴ ↩¹⁵ ↩¹⁶ ↩¹⁷ ↩¹⁸ ↩¹⁹ ↩²⁰ ↩²¹

2. What is a network gateway? - Cisco description of gateways, protocol translation, and their relationship to routers and OSI-layer functions. ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸

3. What Is an Ethernet Switch? - Cisco explanation of Ethernet switching, dedicated links, and the difference between switches and hubs. ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷

Hub versus switch: the critical operational difference

The hub–switch distinction is one of the most important comparisons in introductory networking.2 Although both devices connect multiple end systems inside a LAN, they behave very differently.

A hub repeats incoming signals to every port.2 This means:

• all connected nodes are in one shared collision domain.
• the medium behaves as shared Ethernet, typically half-duplex.2
• if two devices transmit simultaneously, a collision may occur and retransmission logic is needed.

A switch, by contrast, forwards frames based on MAC address learning. This means:

• each port is a dedicated collision domain.2
• hosts can usually operate in full-duplex mode with the switch.2
• traffic is directed only to the intended destination port, reducing unnecessary propagation and improving throughput.

This difference can be visualized in terms of contention:

$$\text{Hub network: many hosts} \rightarrow 1 \text{ shared collision domain}$$ $$\text{Switch network with } n \text{ active ports} \rightarrow n \text{ separate collision domains}$$

In a four-host example, one hub still yields one collision domain, but a four-port switch yields four dedicated collision domains. That architectural difference is why hubs are largely obsolete in modern LAN design and switches are standard infrastructure.

Footnotes

1. Collision and Broadcast Domain - Educational explanation of collision domains, broadcast domains, and how hubs, bridges, switches, and routers affect them. ↩ ↩² ↩³ ↩⁴ ↩⁵

2. What Is an Ethernet Switch? - Cisco explanation of Ethernet switching, dedicated links, and the difference between switches and hubs. ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸

3. Recognize the purpose & functions of various network devices:Router, Switch, Hub & Bridge. - Cisco-oriented explanation of hubs, switches, bridges, half-duplex behavior, and collisions. ↩ ↩²

4. IEEE 802.3 Ethernet Overview - IEEE overview summarizing Ethernet history, speeds, and key standards including $802.3u$, $802.3z$, and $802.3ab$. ↩

Design implications for modern wired LANs

Modern wired LANs overwhelmingly favor switched Ethernet, not hub-based Ethernet. The reasons are architectural as well as performance-oriented:

1. Collision isolation: each switch port is its own collision domain.2
2. Full-duplex links: simultaneous send/receive is supported, so CSMA/CD is no longer required on those links.2
3. Higher effective throughput: traffic is sent only where needed rather than repeated everywhere.
4. Structured scalability: multiple switches can be interconnected, while routers provide controlled boundaries between IP networks.

This also clarifies the role of the classic device hierarchy in enterprise network design:

• Repeaters and hubs are legacy Layer-$1$ devices, useful mainly for conceptual understanding.
• Bridges introduced selective Layer-$2$ forwarding and collision segmentation.
• Layer-$2$ switches became the standard LAN building block.
• Routers connect separate networks and constrain broadcast propagation.2
• Gateways enable interoperability across unlike protocols, services, or systems.

A concise comparison of domain behavior is:

From a theory perspective, the transition from hubs to switches marks a shift from shared-medium LAN behavior to microsegmented Ethernet. That shift is one of the main reasons Ethernet scaled successfully from $10$ Mbps classroom examples to modern gigabit campus access networks.2

Footnotes

1. What Is an Ethernet Switch? - Cisco explanation of Ethernet switching, dedicated links, and the difference between switches and hubs. ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸

2. Collision and Broadcast Domain - Educational explanation of collision domains, broadcast domains, and how hubs, bridges, switches, and routers affect them. ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸

3. IEEE 802.3 Ethernet Overview - IEEE overview summarizing Ethernet history, speeds, and key standards including $802.3u$, $802.3z$, and $802.3ab$. ↩ ↩²

4. What is a network gateway? - Cisco description of gateways, protocol translation, and their relationship to routers and OSI-layer functions. ↩ ↩² ↩³