33  Layered Models: Review

In 60 Seconds

This review consolidates your understanding of the OSI (7-layer) and TCP/IP (4-layer) network models, covering encapsulation, addressing at different layers (MAC, IP, port), ARP resolution, and how these traditional models map to IoT-specific reference architectures. It is organized into two focused companion chapters: model comparison and supplementary resources.

33.1 Overview

Network layering models are the foundation for all networking knowledge. They define how data moves from an application on one device to an application on another, passing through well-defined layers of abstraction. This review brings together the key concepts you have studied – OSI layers, TCP/IP mapping, encapsulation, addressing, and IoT-specific reference models – so you can see the complete picture before moving on to protocol-specific chapters.

Chapter Navigation

This review has two companion chapters for deeper study:

1. OSI and TCP/IP Comparison – Model fundamentals, key concepts, and detailed summary
2. Resources and Visual Gallery – Videos, tools, and diagrams

33.2 Learning Objectives

By the end of this review, you will be able to:

• Compare OSI and TCP/IP: Differentiate the 7-layer OSI model from the 4-layer TCP/IP model, mapping each OSI layer to its TCP/IP equivalent
• Apply Encapsulation: Trace data flow through network layers, demonstrating how each layer adds its own header during transmission and strips it upon receipt
• Explain Addressing: Distinguish MAC (Layer 2), IP (Layer 3), and port (Layer 4) addressing and explain the role each plays in delivering data across networks
• Analyze ARP: Explain how ARP resolves IP addresses to MAC addresses and predict the behavior of ARP in broadcast and unicast scenarios
• Evaluate IoT Models: Compare traditional OSI/TCP/IP models with IoT-specific reference architectures (Cisco 7-level, ITU-T, IEEE 2413) and select the appropriate model for a given analysis task
• Calculate Protocol Overhead: Apply the packet size formula to calculate payload efficiency for different IoT protocol stacks and compare results to justify protocol selection

33.3 Prerequisites

Required Chapters:

• Layered Models Fundamentals – Protocol stacks
• Networking Fundamentals – Network basics
• IoT Protocols Overview – Protocol context

Technical Background:

• OSI model layers
• TCP/IP model
• Encapsulation concepts

Estimated Time: 30 minutes (all sections)

33.4 Quick Reference

Layer	OSI	TCP/IP	IoT Example
7	Application	Application	MQTT, CoAP
6	Presentation	Application	JSON, CBOR
5	Session	Application	–
4	Transport	Transport	UDP, TCP
3	Network	Internet	IPv6, 6LoWPAN
2	Data Link	Link	802.15.4, BLE
1	Physical	Link	Radio, Ethernet PHY

Note that the TCP/IP model merges OSI layers 5-7 into a single Application layer and OSI layers 1-2 into a single Link layer, resulting in four layers instead of seven.

Putting Numbers to It

Layering decisions directly change packet size, airtime, and battery drain.

\[ \text{Total packet size} = P + H_{\text{L2}} + H_{\text{IP}} + H_{\text{transport}} + H_{\text{app}} \]

Worked example: Compare a 12-byte telemetry message on two stacks.

For CoAP over UDP over 6LoWPAN (with IPHC/NHC header compression on an 802.15.4 link):

Component	Header Size
802.15.4 MAC (short addresses)	11 bytes
6LoWPAN IPHC (compressed IPv6)	7 bytes
NHC-compressed UDP	4 bytes
CoAP (minimal, NON message)	4 bytes
Total overhead	26 bytes
Total packet	38 bytes

For MQTT over TCP over full IPv6 on 802.15.4:

Component	Header Size
802.15.4 MAC	11 bytes
Full IPv6	40 bytes
TCP (no options)	20 bytes
MQTT PUBLISH (minimal)	4 bytes
Total overhead	75 bytes
Total packet	87 bytes

Payload efficiency is \(12/38 \approx 31.6\%\) for CoAP/UDP/6LoWPAN versus \(12/87 \approx 13.8\%\) for MQTT/TCP/IPv6. For the same 12 bytes of sensor data, the uncompressed stack transmits about \(87/38 \approx 2.3\times\) more bytes – directly impacting battery life for constrained IoT devices.

Try It: Protocol Stack Overhead Calculator

viewof payload = Inputs.range([1, 127], {value: 12, step: 1, label: "Payload size (bytes)"})

viewof stack = Inputs.select(
  ["CoAP/UDP/6LoWPAN", "MQTT/TCP/IPv6", "Custom"],
  {value: "CoAP/UDP/6LoWPAN", label: "Protocol stack"}
)

viewof hL2 = Inputs.range([0, 50], {
  value: stack === "CoAP/UDP/6LoWPAN" ? 11 : stack === "MQTT/TCP/IPv6" ? 11 : 11,
  step: 1, label: "L2 header (bytes)"
})

viewof hIP = Inputs.range([0, 60], {
  value: stack === "CoAP/UDP/6LoWPAN" ? 7 : stack === "MQTT/TCP/IPv6" ? 40 : 7,
  step: 1, label: "IP header (bytes)"
})

viewof hTransport = Inputs.range([0, 60], {
  value: stack === "CoAP/UDP/6LoWPAN" ? 4 : stack === "MQTT/TCP/IPv6" ? 20 : 4,
  step: 1, label: "Transport header (bytes)"
})

viewof hApp = Inputs.range([0, 30], {
  value: stack === "CoAP/UDP/6LoWPAN" ? 4 : stack === "MQTT/TCP/IPv6" ? 4 : 4,
  step: 1, label: "App header (bytes)"
})

totalOverhead = hL2 + hIP + hTransport + hApp
totalPacket = payload + totalOverhead
efficiency = ((payload / totalPacket) * 100).toFixed(1)

html`<div style="background:#f8f9fa; border:1px solid #dee2e6; border-radius:8px; padding:16px; margin-top:12px; font-family:system-ui,sans-serif;">
<h4 style="margin:0 0 12px; color:#2C3E50;">Results</h4>
<table style="width:100%; border-collapse:collapse;">
<tr><td style="padding:4px 8px;">Total overhead</td><td style="padding:4px 8px; font-weight:bold;">${totalOverhead} bytes</td></tr>
<tr><td style="padding:4px 8px;">Total packet size</td><td style="padding:4px 8px; font-weight:bold;">${totalPacket} bytes</td></tr>
<tr><td style="padding:4px 8px;">Payload efficiency</td><td style="padding:4px 8px; font-weight:bold; color:${efficiency > 50 ? '#16A085' : efficiency > 25 ? '#E67E22' : '#E74C3C'};">${efficiency}%</td></tr>
</table>
<div style="margin-top:12px; background:#e8e8e8; border-radius:4px; height:24px; position:relative;">
  <div style="background:${efficiency > 50 ? '#16A085' : efficiency > 25 ? '#E67E22' : '#E74C3C'}; height:100%; border-radius:4px; width:${efficiency}%; transition:width 0.3s;"></div>
  <span style="position:absolute; top:2px; left:8px; font-size:12px; color:#2C3E50;">Payload: ${payload}B</span>
  <span style="position:absolute; top:2px; right:8px; font-size:12px; color:#2C3E50;">Overhead: ${totalOverhead}B</span>
</div>
<p style="margin:8px 0 0; font-size:13px; color:#7F8C8D;">
802.15.4 maximum payload is 127 bytes. ${totalPacket > 127 ? '<strong style="color:#E74C3C;">Warning: packet exceeds 802.15.4 MTU -- fragmentation required.</strong>' : 'Packet fits within 802.15.4 MTU.'}
</p>
</div>`

For Beginners: How to Use This Review

What is this chapter? This is a consolidation chapter that ties together everything you have learned about network layering models. Think of it as a map showing how all the pieces connect.

When to use this chapter:

• After studying OSI and TCP/IP models
• Before diving into specific protocols (MQTT, CoAP, etc.)
• When you need to understand where protocols fit in the stack

Key Models to Know:

Model	Layers	Use
OSI	7 layers	Reference model for understanding
TCP/IP	4 layers	Practical internet protocol suite
IoT Stack	Varies	Application-specific combinations

Why This Matters for IoT: Understanding layers helps you choose the right protocol for each part of your IoT system – from physical connections to application messaging. For example, a battery-powered sensor needs lightweight protocols (CoAP over UDP) while a cloud dashboard can afford heavier ones (MQTT over TCP).

Recommended Path:

1. Start with OSI and TCP/IP Comparison
2. Explore Resources and Visual Gallery

Related Chapters

Deep Dives:

• Layered Models Fundamentals – OSI and TCP/IP details
• Networking Fundamentals – Network architecture basics

Layer Implementations:

• Routing Fundamentals – Network layer (Layer 3)
• Transport Fundamentals – Transport layer (Layer 4)
• Application Protocols – Application layer (Layer 7)

IoT Protocol Mapping:

• MQTT Architecture – MQTT in OSI model
• CoAP Architecture – CoAP layer mapping
• 6LoWPAN Stack – Adaptation layer

Comparisons:

• IoT Protocols Review – Cross-layer protocol analysis
• Protocol Selection – Choosing protocols

Learning:

• Quizzes Hub – Networking model assessments
• Videos Hub – OSI model tutorials

33.5 Chapter Contents

33.5.1 Part 1: OSI and TCP/IP Comparison

Read the full chapter

Covers:

• Layer functions as questions (WHAT, HOW reliably, WHERE, HOW to transmit)
• Key concepts: encapsulation, decapsulation, addressing
• OSI 7-layer vs TCP/IP 4-layer mapping
• IoT reference models (Cisco 7-level, ITU-T, IEEE 2413)
• Common misconception: “More layers = better security”

33.5.2 Part 2: Resources and Visual Gallery

Explore resources

Covers:

• Video tutorials on OSI and TCP/IP models
• Interactive tools: subnet calculators, OUI lookups
• AI-generated visual diagrams for different learning styles
• Next steps and advanced topics

Decision Framework: Choosing Between OSI and TCP/IP Models for IoT Protocol Analysis

When to use the OSI 7-layer model:

Scenario	Why OSI	Example
Troubleshooting complex issues	Granular layer separation identifies exact failure point	“Application works but data corrupts” – check Presentation layer encoding
Teaching network concepts	Clear separation of concerns aids learning	Session layer (L5) manages connection state independently
Protocol design	Explicit definition of each layer’s responsibilities	Designing a new IoT protocol stack

When to use the TCP/IP 4-layer model:

Scenario	Why TCP/IP	Example
Implementing real systems	Matches actual protocol implementations	Building MQTT clients – Application layer directly uses TCP
Debugging network stacks	Aligns with Linux networking layers	Analyzing tcpdump output shows 4-layer structure
Performance analysis	Fewer layers = clearer overhead attribution	Measuring protocol efficiency

Decision Rule: Use OSI for conceptual understanding and detailed troubleshooting. Use TCP/IP for implementation, debugging, and real-world protocol analysis.

Real Example: A LoRaWAN gateway fails intermittently. OSI analysis reveals:

• L1 Physical: Radio signal OK
• L2 Data Link: Frame Check Sequence errors – Problem identified at L2
• Fix: Adjust spreading factor to improve error correction

TCP/IP would group L1 and L2 into “Link Layer”, making the root cause harder to isolate.

Common Mistake: Assuming More Layers Means Better Security

The Mistake: “OSI has 7 layers, so it must be more secure than TCP/IP’s 4 layers.”

Why It’s Wrong: Security is implemented by protocol features (encryption, authentication), not by the number of abstraction layers. Both OSI and TCP/IP are reference models for organizing protocols – they do not inherently provide security.

The Reality:

• TLS (Transport Layer Security) provides encryption in both models
• IPsec (Network Layer Security) works identically in OSI and TCP/IP
• Application-layer encryption (e.g., MQTT payload encryption) is independent of the model

The Fix: Evaluate security based on implemented protocols (TLS 1.3, AES-256, certificate-based authentication), not on the number of layers in the reference model.

33.6 Review Activities

33.6.1 Match Protocols to Layers

33.6.2 Order the Encapsulation Steps

Common Pitfalls

1. Reviewing Without Attempting to Draw the Models From Memory

Reading about OSI layers is passive. Fix: close the notes and draw both OSI and TCP/IP models from memory, labelling each layer with one protocol and one function.

2. Not Connecting the Review to Practical IoT Scenarios

Reviewing abstract layer definitions without connecting them to concrete IoT examples (e.g., “6LoWPAN adapts IPv6 at the Network layer for IEEE 802.15.4 links”) leaves knowledge fragile. Fix: for each review topic, write one sentence connecting it to a real IoT protocol or product.

3. Spending All Review Time on Well-Known Layers

Students often spend most review time on layers they already understand (Application, Transport) and neglect the Data Link and Physical layers. Fix: allocate review time proportionally to how well you understand each layer, spending more on weaker areas.

🏷️ Label the Diagram

Code Challenge

33.7 Summary

This review consolidated layered network models and their practical application to IoT systems:

• OSI (7-layer) and TCP/IP (4-layer) models provide theoretical and practical frameworks for organizing network protocols
• Encapsulation proceeds top-down during transmission, with each layer adding its own header around the payload from the layer above
• Addressing operates at multiple layers: MAC at Layer 2 for local hop-by-hop delivery, IP at Layer 3 for end-to-end routing, and port numbers at Layer 4 for process demultiplexing
• ARP resolves IP to MAC using broadcast requests and unicast replies on the local network segment
• IoT reference models extend beyond traditional networking with edge computing, analytics, and business layers
• Protocol stack overhead varies significantly – compressed stacks like CoAP/UDP/6LoWPAN achieve roughly 2x better payload efficiency than uncompressed alternatives

33.8 What’s Next

Topic	Chapter	Description
OSI and TCP/IP Comparison	OSI and TCP/IP Comparison	Detailed model fundamentals, key concepts, encapsulation, and ARP
Visual Gallery and Resources	Resources and Visual Gallery	Videos, tools, and AI-generated diagrams for all learning styles
Encapsulation and PDUs	Encapsulation and PDUs	Deep dive into protocol data units and header structure
Layer 3 Routing in IoT	Routing Fundamentals	How RPL and IPv6 routing operate in constrained IoT networks
Layer 4 Transport	Transport Fundamentals	TCP vs UDP trade-offs and their impact on IoT protocol design
Application-Layer Protocols	Application Protocols	MQTT, CoAP, and HTTP mapped to the application layer