Show understanding of circuit switching

14.2 Circuit Switching and Packet Switching

Learning Objective

Show a clear understanding of circuit switching and packet switching – their operation, key characteristics, advantages and disadvantages, the protocol‑stack perspective, and the contexts in which each is used.

1. Circuit Switching

1.1 What is Circuit Switching?

Circuit switching creates a dedicated communication path between two endpoints for the whole duration of a session. The path is formed from a series of physical or logical channels that remain reserved exclusively for that call until it is released.

1.2 Signalling Protocols

• SS7 (Signalling System 7) – the primary out‑of‑band signalling system used in public‑switched telephone networks (PSTN).
• ISUP (ISDN User Part) – a layer of SS7 that sets up, manages and releases voice circuits.

These protocols operate in the control plane and are separate from the data‑plane that carries the voice or video stream.

1.3 How Circuit Switching Works

1. Connection establishment – The signalling protocol reserves a continuous route through each switching node.
2. Data transfer – Once the circuit is set up, data flows continuously at a constant rate; no further routing decisions are needed.
3. Connection release – When the session ends, the reserved resources are freed for other users.

1.4 Protocol‑Stack Perspective

In a circuit‑switched network the data plane works essentially at the Physical and Data Link layers. There is no TCP/IP (or OSI Network/Transport) processing for the voice stream; the circuit itself guarantees delivery.

1.5 Key Characteristics

• Fixed bandwidth is allocated for the whole session.
• Latency is low and predictable because the path never changes.
• Resource utilisation can be poor if the user does not use the full bandwidth.

1.6 Simple Mathematical Model

If a link has total capacity C and supports n simultaneous circuits, each circuit receives an equal share B:

When a circuit is idle for a fraction p of the time, its effective utilisation U is:

These formulas illustrate why a circuit‑switched network can waste capacity when traffic is bursty.

1.7 Advantages

• Guaranteed bandwidth – No contention once the circuit is established.
• Predictable performance – Constant delay and jitter.
• Simplicity of protocol – Data is transmitted as a steady stream; no per‑packet routing decisions.

1.8 Disadvantages

• Inefficient utilisation – Idle periods waste the reserved bandwidth.
• Scalability limits – The number of simultaneous circuits is bounded by the link’s total capacity.
• Setup time – Establishing a circuit adds initial latency before any data can be sent.

1.9 Typical Use Cases

• Traditional public‑switched telephone networks (PSTN).
• Dedicated conference lines, emergency‑service radios, and other real‑time voice/video links that require strict QoS.

1.10 Illustrative Example

A 1 Gbps link can support 100 simultaneous voice calls, each needing 10 Mbps. If only 30 calls are active, 70 × 10 Mbps = 700 Mbps remain unused – a clear case of resource waste.

2. Packet Switching

2.1 What is Packet Switching?

Packet switching divides data into small, self‑contained units called packets. Each packet carries a header with routing information (source/destination addresses, sequence number, etc.) and is forwarded independently through the network. Resources are shared among many users; a link’s capacity is used only when packets are present.

2.2 How Packet Switching Works

1. Segmentation – The sender breaks the message into packets, adds a header, and stores them in a transmission buffer.
2. Routing (store‑and‑forward) – Each packet arrives at a router, which examines the header, looks up the next hop, and forwards the packet onto the appropriate outgoing link.
3. Queuing – If the chosen link is busy, the router temporarily stores the packet in a buffer.
4. Reassembly – The destination collects packets, orders them using the sequence numbers, checks for errors, and reconstructs the original message.

2.3 Role of the Router (mapped to OSI/Internet layers)

• Network layer (OSI 3 / Internet layer) – Header inspection (IP address), routing‑table lookup, TTL decrement.
• Data Link layer (OSI 2) – Frame encapsulation for the outgoing link, error detection (CRC).
• Transport layer (OSI 4 / Transport) – Congestion control (e.g., TCP’s sliding window) and flow control.

2.4 Concrete Router Example

Assume a router receives an IP packet with:

• Source IP = 10.0.0.5
• Destination IP = 192.168.1.20
• TTL = 64

The router:

1. Decrements TTL to 63 (Network layer).
2. Looks up 192.168.1.20 in its routing table and finds the next‑hop address 172.16.0.1 (Network layer).
3. Encapsulates the packet in an Ethernet frame with the MAC address of the next hop (Data Link layer).
4. Places the frame in the output queue; if the link is busy, the packet waits until the link becomes free (Transport‑layer congestion handling may trigger TCP flow‑control signals).

2.5 Protocol‑Stack Perspective

Packet‑switched networks use the full TCP/IP (or OSI) model:

• Application layer – e.g., HTTP, SMTP, VoIP.
• Transport layer – TCP (reliable) or UDP (unreliable).
• Internet layer – IP addressing and routing.
• Link layer – Ethernet, Wi‑Fi, PPP, etc.

2.6 Key Characteristics

• Bandwidth is shared; a link is used only when packets are present.
• Latency is variable – it depends on propagation, transmission, queuing and processing delays.
• Efficient for bursty traffic because resources are allocated on demand.
• Supports statistical multiplexing – many users share the same physical links.

2.7 Simple Mathematical Model (Packet Delay)

The total delay D for a packet can be approximated as:

• d_prop – Propagation delay (distance ÷ signal speed).
• d_trans – Transmission delay = packet length ÷ link bandwidth.
• d_queue – Time spent waiting in router buffers (depends on traffic load).
• d_proc – Processing delay inside the router.

2.8 Advantages

• High utilisation – Links are used only when there is data to send.
• Scalability – Large numbers of users can share the same infrastructure.
• Robustness – If one path fails, packets can be rerouted dynamically.
• Flexibility – Supports many different services (email, web, streaming, VoIP) over the same network.

2.9 Disadvantages

• Variable latency and jitter – Queuing and routing delays can fluctuate.
• Potential for packet loss – Congested routers may drop packets, requiring retransmission.
• Complex protocols – Need error detection, flow control, and congestion control (e.g., TCP).

2.10 Typical Use Cases

• Internet traffic – email, web browsing, file transfer, streaming video.
• Voice over IP (VoIP) – voice is packetised and sent over the same network as data.
• Mobile data networks (4G/5G) and LAN/WAN infrastructures.

2.11 Illustrative Example

Sending a 120 KB email over a 100 Mbps link:

1. The email is split into 1500‑byte packets (≈ 80 packets).
2. Each packet is routed independently; some may travel via Router A, others via Router B.
3. Because the link is shared, packets are interleaved with traffic from other users, but the total bandwidth used at any instant is only the sum of the packets currently on the link.
4. At the receiver, packets are reordered using their sequence numbers and reassembled into the original 120 KB message.

3. Comparison of Circuit Switching and Packet Switching

4. Modern Hybrid Approaches

• VoIP – Voice is packetised and sent over a packet‑switched network, but Quality‑of‑Service (QoS) mechanisms such as DiffServ, MPLS or traffic‑class marking are used to emulate the low‑latency guarantees of circuit switching.
• Cellular networks – 4G/5G employ packet switching for data, while voice may be carried over a dedicated logical channel (circuit‑like) or via VoIP (packet).
• Software‑Defined Networking (SDN) – Allows dynamic creation of virtual circuits on top of a packet‑switched fabric, blending the benefits of both techniques.

5. Network Devices & Topologies (relevant to both switching types)

6. Suggested Diagrams (for classroom use)

• Figure 1 – Circuit Switching: A straight line of switches showing a single, dedicated path from Sender A to Receiver B.
• Figure 2 – Packet Switching: Multiple packets travelling from Sender C to Receiver D, taking different routes through a mesh of routers.
• Figure 3 – Router Header Processing: Shows an IP packet entering a router, TTL decrement, routing‑table lookup, and Ethernet encapsulation.

7. Summary

Circuit switching guarantees a fixed, low‑latency channel by reserving resources for the whole session, making it suitable for applications that need strict QoS. However, it can waste bandwidth when traffic is intermittent. Packet switching shares network resources among many users, maximising utilisation and providing robustness, but introduces variable latency and requires more sophisticated protocols that span the full TCP/IP (or OSI) stack. Modern networks often combine both ideas – e.g., VoIP over IP with QoS, or SDN‑created virtual circuits – so a solid grasp of the trade‑offs between the two paradigms is essential for designing and analysing contemporary communication systems.