Network Topologies

Why This Matters

When the AP exam asks about networks, you're being tested on more than just naming shapes. You need to understand how network design choices affect fault tolerance, scalability, and performance. The College Board specifically emphasizes that networks should have redundant paths so that if one connection fails, data can still reach its destination. Every topology represents a different trade-off between these competing goals.

Network topologies are the blueprint for how computing devices connect and communicate. The exam frequently asks about single points of failure, redundancy, and why the Internet's design makes it so resilient. Don't just memorize which topology looks like what. Know why each design succeeds or fails when connections break, and which scenarios call for each approach.

---

Single-Path Architectures: Simple but Fragile

These topologies route data through a single pathway, making them cost-effective but vulnerable. When any link in the chain breaks, communication stops entirely.

Bus Topology

• Single central cable connects all devices. Data travels along the bus, and each device checks whether incoming packets are addressed to it.
• Single point of failure (SPOF). If the main cable breaks at any point, the entire network goes down. This directly violates fault tolerance principles.
• Limited scalability. As you add more devices, signal degradation increases and collisions become more frequent, making bus impractical for growing networks.

Ring Topology

• Circular pathway where each device connects to exactly two neighbors. Data travels in one direction around the loop (or both directions in a dual-ring setup).
• Ordered transmission control. Devices take turns transmitting (historically using a token-passing scheme), which avoids collisions. However, a single device or link failure can disrupt the entire ring.
• Predictable performance for small networks. Traffic flows in an orderly way, but adding or removing devices means reconfiguring the loop.

---

Centralized Architectures: Easy Management, Central Risk

These topologies funnel all traffic through a central device. The hub or root node simplifies management but creates a critical vulnerability.

Star Topology

• A central hub or switch connects to every device individually. Each connection is independent, so if one device or its cable fails, the rest of the network keeps working.
• Centralized management simplifies troubleshooting. You can isolate problems to a single connection. The trade-off is that the central hub itself becomes a single point of failure for the entire network.
• Easy to scale. Adding or removing a device just means plugging into (or unplugging from) the hub, which is much simpler than reconfiguring a bus or ring.

Tree Topology

• Hierarchical structure combining star and bus characteristics. A root node branches into multiple levels, with each branch potentially forming its own star.
• Segmentation supports large organizations. Different departments or floors can occupy separate branches, keeping the network organized and manageable.
• Root node failure is catastrophic. If the top-level node goes down, all downstream devices lose connectivity. Failures at intermediate nodes cut off everything below them as well.

---

Redundant Architectures: Built for Fault Tolerance

These topologies include multiple paths between devices, directly supporting the Internet's design principle that data should find alternate routes when connections fail.

Mesh Topology

• Every device interconnects with multiple (or all) other devices. If one path fails, data automatically routes through an alternative connection.
• Highest redundancy and fault tolerance. This exemplifies the Internet's resilience: when a link goes down, subsequent data takes a different route to reach its destination.
• Full mesh vs. partial mesh. In a full mesh, every device connects to every other device. With $n$ devices, that requires $\frac{n(n-1)}{2}$ connections, which gets expensive fast. Partial mesh strategically places redundant links only where they matter most, keeping costs manageable while still providing fault tolerance.

Hybrid Topology

• Combines two or more topology types, such as star-mesh or star-bus configurations tailored to specific network needs.
• Balances competing priorities. You can use mesh redundancy in critical areas (like a data center) while using simpler, cheaper topologies (like star) in less critical areas.
• Complex design and management. Requires careful planning to ensure redundant paths exist where fault tolerance is most essential.

---

Quick Reference Table

---

Self-Check Questions

1. Which two topologies both suffer from single points of failure but at different levels (cable vs. device)?

2. If a question describes a network where "subsequent data will be sent via a different route if a connection fails," which topology best demonstrates this principle, and why?

3. Compare and contrast star and tree topologies: what management advantage do they share, and how do their failure vulnerabilities differ?

4. A company needs high fault tolerance in their data center but wants to minimize cabling costs in office areas. Which topology approach would you recommend, and what concept does this illustrate?

5. Why does the College Board emphasize redundancy and multiple paths in network design? Identify which topologies meet this requirement and which do not.