Fiber Network Redundancy: Ring vs Star Topology

Network Topology Fundamentals

The physical topology of a fiber optic network determines how data flows between nodes, how the network responds to cable or equipment failures, and how much fiber cable is required. For industrial SCADA networks connecting substations, pump stations, valve sites, and control centers, the topology choice directly impacts system availability, restoration time, and total cost. The two dominant topologies are ring and star (also called hub-and-spoke), with hybrid designs combining elements of both.

Ring Topology

Architecture

In a ring topology, fiber cable connects each node to two adjacent nodes, forming a continuous loop. Data can travel in either direction around the ring to reach any node. If a fiber cable is cut at any point, traffic automatically reroutes in the opposite direction around the ring to maintain connectivity to all nodes. This inherent redundancy makes ring topology the preferred architecture for SCADA networks where communication availability is critical.

Ring Protocols

Ring redundancy requires network protocols that manage path switching:

• RSTP (Rapid Spanning Tree Protocol): IEEE 802.1w standard. Convergence time of 1-5 seconds. Widely supported by all managed switch vendors. Suitable for SCADA networks where sub-second failover is not required.
• MRP (Media Redundancy Protocol): IEC 62439-2 standard. Convergence time under 200ms with MRP configured as ring manager. Designed specifically for industrial ring networks.
• eRSTP/Turbo Ring: Vendor-specific enhanced ring protocols from Moxa, Hirschmer, and others. Convergence times of 20-50ms. Proprietary but widely used in industrial SCADA networks.
• PRP/HSR: IEC 62439-3. Zero-loss redundancy by sending duplicate frames on both ring paths simultaneously. Required for protection and control applications where any data loss is unacceptable.

Ring Topology Advantages

• Single cable cut does not cause any node to lose connectivity
• Automatic failover without manual intervention
• Uses less total fiber cable than a star topology when nodes are geographically distributed along a line (pipeline, transmission corridor)
• Scales easily by inserting new nodes into the ring

Ring Topology Limitations

• All traffic shares the ring bandwidth. A heavily loaded node affects all other nodes on the ring.
• Under normal operation, half the ring bandwidth is reserved for the redundant path (unless using PRP/HSR which uses both paths simultaneously)
• Dual cable cuts that segment the ring into two isolated sections cause a complete network split
• Ring configuration is more complex than star, requiring proper protocol setup at each node

Star Topology

Architecture

In a star topology, each remote node has a dedicated fiber link back to a central hub (typically the control center or a core switch location). There is no interconnection between remote nodes. Each link operates independently, and a cable failure affects only the single node connected by that link.

Star Topology Advantages

• Each node has dedicated bandwidth not shared with other nodes
• A cable failure isolates only one node, not the entire network
• Simpler network configuration with no ring protocols required
• Easier troubleshooting since each link is independent
• Better suited for networks with a central data aggregation point

Star Topology Limitations

• No inherent redundancy. A single cable cut disconnects the affected node completely.
• Requires more total fiber cable when nodes are geographically distributed (each node needs its own home run to the hub)
• Central hub is a single point of failure for the entire network
• Adding redundancy requires duplicate cables on separate physical routes, doubling the fiber cable cost

Hybrid Topologies

Many industrial SCADA networks use hybrid topologies that combine ring and star elements:

• Ring backbone with star spurs: A fiber ring connects major facilities (substations, compressor stations, treatment plants), with individual star connections from ring nodes to nearby smaller sites (valve stations, well pads, pump stations).
• Dual star (redundant star): Two independent star networks from separate hub locations, each reaching every node via a different physical route. Provides ring-equivalent redundancy with star simplicity, but at higher cable cost.
• Interconnected rings: Multiple rings connected at two or more points, providing redundancy both within and between rings. Used in large utility networks with hundreds of nodes.

Topology Selection Criteria

The right topology depends on several factors:

• Availability requirements: If single-point-of-failure is unacceptable, ring or dual-star is required
• Geographic layout: Linear facilities (pipelines, transmission lines) favor rings. Centralized facilities (plant sites, campuses) favor stars
• Fiber cable cost: Compare total cable length for each topology option. Ring is often cheaper for linear layouts, star for clustered layouts
• Bandwidth per node: If individual nodes need high dedicated bandwidth, star topology avoids ring bandwidth sharing
• Operational complexity: Star is simpler to operate and troubleshoot. Ring requires proper protocol configuration and testing

NFM Consulting Network Design

NFM Consulting designs fiber network topologies based on rigorous analysis of each client's availability requirements, geographic constraints, and budget. We model multiple topology options with fiber cable quantity estimates, equipment costs, and availability calculations to identify the optimal design. Our network designs include detailed specifications for ring protocols, failover testing procedures, and maintenance access plans that ensure the fiber network delivers the reliability that SCADA systems demand.