Time-Sensitive Networking (TSN)

The foundational mechanism for all TSN capabilities, establishing a distributed sense of time across every device in the network. TSN relies on the IEEE 802.1AS profile, a simplified subset of the Precision Time Protocol (PTP) (IEEE 1588). A grandmaster clock, selected via the Best Master Clock Algorithm (BMCA), distributes a reference time to all TSN bridges and end-stations. This synchronization achieves accuracy in the sub-microsecond range, ensuring that every node agrees on when a specific cycle or time slot begins. Without this shared clock, scheduled traffic and latency guarantees would be impossible.

TSN enforces determinism by controlling when frames are transmitted. The key standard is IEEE 802.1Qbv, known as the Time-Aware Shaper. It divides communication into cyclical time slices, creating protected windows for time-critical traffic while non-critical traffic is blocked. A gate mechanism at each egress port opens and closes according to a pre-configured schedule. This ensures a high-priority control frame is never delayed by a large, low-priority data burst, effectively eliminating congestion-based jitter for scheduled streams.

Frame preemption, defined in IEEE 802.1Qbu, solves the problem of a time-critical frame waiting for a large, non-critical frame to finish transmission. It allows the transmission of a standard Ethernet frame to be paused, split into fragments, and resumed after an express high-priority frame has been sent. This mechanism dramatically reduces the worst-case latency for urgent control data, especially on lower-speed links where the transmission time of a single maximum-size frame could otherwise violate a control loop's deadline.

TSN provides proactive redundancy for ultra-high-reliability applications through IEEE 802.1CB (Frame Replication and Elimination for Reliability, or FRER). The source sends duplicate copies of a critical stream across multiple disjoint paths in the network. The destination eliminates the redundant copy, ensuring seamless failover with zero recovery time. This is combined with explicit path control via IEEE 802.1Qca, which allows a central controller to pre-engineer the exact routes through the network, avoiding single points of failure and guaranteeing path diversity.

To provide a guaranteed quality of service, TSN uses a centralized model for resource management. A Centralized Network Configuration (CNC) controller uses protocols like IEEE 802.1Qcc to discover the network topology and register stream requirements from talkers and listeners. The CNC calculates the necessary schedules and path reservations, then deploys these configurations to all TSN bridges. This contrasts with traditional IT networking's distributed control, enabling a fully deterministic and managed data plane where bandwidth and latency are contractually guaranteed for each critical stream.

TSN integrates with established industrial redundancy protocols to achieve bumpless failover. High-availability Seamless Redundancy (HSR) and Parallel Redundancy Protocol (PRP), standardized in IEC 62439-3, operate at the end-node level. A source node duplicates every frame and sends it over two independent, parallel networks. The destination node accepts the first frame and discards the duplicate. This provides zero-switchover-time redundancy that is completely transparent to the application, a critical requirement for safety-related systems and high-speed motion control.

The foundational synchronization mechanism for TSN, defined by IEEE 1588. PTP establishes a master-slave clock hierarchy across the network, distributing a grandmaster clock's time to all participating nodes with sub-microsecond accuracy. This absolute time reference is non-negotiable for TSN's time-aware shapers to execute coordinated, collision-free transmission schedules. Without PTP, the deterministic guarantees of TSN collapse.

A decentralized, data-centric middleware standard from the Object Management Group (OMG). DDS implements a global data space where publishers and subscribers are fully decoupled and anonymous. It offers rich Quality of Service policies for reliability and durability. In a TSN network, DDS's native real-time publish-subscribe model maps directly onto TSN's traffic classes, allowing complex autonomous systems to communicate deterministically without a central broker.

An architectural approach that decouples the control plane from the data plane. A centralized SDN controller has a global view of the network topology and programs the forwarding rules of TSN bridges dynamically. This integration is critical for centralized network configuration (CNC) in TSN, where the SDN controller computes and deploys the gating schedules for each queue on every bridge, enabling dynamic reconfiguration of deterministic paths.

A consolidated computing architecture where functions of different safety importance share a single hardware platform. TSN is the enabling network fabric for these systems, providing temporal and spatial isolation for traffic classes:

• Safety-critical control frames (high priority, bounded latency)
• Best-effort diagnostic streams (low priority)
• Audio/Video bridging traffic (bounded jitter) This consolidation eliminates the need for physically separate networks.

An international standard for distributed industrial automation that defines an event-driven function block model. Unlike the scan-based execution of IEC 61131-3, IEC 61499 applications are inherently distributed across networked devices. TSN provides the deterministic communication backbone that makes this distribution viable, ensuring that event-triggered function block invocations and their associated data exchanges complete within guaranteed time windows across the network.