Dynamic optical ethernet path reconfiguration for substation secondary systems based on software-defined networking

Abstract

The optical communication network of a substation secondary system carries protection, control, sampled-value, and monitoring traffic whose delay and availability directly affect safe grid operation. Static optical Ethernet paths cannot adapt quickly to fiber cuts, optical-transceiver alarms, or burst congestion, while controller-only rerouting may introduce excessive recovery latency. This paper proposes a software-defined networking (SDN) method for dynamic optical Ethernet path reconfiguration in digitally interconnected substations. A global controller continuously estimates link delay, residual bandwidth, utilization, packet loss, and reliability; generates a bounded set of candidate paths; and evaluates a normalized multi-constraint cost model. A lightweight feasibility-first heuristic accelerates the mixed-integer decision and installs flow rules through the southbound interface. An event-driven simulation containing 20 OpenFlow switches, 36 full-duplex fiber-optic Ethernet links, and mixed International Electrotechnical Commission (IEC) 61850 traffic was evaluated under 1000 random single-link failures and 200 double-link failures. The proposed method achieved a mean recovery time of 25.0 ms, a post-failure latency of 2.69 ms, packet loss of 0.86%, and 98.6% deadline satisfaction. Compared with reactive k-shortest-path rerouting, recovery time and packet loss decreased by 67.3% and 73.2%, respectively. The results show that topology-aware fiber-path reconstruction and explicit reliability constraints can maintain deterministic optical Ethernet service without excessive control-plane signaling.