Cross-Layer Latency Compression and Spectrum Compaction Framework for Event-Triggered Traffic in HRLLC Industrial Network

The convergence of hyper-reliable and low-latency communication (HRLLC) with industrial networking has been envisioned as the potential candidate for cyber-physical systems deployments, which increases the flexibility without moderating the stringent requirements of latency and reliability. Throughout the industrial network, event-triggered (ET) traffic bears serious performance degradation of timeliness, stemming from the resource preemption by time-critical time-triggered (TT) traffic under prioritized scheduling mechanisms. Furthermore, wireless links changes and finite blocklength regime in HRLLC result in transmission error, violation of latency requirement, and conservative resource allocation. Consequently, a cross-layer optimization framework is established to compress the total end-to-end (E2E) latency and bandwidth of ET flows under quality-of-service (QoS) constraints. The reliability is characterized with the decoding error probability with frequency diversity in finite blocklength regime, and the timeliness is characterized with the experienced E2E latency considering time offsets and routing of TT flows. Through spatial non-overlapping path planning that balances shorter length and fewer links contention, and physical optimal resource allocation that refines QoS-aware assignments of bandwidth, the number of subchannels, and decoding error probability threshold, the total latency compression and spectrum compaction can be achieved. Simulation results show the distinct performance gain of proposed cross-layer optimization framework compared with other methods.