Weighted Round Robin Configuration for Worst-Case Delay Optimization in Network-on-Chip

We propose an approach for computing the end-to-end delay bound of individual variable bit-rate flows in a FIFO multiplexer with aggregate scheduling under Weighted Round Robin (WRR) policy. To this end, we use network calculus to derive per-flow end-to-end equivalent service curves employed for computing Least Upper Delay Bounds (LUDBs) of individual flows. Since real time applications are going to meet guaranteed services with lower delay bounds, we optimize weights in WRR policy to minimize LUDBs while satisfying performance constraints. We formulate two constrained delay optimization problems, namely, Minimize-Delay and Multiobjective optimization. Multi-objective optimization has both total delay bounds and their variance as minimization objectives. The proposed optimizations are solved using a genetic algorithm. A Video Object Plane Decoder (VOPD) case study exhibits 15.4% reduction of total worst-case delays and 40.3% reduction on the variance of delays when compared with round robin policy. The optimization algorithm has low run-time complexity, enabling quick exploration of large design spaces. We conclude that an appropriate weight allocation can be a valuable instrument for delay optimization in on-chip network designs