The sensory traffic in Machine-to-Machine (M2M) communications has fairlyheterogeneous service delay requirements. Therefore, we study thedelay-performance of a heterogeneous M2M uplink from the sensors to a M2Mapplication server (AS) via M2M aggregators (MA). We classify the heterogeneousM2M traffic aggregated at AS into multiple Periodic Update (PU) and EventDriven (ED) classes. The PU arrivals are periodic and need to be processed by aprespecified firm service deadline whereas the ED arrivals are random with firmor soft real-time or non real-time service requirements. We use step andsigmoidal functions to represent the service utility for PU and ED packetsrespectively. We propose a delay efficient multiclass packet schedulingheuristic that aims to maximize a proportionally fair system utility metric.Specifically, the proposed scheduler prioritizes service to ED data whileensuring that the PU packets meet their service deadline. It also minimizessuccessive PU failures for critical applications by penalizing theiroccurrences. Furthermore, the failed PU packets are immediately cleared fromthe system so as to reduce network congestion. Using extensive simulations, weshow that the proposed scheduler outperforms popular packet schedulers and theperformance gap increases with heterogeneity in latency requirements and withgreater penalty for PU failures in critical applications.
展开▼