Dynamic Scheduling Algorithm for Delay-Sensitive Vehicular Safety Applications in Cellular Network

The vehicular safety applications disseminate the burst messages during an emergency scenario, but efforts to reduce the delay of communication are hampered by wireless access technology. As conventional VANET (Vehicular ad-hoc network) connected intermittently, the LTE (Long Term Evolution) based framework has been established for the vehicular communication environment. However, resource allocation which is affected by many factors, such as power, PRB (physical resource block), channel quality, are challenging to guarantee the safety services QoS (Quality of Service) in LTE downlink for OFDM (Orthogonal Frequency Division Multiplexing). In order to solve the problem of safety message dissemination in LTE vehicular network, we proposed a delay-aware control policy by leveraging a cross-layer approach to maximize the system throughput. First, we model the resource allocation problem using the queuing theory based on the MISO (multi-input single-output). Second, the method casts the problem of throughput and latency for dynamic communication system into a stochastic network optimization problem, and then makes tradeoffs between them by Lyapunov optimization technique. Finally, we use the improved the branch and bound algorithm to search for the optimal solution in system capacity region for these decomposed subproblems. The simulation results show that our algorithm can guarantee the delay while maximum system throughput.