TD-LTE交通安全预警通信系统调度算法研究

Research on scheduling algorithm for TD- LTE traffic safety warning system

为向各类交通安全预警信息流的可靠传输提供更好的通信链路支持,将具有更高数据传输速率和更好频谱利用率的第4代移动通信技术(4 G)——时分长期演进(TD-LTE)技术引入交通安全预警通信系统。交通安全预警管理中心首先通过公网/专网将各类交通预警信息流递交至监管车辆隶属的TD-LTE基站(eNodeB)处,再由eNodeB将其进行分类映射,并快速、可靠的下达至监管车辆车载终端处,从而通过主动预警的方式有效避免和减少交通事故,保障生命安全,提高交通运行效率。在此基础上,建议一种适用于TD-LTE交通安全预警通信系统的时频资源调度模型及相关调度算法,建议算法在已有比例公平(PF)算法基础上,将时延调节因子引入时频资源调度函数,使得系统能够根据各类交通安全预警信息流的目标时延限定要求及其所在分类器的队头分组的等待时间,自适应地调整时频资源调度匹配指数,并总是为即将超出自身时延限定要求的交通安全预警信息流赋予优先获取时频资源的调度机会。仿真结果表明,建议算法与典型的PF算法均具有较好的公平性,建议算法能够将视频预警信息流和话音预警信息流的丢包率分别平均降低29%和48%,从而为监管车辆提供更加可靠地安全预警保障。

In order to provide all kinds of traffic safety warning information flow with better communica- tion link support for reliable transmission, this paper introduced the fourth generation mobile communication technology（4 G）, Time Division Long Term Evolution （TD -LTE ）technology into the traffic safety warning system, with higher transmission rate and better spectrum efficiency. Firstly, the traffic safety warning management center delivered all kinds of traffic safety warning information flow through the public/private network, to the TD - LTE base station （eNodeB） which the regulatory vehicles belonged to. Secondly, all kinds of traffic safety warning information flow were classified by the eNodeB, and then scheduled to the related regulatory vehicle terminals reliably as fast as possible, thus the traffic safety warning system can avoid and reduce traffic accidents, ensure the security of life, and improve the traffic efficiency effectively by means of active warning. At the same time, a time-frequency resource scheduling model and its corresponding scheduling algorithm for TD - LTE traffic safety warning system were proposed. Based on the existing Proportional Fair （PF） algorithm, this proposed scheduling algorithm introduced a kind of delay adjustment factor into the time-frequency resource scheduling function, which could make the system adaptively adjust this function value, according to the target delay limits of various traffic safety warning information and the waiting time of the packet which was in the head of line of every classifier. As a result, if the waiting time of the traffic safety warning information flow is beyond the target delay limit, this system would give it priority to obtain time-frequency resources for scheduling. Simulation results show that both the proposed algorithm and the typical PF algorithm have better performance of fairness, and the proposed algorithm can improve packet loss rate of the video warning infor- mation flow and the voice warning information flow by an average of 29% and 48% ,respectively,which provide more reliable security warning guarantee for the regulatory vehicles.