针对交通监控中突发数据实时性问题,提出一种基于模糊聚类的媒体访问控制(FC-MAC)协议。该协议采用时分多址(TDMA)和改进的载波监听多路访问冲突避免(CSMA/CA)交替工作的方式,既保证了普通周期数据的传递,又增强了突发数据的实时性。在CSMA/CA阶段,提出模糊聚类分析的方法,根据因素向量聚类簇内节点,使节点突发数据具有不同的优先级,优先级高的突发数据更早接入信道完成传输;同时,根据该协议的时隙分配策略,提出一种基于分层随机延迟的方法,减少同一时段内竞争接入Sink节点的簇头数量,降低簇头节点之间因退避而产生的数据延迟。仿真结果表明:FC-MAC在能量消耗上介于混合型Z-MAC协议与调度型S-LMAC协议之间;在突发数据平均时延减少的情况下,网络吞吐量比Z-MAC提高了11.2%,比S-LMAC提高了21.3%,并且对网络业务流量具有更好的适应性。
In order to solve the real-time issue of burst data in traffic monitoring, an Medium Access Control (MAC) protocol based on fuzzy clustering called FC-MAC was proposed. This protocol works by alternating between Time Division Multiple Access (TDMA) and improved Carrier Sense Multiple Access with Collision Avoidance (CSMA/CA), which not only ensures periodic data transmission, but also enhances the real-time property of burst data. The method of fuzzy clustering was introduced to the phase of CSMA/CA, the nodes in a cluster were clustered according to factor vectors and assigned with different levels of priority according to their vectors, and burst data with higher priority was transmitted earlier. In addition, according to the timing slot allocation strategy of FC-MAC, a method of hierarchical random delay was presented to reduce the number of nodes accessing to sink node simultaneously and decrease the data delay caused by backoff mechanism. Simulation results show that the energy consumption of FC-MAC is between Z-MAC and S-LMAC. In the case of reducing delay of burst data, FC-MAC increases the network throughput by 11.2% and 21.3% respectively compared to Z-MAC and S-LMAC, and it is more adaptive to the change of network traffic.