基于数据分类和最小时延的LWA网络流量控制算法

无线异构网络已经成为未来无线通信网络演进的关键路径之一,多种无线网络互相融合亦是大势所趋.长期演进(longterm evolution,LET)与WLAN网络融合系统(LTE-WLAN aggregation,LWA)在改善系统频谱效率与能量效率以及提升网络系统容量与数据传输速率方面发挥着重要的作用.单一性地将LWA系统中分组数据集中协议(packet data convergence protocol,PDCD)数据单元分配给某个网络会增加网络负载和时延,所以LWA网络合理地流量控制方案愈发重要.为了提高系统吞吐量和能量效率,同时降低演进型基站(evolved nodeB,eNB)负载,提出一种基于数据分类和最小时延(data cassification and minimum delay,DCMD)的LWA网络流量控制算法.所提算法通过对传输数据进行分类并选择时延最小的网络传输对时延敏感的数据,而时延较大的网络则传输时延不敏感的数据,以此实现流量控制.仿真结果表明,所提算法能有效地提升系统的吞吐量、边缘地区用户终端(user e-quipment,UE)吞吐量和用户能量效率.

Data classification and minimum delay based flow control algorithm for the LWA network

Wireless heterogeneous network is becoming one of the key evolutions for wireless communication in the future, and the aggregation of various wireless networks is viewed as a major trend. The LWA system plays an important role in improving system spectrum efficiency and energy efficiency, and enhancing the network capacity and data transmission rate. Due to the fact that uniformly assigning PDCP data units to either LTE or WiFi nodes in the LWA network results in increased network load and latency, it is more important for the LWA system to operate within a reasonable flow control scheme. In order to ameliorate system throughput and energy efficiency, and alleviate the eNBs'' load burden, a data classification and minimum delay (DCMD) based flow control algorithm is proposed in this paper. The proposed algorithm first categorizes the data packets in terms of traffic types and evaluates the delays they may encounter in different flow path in LWA network. Inherently, the delay-sensitive data packets are directed to either LTE or WiFi network depending on whether the selected network is helpful for potentially imposing the smallest delay in transmission. In this sense, different traffic data, either delay-sensitive or delay-insensitive, will be guided along a different flow path. Simulation results show that the proposed algorithm can effectively improve the overall throughput of the system, the UE throughput in the edge region, and the system energy efficiency as well.