Computer Engineering and Applications ›› 2020, Vol. 56 ›› Issue (14): 93-97.DOI: 10.3778/j.issn.1002-8331.1904-0450
Previous Articles Next Articles
WANG Yaming, TANG Junyong, LEI Juchao
Online:
Published:
王亚铭,唐俊勇,雷聚超
Abstract:
In view of the large interference of the wireless network link and the high bit error rate, the TCP Westwood algorithm(TCPW) has excessive feedback on the packet when estimating the bandwidth, and lacks the shortcomings of distinguishing the type of packet loss during transmission. A TCPW congestion control optimization algorithm TCPW-F is proposed. The algorithm constructs the congestion factor by using the transmission rate and the like as the basis for judging the type of packet loss, and further adjusts the congestion window when the noise packet is lost, avoids the window drop caused by the unrelated noise packet loss, and improves the transmission efficiency of the window in this case. The simulation results show that the TCPW-F algorithm performs better in terms of delay performance, and the jitter per unit time tends to be stable. The packet generation rate is increased under the same channel bandwidth. The real-time throughput of the improved algorithm is significantly higher than that of the original algorithm. It has a certain noise packet loss sensing capability, and the TCP transmission quality of the wireless network is greatly improved.
Key words: wireless network, packet loss type, congestion control, NS3, Westwood
摘要:
针对无线网络链路干扰大、误码率高等特点,以及TCP Westwood算法(TCPW)存在估算带宽时过度依赖包的反馈,缺乏区分传输过程中丢包类型的缺点等问题,提出一种TCPW拥塞控制优化算法——TCPW-F。该算法利用发送速率等构建拥塞因子[F]作为判断丢包类型的依据,同时对判定发生噪声丢包时的拥塞窗口进一步调整,避免噪声丢包引起的窗口下降,提高该情况下窗口的发送效率。仿真结果表明,TCPW-F算法在时延性能方面表现更优,单位时间抖动趋于稳定的速度更快。在同一信道带宽下增大包生成速率,改进算法的实时吞吐量明显高于原算法,具备一定的噪声丢包感知能力,无线网络的TCP传输质量获得较大改善。
关键词: 无线网络, 丢包类型, 拥塞控制, NS3, Westwood
WANG Yaming, TANG Junyong, LEI Juchao. Wireless Network Congestion Algorithm for Noise Packet Loss Sensing[J]. Computer Engineering and Applications, 2020, 56(14): 93-97.
王亚铭,唐俊勇,雷聚超. 面向噪声丢包感知的无线网络拥塞算法[J]. 计算机工程与应用, 2020, 56(14): 93-97.
0 / Recommend
Add to citation manager EndNote|Ris|BibTeX
URL: http://cea.ceaj.org/EN/10.3778/j.issn.1002-8331.1904-0450
http://cea.ceaj.org/EN/Y2020/V56/I14/93