基于数据链路层协作机制的SCTP协议改进及其仿真  被引量:1

Design and simulation on improving SCTP performance based on a collaborative approach

在线阅读下载全文

作  者:韦冬[1] 杨寿保[1] 鲁黎[1] 

机构地区:[1]中国科学技术大学计算机科学与技术系,安徽合肥230026

出  处:《中国科学技术大学学报》2006年第8期861-866,共6页JUSTC

基  金:国家自然科学基金(60273041);国家"863"计划(2002AA104560)资助

摘  要:流控制传输协议(stream control transmission protocol,SCTP)是IETF提出的一种用于有线网络的传输协议,该协议没有考虑无线网络具有的高误码率和较大延迟变化范围的特点.通过引入与802.11MAC层的协作机制对SCTP协议进行改进,使其可以准确地侦测到因无线链路误码而导致的数据报丢失,进而采用错误恢复算法进行处理,以避免拥塞窗口的盲目缩减.仿真实验表明:在异构网络中,改进后的协议能够显著地改善性能,获得更大的吞吐率.Stream control transmission protocol (SCTP) is a transport layer protocol proposed by IETF. It could be used for transport tasks on the IP-based network. SCTP was designed for wired networks as well as the well-known transport layer protocol TCP. The characteristics of wireless networks including high bit-error-rate and large delay variation were not considered in SCTP. SCTP was improved by introducing the collaboration with 802.11 MAC layer which made SCTP detect the packet losses caused by wireless bit error sensitively. When wireless loss occurred, the data link layer would retransmit the lost frame. If the number of transmission attempts reached the threshold, the frame would be discarded. And the data link layer would create an error report chunk for the upper layer-SCTP. SCTP could collect the information before SACK chunk returned or timeout process began. So SCTP modified the threshold of reduplicate SACKs which could trigger the fast retransmit phase which could deal with the situation using the recovery algorithm to avoid unnecessary shrinkage of the congestion window. Simulation results show that improved SCTP achieves evidently much better performance and larger goodput in heterogeneous networks. And the improved protocol also does well in fairness and friendliness compared with the original protocol.

关 键 词:拥塞控制 无线网络 流控制传输协议 无线链路误码 

分 类 号:TP393[自动化与计算机技术—计算机应用技术]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象