Adaptive Code In IEEE 802.11n Long-Distance Wireless Links

Long-Distance wireless Mesh networks is very suitable for deployment in remote or sparsely populated areas to access to broadband.But in long distance link,Long-Distance wireless Mesh networks,which based on IEEE 802.11 n,has a low utilization of channel and a poor performance in network,due to the limitation of original transmission of Stop-and-wait protocol and the loog time of retransmission after timeout.In this paper,based on large amounts of long-distance link experimental data,we have established RSSI(Received Signal Strength Indication)and IntraBER(Intraframe Bit Error Rate)model and RSSI and PDR(Packet Delivery Ratio)model for each MCS(Modulation and Coding Scheme).Therefore,based on the models,we come up with Adaptive Code,which uses RS(Reed-Solomon)code as intraframe code and LT(Luby Transform)code as interframe code and could dynamically adjust the intraframe and the interframe codingrate based on the current link status,to deal with bit error and packet loss with the minimum redundancy to avoid the cost of retransmission and improve the link throughput.Meanwhile,a new MAC layer based on Adaptive Code is designed,which change the transmission of traditional MAC layer.It transmits continuous frames and uses cumulative ACK manner to improve the utilization of channel.At last,we combines Adaptive Code and rate selection to achieve the best link throughput by selecting the MCS and intraframe/interframe codingate,when the status of link dynamic changes.Based on the real trace measured by test bed,we use Matlab to realize Adaptive Code.The simulation results show that in the fixed MCS,Adaptive Code gets 100% of decoding rate without retransmission,the coding rate stands at above 0.8 and the effective throughput of Adaptive Code increases by about 22% compared with the max measured effective throughput;the effective throughput of Adaptive Code is about 1.7 times compared with RainbowRate.