CAJ | 학술논문

提出了一种基于时延的模糊控制TCP拥塞控制算法(DFCC), 采用当前RTT变化和历史变化相结合的方法,通过一个模糊控制器调节CWND,从而对网络拥塞情况有了更好的估计和反应,有效解决了突发拥塞降低吞吐量的问题.仿真试验表明,改进的算法更加准确地监测到网络拥塞,提高了网络的吞吐量.
제출료일충기우시연적모호공제TCP옹새공제산법(DFCC), 채용당전RTT변화화역사변화상결합적방법,통과일개모호공제기조절CWND,종이대망락옹새정황유료경호적고계화반응,유효해결료돌발옹새강저탄토량적문제.방진시험표명,개진적산법경가준학지감측도망락옹새,제고료망락적탄토량.
A fuzzy TCP congestion control algorithm is proposed. The algorithm achieves better evaluation on network congestion by the way of combining the current and historical RTT changes which are used to adjust congestion window (CWND) with fuzzy controller, so that the problem of low throughput under abrupt congestion situations is solved effectively. The performance of the algorithm has been tested and evaluated on NS simulator and the simulation results demonstrate that the algorithm inspects network congestion more exactly and accordingly increases network throughput.