This paper highlights the usefulness of tampered-TCP whose congestion control mechanism has been modified by malicious users for higher than the normal TCP throughput. The paper focuses on a tampered-TCP in which the increase and decrease ratio of the congestion window size were changed during the congestion avoidance phase. The performance of the tampered-TCP was examined based on a mathematical analysis and simulation given the coexisting of tampered-TCP and TCP Reno connections in the network.
This paper focuses on upper-layer applications requiring constant throughput, and proposed the TCP congestion control mechanism for achieving the required throughput with a high probability. Through simulation evaluations, it demonstrates that the proposed mechanism for one connection can achieve the required throughput with a high probability, even when there is almost no residual bandwidth of the network path, and that the extended mechanism performs effectively to provide the required throughput for multiple TCP connections.
This research paper performed an experimental study to assess the performance of high-speed TCP and parallel TCP variants in terms of utilization, throughput and fairness. Based on these experiments, a refined gHSTCP algorithm was proposed for its application in a real network. The results indicate that gHSTCP can offer a better tradeoff between utilization and fairness on LFNs. In the present paper, the performance of gHSTCP is evaluated only when the TailDrop mechanism is deployed at routers. Active Queue Management (AQM), such as Random Early Detection (RED), is an important queue management mechanism.
This paper analyzes the problem of packet collision that arises when TCP is used in ad hoc networks. It proposes a technique that combines data and ACK packets, and has shown through simulation that this technique can make radio channel utilization more efficient. In the simulation, the technique improved the TCP performance by up to 60% and by about 10% even when the network load was very high.