ABSTRACT:

Most normal implementations of TCP perform poorly when packets are reordered. In this paper, we tend to propose a replacement version of TCP that maintains high throughput when reordering occurs and nevertheless, when packet reordering does not occur, is friendly to other versions of TCP. The proposed TCP variant, or TCP-PR, will not rely on duplicate acknowledgments to detect a packet loss. Instead, timers are maintained to stay track of how long ago a packet was transmitted. In case the corresponding acknowledgment has not yet arrived and also the elapsed time since the packet was sent is larger than a given threshold, the packet is assumed lost. Because TCP-PR will not rely on duplicate acknowledgments, packet reordering (together with out-or-order acknowledgments) has no result on TCP-PR's performance. Through extensive simulations, we tend to show that TCP-PR performs consistently higher than existing mechanisms that try to make TCP a lot of strong to packet reordering. In the case that packets are not reordered, we tend to verify that TCP-PR maintains the identical throughput as typical implementations of TCP (specifically, TCP-SACK) and shares network resources fairly. Furthermore, TCP-PR only requires changes to the TCP sender facet creating it easier to deploy.