Packet loss models of the Transmission Control Protocol

2006 ◽  
Author(s):  
Kaiyu Zhou
Author(s):  
Sandeep Kumar Gonnade ◽  
Naresh Kumar Nagwani

When the packet is reordered the most standard implementation of the Transmission control protocol (TCP) gives poor performance. This paper proposes a new version of the TCP which gives the high throughput when the packet reordering occurs and in another case if the packet reordering is not occurs then in that case also it is friendly to other version of the TCP. Transmission control protocol constant Packet Reordering (TCP-CPR) does not depend or rely on the duplicate acknowledgement to detect the packet loss. Instead the timer is used to maintain how long packet is transmitted. In this case timer is used to keep the track how long packets are transmitted. If acknowledgement are not received within the appropriate time then packet assume to loss because of the TCP-CPR does not depend on the duplicate acknowledgement. Packet reordering has does not effect on the performance of TCP-CPR. Through the performance of the TCP-CPR consistently better than existing mechanism that make the try to make the TCP more robust to packet reordering. In case where packets are not reordered, it’s verified that TCP-CPR maintains the same throughput as the typical implementation of TCP.


2006 ◽  
Vol 38 (4) ◽  
pp. 1132-1154
Author(s):  
Teunis J. Ott

The ‘square root formula’ in the Internet transmission control protocol (TCP) states that if the probability p of packet loss becomes small and there is independence between packets, then the stationary distribution of the congestion window W is such that the distribution of W√p is almost independent of p and is completely characterizable. This paper gives an elementary proof of the convergence of the stationary distributions for a much wider class of processes that includes classical TCP as well as T. Kelly's ‘scalable TCP’. This paper also gives stochastic dominance results that translate to a rate of convergence.


Author(s):  
Agus Halid ◽  
Reza Pulungan

AbstrakStream Control Transmission Protocol (SCTP) merupakan protokol yang mirip dengan Transmission Control Protocol (TCP) dan User Datagram Protocol (UDP). SCTP merupakan protokol yang bersifat reliable dan connectionless. Protokol ini memiliki kemampuan multistreaming dan multihoming dalam melakukan transmisi data. Penelitian ini merupakan pemodelan terhadap SCTP menggunakan simulator OPNET yang dapat menjadi akselerasi bagi peneliti dalam bidang jaringan. SCTP pada simulator dibangun dengan melakukan modifikasi terhadap TCP. Pemodelan dimulai dengan membangun skenario jaringan dan menentukan bandwidth pada jalur yang akan dilewati oleh paket data.Modifikasi ukuran window dalam penelitian ini menggunakan nilai 1 MMS, 2 MMS hingga 10 MMS pada pengendali kemacetan. Tujuannya adalah untuk melihat pengaruh modifikasi ukuran window terhadap nilai packet loss, delay dan throughput. Hasil pengukuran memperlihatkan bahwa nilai throughput tertinggi terdapat pada Skenario Kedua sebagaimana diperlihatkan pada Tabel 6.4 dengan nilai throughput sebesar 433.566,0244 bit/s. Penggunaan ukuran window dalam pengendali kemacetan dimaksudkan untuk menghindari banjir data pada sisi endpoint yang dapat menyebabkan packet loss. Kata kunci—Pengendali kemacetan, throughput, delay, packet loss, ukuran window, multihoming, SCTP  Abstract Stream Control Transmission Protocol (SCTP) is a protocol that is similar to the Transmission Control Protocol (TCP) and User Datagram Protocol (UDP). SCTP is a protocol that is both reliable and connectionless. This protocol has the ability multistreaming and multihoming in the transmit data.This research is the modeling of the SCTP using OPNET simulator that can be accelerated for researchers in the field of networking. SCTP on the simulator was built to perform modifications to TCP. Modeling starts with building a network scenarios and determine the bandwidth on the path that will be passed by data packets.Modification of window size in this research using 1 MMS, 2 MMS up to 10 MMS on congestion control. The aim is to disclose the effect of modification of the window size to the value packet loss, delay and throughput. The measurement results show that the throughput rate is highest in the Second Scenario as shown in Table 6.4 with throughput value of 433.566,0244 bits/s. Using window size in congestion control is intended to prevent a flood of data on the endpoint that can lead to packet loss. Keywords—Congestion control, throughput, delay, packet loss, window size, multihoming, SCTP 


Sign in / Sign up

Export Citation Format

Share Document