Modeling the Stationary Behavior of TCP Reno Connections.
ABSTRACT In this paper, we outline a methodology that can be applied to model the behavior of TCP Reno flows. The proposed methodology
stems from a Markovian model of a single TCP source, and eventually considers the superposition and interaction of several
such sources using standard queueing analysis techniques. Our approach allows the evaluation of such performance indices as
throughput, queueing delay and segment loss of TCP flows. The results obtained through our model are validated by means of
simulation, under different traffic settings.
- SourceAvailable from: Carla-Fabiana Chiasserini
[Show abstract] [Hide abstract]
- "The postulated model is quite accurate in predicting TCP behavior , as shown in  by comparison with simulation results. "
ABSTRACT: Wireless access to the Internet requires that information integrity be preserved while transmitting data over the radio channel. Automatic repeat request (ARQ) schemes and transport control protocol (TCP) are often used as error-control techniques at the link layer and at the transport layer, respectively. In this paper, we study the interactions between an ARQ protocol and TCP when the last hop of the connection path is wireless. Link layer retransmissions are used to hide losses over the wireless link to TCP and to avoid that TCP interprets these losses as signs of network congestion. By adapting the protocol parameter setting at the link and transport layers to the time-varying characteristics of the wireless channel, we obtain a reliable communication link in spite of disadvantageous radio channel conditions. The number of retransmissions of the ARQ scheme and the TCP segment size can be set to proper values that, given certain propagation conditions, enable the desired quality of service (QoS) at the transport layer. In order to derive the protocol parameter setting fulfilling the QoS requirements, we develop detailed models of the ARQ protocol and TCP by using standard Markovian techniques. Through these models, the impact of the link layer functions on the TCP performance is evaluated in the presence of different wireless environments.IEEE Transactions on Vehicular Technology 12/2002; 51(6-51):1608 - 1620. DOI:10.1109/TVT.2002.804863 · 2.64 Impact Factor
[Show abstract] [Hide abstract]
- "For the CTMC model, which describes the TCP behavior , we employ the approach proposed in . "
ABSTRACT: Wireless access to the Internet requires that information integrity is preserved while transmitting data over the radio channel. ARQ schemes and TCP are often used as error-control techniques at the link layer and at the transport layer, respectively. We study the interactions between an ARQ protocol and TCP when a data traffic connection includes both wired and wireless links. By using standard Markovian techniques, we analyze the impact of different parameter settings of the ARQ scheme and of the radio channel conditions on the TCP performancePersonal, Indoor and Mobile Radio Communications, 2001 12th IEEE International Symposium on; 10/2001
- "Similarly, all the other transition rates are computed. For further details see . "
Conference Paper: Improving TCP over wireless through adaptive link layer setting[Show abstract] [Hide abstract]
ABSTRACT: Consider a communication link where the last hop is wireless and TCP is used as transport protocol over the end-to-end connection. We study the capability of the link layer to hide losses over the wireless link to TCP in spite of the time varying transmission quality. We focus on link-layer retransmission mechanisms and determine their parameter setting in such a way that a reliable communication link is provided. In particular, we choose a significant QoS metric at the transport layer and fixed its targeted value, and we adapt the maximum number of link-layer transmissions to the characteristics of the wireless link so that the desired QoS at the transport layer is provided. Results showing the impact of the link-layer retransmissions on the TCP performance are derived by using analytical models based on Markovian techniquesGlobal Telecommunications Conference, 2001. GLOBECOM '01. IEEE; 02/2001