Behaviour of TCP in the European ATM Pilot

Bonaventure, Olivier; Klovning, Espen; Danthine, André

doi:10.1016/0140-3664(96)01058-4

Cited by 8 publications

(7 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We choose 1000 cells as the minimum value for the MBS as we have already discussed in a previous paper [BKD96] how TCP behaves when a UPC enforces a MBS corresponding to only a few packets.…”

Section: Measurements With Tcpmentioning

confidence: 99%

“…When a TCP sender receives three duplicate acknowledgements pointing to the same data packet, it assumes that the first unacknowledged packet has been lost, and this packet is selectively retransmitted. This solution works weil provided that the window size is sufficiently large [BKD96], 8Jld that no more than a single data packet is lost within a single window. If these conditions are not fulfilled, TCP will have to wait for the expiration of its retransmission timer, and this may cause large idle times.…”

Section: Other Retransmission Strategiesmentioning

confidence: 99%

“…It includes the slow-start and congestion avoidance algorithms, but Iacks the window extensions and timestamp options defined in [JBZ92]. The implementation of the fast retransrnit algorithm [Ste94] is incomplete in SunOS 4.1.3_U1 [BKD96], but this bug had been fixed for our measurements. The ASX-200BX is a 2.5 Gbps bus-based non-blocking outputbuffered ATM switch.…”

Section: Measurement Environmentmentioning

confidence: 99%

“…Instead we measured the Inter-AAL-SDU time with a modified (but older-release 2.2.9) Fore ATM driver. This modified driver [BKD96] is able to log in a kerne! table a timestamp and some related information for each AAL-SDU sent or received by the driver.…”

Section: •mentioning

confidence: 99%

“…Cells from the last packets of this burst were discarded by the UPC, and TCP had to wait for the expiration of the retransmission timer to retransmit the lost packets. As the minimum value for TCP's retransmission timer is two 500 msec periods, the sender had to wait for approximately one second between two bursts of packets [BKD96].…”

Section: Explanation For the Large Idle Timesmentioning

confidence: 99%

See 4 more Smart Citations

Is VBR a solution for an ATM LAN ?

Bonaventure

Klovning

Danthine

1997

Protocols for High-Speed Networks V

Self Cite

View full text Add to dashboard Cite

W e present and discuss detailed measurem!'lnts showing how TCP behaves in an A TM LAN where a Variable Bit Rate (VBR) traffic contract is enforced by a Usage Parameter Control (UPC) mechanism. These measurements show that TCP has a Iot of difficulties to adapt its behaviour to such a traffic contract. In our environment, TCP only achieved a 10% utilisation of a 10 Mbps VBR VC. We then discuss and eva!uate possible solutions to improve this low utilisation of the reserved bandwidth.

show abstract

Section: Measurements With Tcpmentioning

confidence: 99%

Section: Other Retransmission Strategiesmentioning

confidence: 99%

Section: Measurement Environmentmentioning

confidence: 99%

Section: •mentioning

confidence: 99%

Section: Explanation For the Large Idle Timesmentioning

confidence: 99%

See 3 more Smart Citations

Is VBR a solution for an ATM LAN ?

Bonaventure

Klovning

Danthine

1997

Protocols for High-Speed Networks V

Self Cite

View full text Add to dashboard Cite

show abstract

Behaviour of XTPX in the European ATM pilot

Klovning

Bonaventure

1996

Trans Emerging Tel Tech

View full text Add to dashboard Cite

Abstract. We present and discuss measurements performed with XTPX over the European ATM Pilot. The main characteristic of this wide area ATM network is that the traffic contract is enforced by a Usage Parameter Control (UPC) mechanism. When the ATM traffic is compliant with the traffic contract, XTPX achieves a good utilisation of the link. However, when the ATM traffic is not entirely compliant, the UPC discards cells and the link utilisation becomes very low.

show abstract

RACE “MARS” project for the experimentation of multimedia interactive services: Architecture solutions and results

Bertin¹,

Sousa

Figuera

et al. 1997

Trans Emerging Tel Tech

View full text Add to dashboard Cite

This paper describes the architecture implemented in five different countries to vaIidate, by means of tests carried out with expert users, a service architecture for the interactive retrieval of multimedia information. The service architecture definition (which corresponds to the DAVIC 1.0 Specifications) and all activities for the set-up and running of the experiments, have been carried out within the RACE R2120 "Multimedia Audiovisual Retrieval Services" (MARS) project. A unique server and terminal architecture has been developed and put in opention at each site, while a combination of different public network solutions were used for the terminal access side. All national experiments made use of an ATM-based transport network. The interconnections between the five countries were provided by the European ATM Pilot (EM). The minimal testbed configuration implemented in each country consisted of one server and two terminals. After the integration and test phases, this configuration was extensively tested within the national environment to obtain the users' perception of the service.

show abstract

Behaviour of TCP in the European ATM Pilot

Cited by 8 publications

References 15 publications

Is VBR a solution for an ATM LAN ?

Is VBR a solution for an ATM LAN ?

Behaviour of XTPX in the European ATM pilot

RACE “MARS” project for the experimentation of multimedia interactive services: Architecture solutions and results

Contact Info

Product

Resources

About