arXiv subjects

Nicolas Kuhn

Publications and source records attributed to Nicolas Kuhn.

6 recordsLinked to original sources

arXiv2021

Evaluating BDP FRAME extension for QUIC

The first version of QUIC has recently been standardized by the IETF. The framework of QUIC enables the proposition, negociation and exploitation of extensions to adapt some of its mechanisms. As one example, the DATAGRAM extension enables the unreliable transmission of data. The BDP FRAME extension is a method that can improve traffic delivery by allowing a QUIC connection to remember the knowledge of path characteristics and exploit them when resuming a session. This technical report presents the rationale behind fast convergence in SATCOM systems and evaluate the BDP FRAME extension in emulated and live environments.

cs.NI↗

arXiv2019

Google QUIC performance over a public SATCOM access

Google QUIC accounts for almost 10% of the Internet traffic and the protocol is not standardized at the IETF yet. We distinguish Google QUIC (GQUIC) and IETF QUIC (IQUIC) since there may be differences between the two. Both Google and IETF versions run over UDP and cannot be split the way satellite systems usually do with TCP connections. The need for adapting any-QUIC parameters needs to be evaluated. Since GQUIC is available, we analyze its behavior over a satellite communication system. In our evaluations, GQUIC quick connection establishment does not compensate an inappropriate congestion control. The resulting page downloading time doubles when using GQUIC as opposed to the performance with optimized split TCP connections. This paper concludes that specific tuning are required when any-QUIC runs over a high BDP network.

cs.NI↗

arXiv2016

Improving PIE's performance over high-delay paths

Bufferbloat is excessive latency due to over- provisioned network buffers. PIE and CoDel are two recently proposed Active Queue Management (AQM) algorithms, designed to tackle bufferbloat by lowering the queuing delay without degrading the bottleneck utilization. PIE uses a proportional integral controller to maintain the average queuing delay at a desired level; however, large Round Trip Times (RTT) result in large spikes in queuing delays, which induce high dropping probability and low utilization. To deal with this problem, we propose Maximum and Average queuing Delay with PIE (MADPIE). Loosely based on the drop policy used by CoDel to keep queuing delay bounded, MADPIE is a simple extension to PIE that adds deterministic packet drops at controlled intervals. By means of simulations, we observe that our proposed change does not affect PIE's performance when RTT < 100 ms. The deterministic drops are more dominant when the RTT increases, which results in lower maximum queuing delays and better performance for VoIP traffic and small file downloads, with no major impact on bulk transfers.

cs.NI↗

arXiv2013

On the Trade-off Between Spectrum Efficiency with Dedicated Access and Short End-to-End Transmission Delays with Random Access in DVB-RCS2

This paper analyses the performance of TCP over random and dedicated access methods in the context of DVB-RCS2. Random access methods introduce a lower connection delay compared to dedicated methods. We investigate the poten- tial to improve the performance of short flows in regards to transmission delay, over random access methods for DVB- RCS2 that is currently under development. Our simulation experiments show that the transmission of the first ten IP datagrams of each TCP flow can be 500 ms faster with ran- dom access than with dedicated access making the former of interest to carry Internet traffic. Such methods, however, are less efficient in regards to bandwidth usage than dedicated access mecanisms and less reliable in overloaded network conditions. Two aspects of channel usage optimization can be distinguished: reducing the duration of ressource utiliza- tion with random access methods, or increasing the spec- trum efficiency with dedicated access methods. This article argues that service providers may let low-cost users exploit the DVB-RCS2 to browse the web by introducing different services, which choice is based on the channel access method.

cs.NI↗

arXiv2013

Physical Channel Access (PCA): Time and Frequency Access Methods Emulation in NS-2

We present an NS-2 module, Physical Channel Access (PCA), to simulate different access methods on a link shared with Multi-Frequency Time Division Multiple Access (MF-TDMA). This tech- nique is widely used in various network technologies, such as satellite communication. In this context, different access methods at the gateway induce different queuing delays and available capacities, which strongly impact transport layer performance. Depending on QoS requirements, design of new congestion and flow control mechanisms and/or access methods requires evaluation through simulations. PCA module emulates the delays that packets will experience using the shared link, based on descriptive parameters of lower layers characteris- tics. Though PCA has been developed with DVB-RCS2 considerations in mind (for which we present a use case), other MF-TDMA-based appli- cations can easily be simulated by adapting input parameters. Moreover, the presented implementation details highlight the main methods that might need modifications to implement more specific functionality or emulate other similar access methods (e.g., OFDMA).

cs.NI↗

arXiv2012

Enabling Realistic Cross-Layer Analysis based on Satellite Physical Layer Traces

We present a solution to evaluate the performance of transport protocols as a function of link layer reliability schemes (i.e. ARQ, FEC and Hybrid ARQ) applied to satellite physical layer traces. As modelling such traces is complex and may require approximations, the use of real traces will minimise the potential for erroneous performance evaluations resulting from imperfect models. Our Trace Manager Tool (TMT) produces the corresponding link layer output, which is then used within the ns-2 network simulator via the additionally developed ns-2 interface module. We first present the analytical models for the link layer with bursty erasure packets and for the link layer reliability mechanisms with bursty erasures. Then, we present details of the TMT tool and our validation methodology, demonstrating that the selected performance metrics (recovery delay and throughput efficiency) exhibit a good match between the theoretical results and those obtained with TMT. Finally, we present results showing the impact of different link layer reliability mechanisms on the performance of TCP Cubic transport layer protocol.

cs.NI↗