RES E AR C H Open Access
An FPGA-based high-speed network performance
measurement for RFC 2544
Yong Wang
1*
, Yong Liu
2
, Xiaoling Tao
3
and Qian He
3
Abstract
Aiming at the problem that existing network performance measurements have low accuracy for (Request for Comments)
RFC 2544, this paper proposes a high-speed network performance measurement based on field-programmable
gate array (FPGA). The active measurement method is used to generate probe data frames, and a passive measurement
method is employed to count network traffic. According to the statistical laws based on throughput variation, interval
stretching mechanism is used to dynamically adjust interframe gap. When our approach approaches the maximum
throughput, the network performance parameters are achieved. A prototype based on NetFPGA is also implemented
for evaluation. Experimental results show that our approach can be applied in high-speed network and the latency
can be accurate to the nanosecond. Compared with network performance measurement using software to send
probe data frames and a similar work based on FPGA, our approach can be more flexible and the evaluation data
aremoreaccurate.
Keywords: RFC 2544; Active measurement; Passive measurement; Traffic generator; Interval stretching; Network
performance measurement
1 Introduction
The explosive growth in Internet deployment for a con-
stantly growing variety of applications has created a
massive increase in demand for network performance
parameters, such as throughput, latency, and packet loss
rate [1-6], which are very important for providing differ-
entiated network services. Accurate network performance
parameters can help to improve the quality of network
services, including active and passive resource manage-
ment, traffic engineering, as well as providing quality of
service (QoS) guarantees for end-user applications. In par-
ticular, as modern network management systems shift
their focus forward service-level and application-level
management, the network monitoring process requires
more data to be collected in a higher frequency.
With the development of new network applications
and value-added services, network traffic characteristics
have become more and more complex. Different applica-
tions have different flow characteristics and behavioral
characteristics, so it is not enough to analyze only using
mathematical simulation and the classical queuing
theory mode. But through the result of the network per-
formance parameters and its analysis, we could simulate
the Internet environment accurately, which helps us to
optimize the network and design the network equipment
to a certa in extent.
Different network applications have different require-
ments for QoS. For example, file transfer services re-
quire low packet loss rate and high throughput, and
real-time multimedia services demand low latency [7].
Through network measurement, the users can detect
network congestion, locate network performance bottle-
necks, and provide the basis for the network resource
optimization.
Faced with an increasingly serious threat to network
security, the large-scale network measurements are used
to analyze and assess the network perfo rmance in abnor-
mal circumstances so early warning can be provided to
prevent large-scale network attacks. Therefore, network
measurement method also has become an important
mean to prote ct the network security and prevent large-
scale network attacks.
The most frequent used methods are ping and tracer-
oute. By calculating the time between sending the
* Correspondence: wang@guet.edu.cn
1
CSIP Guangxi Center, Guilin University of Electronic Technology, Guilin
541004, China
Full list of author information is available at the end of the article
© 2015 Wang et al. This is an Open Access article distributed under the terms of the Creative Commons Attribution License
(http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium,
provided the original work is properly credited.
Wang et al. EURASIP Journal on Wireless Communications and Networking 2015, 2015:2
http://jwcn.eurasipjournals.com/content/2015/1/2