The increasing speeds and decreasing prices of
network equipment have spurred much interest in applying
networks to many different types of systems. A wide range of
network protocols and tools already exist making its application
attractive. Over the last decade, there has been a noticeable shift
to introduce Internet Protocol (IP) networks into new areas such
as data acquisition systems. These modern network-based test
systems are composed of a large number of recognizable devices
such as switches, routers, and recorders. While these networks
contain components familiar to most, common enterprise
network test tools are not always sufficient to test these
specialized systems. Typically, these specialized network-centric
data systems require tighter timing accuracies, shorter latency
windows, and little to no acceptable packet loss. The unique
features of IP networks present testing challenges that must be
addressed to provide reliable verification of the data acquisition
system. Packet latency, delay, jitter, and loss are all factors that
must be accounted for. Intentionally introducing these types of
impairments to the data system are essential in any full, end-toend
system test. It is not only important to produce these types of
errors, but to mimic the conditions that will exist in the real
system and reliably reproduce them at any time. This paper
describes the shortcomings of those common network tools for
testing specialized network-based data acquisition systems and
the custom workarounds that were necessary to provide the test
capabilities that we required. We will share some of our lessons
learned from real-life examples of testing specialized network
equipment for a flight-test data acquisition system that would
apply to any network-centric data system. The paper will also
describe a portion of our lab setup that allows us to rapidly
reconfigure test networks and deploy custom simulation tools to
support multiple types of data acquisition systems
simultaneously
The increasing speeds and decreasing prices ofnetwork equipment have spurred much interest in applyingnetworks to many different types of systems. A wide range ofnetwork protocols and tools already exist making its applicationattractive. Over the last decade, there has been a noticeable shiftto introduce Internet Protocol (IP) networks into new areas suchas data acquisition systems. These modern network-based testsystems are composed of a large number of recognizable devicessuch as switches, routers, and recorders. While these networkscontain components familiar to most, common enterprisenetwork test tools are not always sufficient to test thesespecialized systems. Typically, these specialized network-centricdata systems require tighter timing accuracies, shorter latencywindows, and little to no acceptable packet loss. The uniquefeatures of IP networks present testing challenges that must beaddressed to provide reliable verification of the data acquisitionsystem. Packet latency, delay, jitter, and loss are all factors thatmust be accounted for. Intentionally introducing these types ofimpairments to the data system are essential in any full, end-toendsystem test. It is not only important to produce these types oferrors, but to mimic the conditions that will exist in the realsystem and reliably reproduce them at any time. This paperdescribes the shortcomings of those common network tools fortesting specialized network-based data acquisition systems andthe custom workarounds that were necessary to provide the testcapabilities that we required. We will share some of our lessonslearned from real-life examples of testing specialized networkequipment for a flight-test data acquisition system that wouldapply to any network-centric data system. The paper will alsodescribe a portion of our lab setup that allows us to rapidlyreconfigure test networks and deploy custom simulation tools tosupport multiple types of data acquisition systemssimultaneously
การแปล กรุณารอสักครู่..
