VisualSim
provides a graphical modeling and simulation environment to conduct
traffic analysis and performance analysis. Using this environment, the user
can construct a queuing or statistical model with the pre-built modeling library.
The model can contain traffic generators, analysis tools and resources to resemble
the different devices. These resource blocks can emulate masters such as
processor and servers; intermediate channels such as Xon-Xoff flow control;
and slave devices. Credit policy can be described with knowledge of the statistics
such utilization and latency of forward devices. The model will generate statistics
for queue utilization, buffer occupancy, throughput and identify design issues such
as buffer overflow and system slow-down/point of failures.
Experiments by customers have shown accuracy
of over 90% with this VisualSim approach. Moreover, this analysis is done prior
to any implementation. Hence the rework cost is almost fully eliminated. This
example shows the application of a backward pressure based on a buffer occupancy
threshold of the Egress queue.
The
VisualSim model associated with this description is provided below.
You can view, change parameter values and run simulation right from
within the Web Browser. No additional software is required. This
shows how you can use a pre-built VisualSim model for doing trade-studies.
To use the models at the links,
click on the GO button to run the simulation. Double-click on any parameter in
the model window to change the parameter value.
Click
here to view the interactive VisualSim
Block Diagram and model (Model 1)
Click
here to view and execute the VisualSim
model (Model 2)
Note:
This model is a good example of QOS analysis where backward
pressure is based on certain algorithm
that depends on the details of the
buffer. This is also good for arbitration
and contention resolution in Ethernet
and Wireless LAN (802.11).
The model is scalable to easily vary the data size,
input traffic rate, threshold levels and queue depths.
Using a parameter value of the model, the number of
Ingress and Egress queues can be changed.
A traffic manager is modeled to
determine efficient parameters for
the forward queue processing
Round-Robin algorithm and to construct
an optimal back pressure to manage
congestion. The goal is to have
a QOS that ensures maximum latency
does not exceed the set threshold.
System Overview
The system is show in Figure 1.
12B frames are transmitted to the
traffic forward queues at the rate
of 200MHz. There are variable number
of forward queues upto a maximum
of 1024. The output from this set
of queues are frames of 36B at 200MHz.
This means that 3 frames are combined
at the forward queue before transmitting.
The combined frames are fed to another
set of queues that are again a variable
number upto a maximum of 64. The
queues in the forward section are
mapped to a specific ouput queue
and will transfer to that output
queue only. When the depth of the
output queue exceeds a threshold,
a back pressure is applied from
the output queues to the forward
queues to stop transmission.
Figure 1: Block Diagram
of the Traffic Management Systems
|
Model
Parameters:
Incoming frame size = 12B Incoming
rate = 200 MHz
Outgoing frame size = 36B Outgoing
rate = 200 MHz
Number
of forward queues = vary between
64-1024 Depth
of forward queues = 30 frames
Number of output queues = vary
between 8-64 Depth
of output queues = 20 frames
Backward
pressure trigger point = 7 frames
in the output queue
The
model is broken down into the following
sections
1. Traffic source: The source has a input rate with a uniform distribution.
2. Setting up the forward and output
queues
3. Writing to the forward queue
4. Removing from the output queues
5. Round-robin algorithm
6.
Applying Backward Pressure
7.
Reading from the forward queue and
placing them in the output queue along
with the estimated processing time
in the output queue: The process of
transferring the data from the forward
queue to the output queue is done
once the backward pressure has been
evaluated and the forward queue is
free to send the data. This is done
by first removing the data from the
Forward Queue, computing processing
delay at the output queue and then
placing the frames in the associated
output port queue along with the processing
time.
8. Statistics has been gathered on the forward
and output queues, and the latency
of each frame at the exit of the output
queue. The latency is computed by
calculating the time from the packet
generation to the exit of the output
queue ("DELTA = TNow - TIME").
