When you select a Session Loading With Mobility test, you can define the cycle used to connect and disconnect sessions, the node that will participate in the handoffs or the MN addresses used for handoffs, when the handoffs will begin, and the number of handoffs performed during the session loading run.
These instructions assume that you are working with a saved test session, and only provide information on defining this test activity. See Related Topics for instructions on configuring the test cases that can use a Session Loading With Mobility test.
TIP: Start with a test case that you have successfully executed with a Capacity Test. |
To set up a Session Loading With Mobility test:
Edit the test case... that will run the test.
Select a Session Loading With Mobility from the Test Activity drop-down list on the General tab.
Click the Settings... button, and a dialog window opens and displays the test activity parameters.
Define the session loading run and the handoff cycle
with Session Pending
Time..., Session
Hold Time..., Handoff Time..., and Number of Handoffs..., considering the type of
load you want to place on the SUT. In this case, the sessions remain
connected for the amount of time required to process the handoffs
plus the hold time.
A cycle with long hold and pending times can present a peak of activity
followed by a lull or a long duration of low to moderate activity.
A cycle with short hold and pending times presents a constant mix
of connect, handoff, and disconnect activities. If you check the Ensure Constant Session Loading...
box, the values you enter will be validated.
TIP: You can configure a test with an evenly balanced load, where sessions are constantly being connected, handed off, and disconnected, by setting the pending time and the handoff cycle plus the hold time to half of the time that it takes to establish all of the MN sessions. Example:Example: Assuming that the test connects 200,000 MN sessions (including all contexts or service instances) at a rate of 1,000 sessions/second, 200 seconds are required to connect all of the sessions. To achieve a balanced load, sessions must be active for 100 seconds and idle (Session Pending Time) for 100 seconds. During the 100 seconds that a session is active, handoffs are performed and then the session is maintained for the hold time. If 2 handoffs are performed with a Handoff Time of 30 seconds, 60 seconds are required to process the handoffs, leaving 40 seconds for the Hold Time. The hold timers in the first sessions start to expire while the remaining 100,000 sessions are being connected and handed off. By the time all of the 200,000 sessions have been connected for the first time, the pending timers in the first sessions start to expire, and they begin to be reconnected at the same rate that the later sessions are being disconnected. The result is a balanced session loading cycle with 100,000 active sessions and 100,000 idle sessions at any given time. Every second 1,000 sessions are connected, 1,000 sessions are handed off, and 1,000 sessions are disconnected. |
The following options are available depending on the test case:
Select the Mobility Type if different types of handoffs are supported.
In CDMA2000 test cases, you can choose Fast Handoff Support... or to Activate Dormant Session on Handoff...
In the UMTS test case, you can choose Perform Serving RNS Relocation Procedure... or Perform PS Handover Procedure... depending on the type of interface being tested.
Click OK to close the test activity window.
Define the handoff nodes on the General tab.
IMPORTANT: The addresses used for handoffs must be unique within the test session. |
If the handoff nodes utilize a control protocol, a separate tab is displayed for the settings associated with the handoff nodes. Define the settings on the handoff protocol tab.
Click OK to apply the test case settings. If you have entered a value outside of the allowable range you will receive an error.