Agilent Technologies Home Page 8960 Series 10 Wireless Communications Test Set
- +
Home | Product Web Site | Contact Us
+
- -
+ -
8960 1xEV-DO Online User's Guide
E1966A, E1976A, E6706B
Search this guide for     
search tips


Feedback
Did the information on this web page help answer your question or solve a problem?

Yes
No
Did not apply

Please provide additional comments about this page.
(Optional, 255 character max)


Email address (optional):

Privacy Notice: We will not sell or give away your email address to any third party.

 

Packet Error Rate (PER) Measurement Description

Last updated: July 22, 2008

How is a PER Measurement Made?

Packet Error Rate (PER) is used to test the performance of an access terminal's receiver. PER is the ratio, in percent, of the number of FTAP or FETAP Test Packets not successfully received by the access terminal (AT) to the number of FTAP or FETAP Test Packets sent to the AT by the test set. See Forward Test Application Protocol (FTAP) , Multi-carrier Test Application Protocol and Forward Enhanced Test Application Protocol (FETAP) .

To perform a packet error rate measurement, the test set sends an FTAP (when current physical layer subtype is subtype 0) or FETAP Test Packet (when current physical layer subtype is subtype 2) or FMCTAP Test Packet (when current physical layer subtype is subtype 3) to the access terminal. See Release A Physical Layer Subtype or Release B Physical Layer Subtype to configure the physical layer subtype. Each packet contains 16 FCS (Frame Check Sequence) bit, which provide information about the packet. If the AT is able to decode the packet and the FCS checks (the information relayed by the FCS matches the packet characteristics), then the packet is successfully received.

The AT sends one or more FTAP/FETAP/FMCTAP Loop Back Packets to the test set to indicate how many FTAP/FETAP/FMCTAP Test Packets were successfully received for each 16-slot "observation interval" (frame). The AT sends one Loop Back Packet (or more, if needed to convey all of the records) containing a record for each FTAP/FETAP/FMCTAP Test Packet in the observation interval. The Loop Back Packets are queued for transmission on the Reverse Traffic Channel, and the AT must provide buffering for at least 8 FTAP/FETAP/FMCTAP Loop Back Packets. An FTAP/FETAP/FMCTAP Loop Back Packet is generated even if no FTAP/FETAP/FMCTAP Test Packets were received during the 16-slot observation interval.

In performing the PER measurement, the test set:

  • always operates with 100% packet activity (see Fixed Settings ). 100% packet activity means that all slots contain Forward Traffic Channel or Control Channel packets (data). The test set fills any empty slots with filler data directed to a random AT other than the AT under test. (0% packet activity means that the source is pulsed off when there is no data to transmit in the slot, and then pulsed on to transmit the MAC and Pilot Channels. This is not supported by the test set.)
  • always implements forced single encapsulation. Forced single encapsulation means that there is always only one valid MAC Packet per Physical Packet, regardless of the size of the Physical Packet. Depending upon data rate, the Physical Packet can be contain between 1 and 4 MAC Packets. If other MAC Packets are contained in the Physical Packet, they are loaded with all "F" data and the indicator bits are set to indicate that these packets are invalid.
  • by default, only directs half of the valid FTAP/FETAP/FMCTAP Test Packets to the AT under test, as per C.S0033 specifications. ( AT Directed Packets is set to 50% by default. However this parameter may be adjusted to optimize PER throughput based on AT performance. If Limited TAP is on while Test Application Protocol is set to FTAP, AT Directed Packets must not be greater than 50%, and RTAP Rate must be set to 9.6 kbps.)

You must establish an FTAP (for Subtype 0 Physical Layer) or FETAP (for Subtype 2 Physical Layer) or FMCTAP (for Subtype 3 Physical Layer) connection to perform packet error rate measurements (see Measuring Packet Error Rate (PER) for detailed procedure).

Confidence Level Testing

Confidence level testing is a feature of PER measurements that applies statistical analysis to PER measurements so that pass/fail test results can be obtained in the shortest possible time.

When confidence level is on, the PER measurement keeps track of the number of packets measured, the number of packet errors and performs a statistical analysis of the probability that the value entered in the PER requirement field will (or will not) be exceeded.

The confidence level parameter is typically set at 95%. With that setting, the PER measurement returns a pass result before the maximum packet count is reached if statistically there is a 95% probability that the percentage of packet errors will not exceed the PER requirement value if the measurement is allowed to reach the maximum packet count. The PER measurement returns a fail result before the maximum packet count is reached if statistically there is a 95% probability that the percentage of packet errors counted will exceed the PER requirement if the measurement is allowed to reach the maximum packet count. If the number of packet errors counted does not allow a reliable prediction of "pass" or "fail," the measurement continues until the number of packets tested equals max packets. "Max packets" and the PER percentage result are then returned.

If confidence level is off, the measurement always runs until the number of packets tested equals the specified maximum packets. "Max packets" and the PER percentage result are then returned.

During a fading test, the PER measurement must run for a sufficient length of time (the minimum number of packets) to allow fading statistics to average out (mean power standard deviation must be less than 0.2 dB). After the minimum number of packets have been measured, the statistical analysis is applied. This can prevent the PER measurement from failure too early due to power uncertainty. The early failure termination is not compatible with some tests (such as requiring faded signals) to achieve reliable results.

PER Measurement Parameters

  • Confidence State - Controls whether the confidence level will be used during the PER measurement. If confidence level is on, the PER measurement will continue until a pass/fail indication occurs, or until maximum packet count is reached. If confidence level is off, the PER measurement will run until maximum packet count is reached.
  • Confidence Level - Sets the required confidence level for the PER requirement (range is 80% to 99.99%).
  • PER Requirement - Sets the PER percentage that must not be exceeded for a "pass" result to be returned (range is 0.1% to 15%).
  • PER Target Slots - Sets the target slot number within which AT should successfully decode the forward test packets. It is used to calculate PER result by comparing with the 'FwdPhysSlots' in the loopback packets sent from AT: if 'FwdPhysSlots' is larger than the setting, the packet is counted as a packet error, otherwise the packet is counted as a valid packet. The range of the setting is 1-16 and the default value is 16.
  • Minimum Packet Count - Sets the minimum number of packets (range is 0 to 10,000,000 frames) to be tested before applying the statistic analysis when the Confidence State is on. It does not apply when Confidence State is off. See Confidence Level Testing for details. The minimum packet count is required for some faded test as specified in the C.S0033 standard.
  • Maximum Packet Count - Sets the maximum number of packets to be counted during the PER measurement (range is 25 to 10,000,000 packets). If the number of packets tested reaches Maximum Packet Count, the measurement will stop.
  • Control Channel Data Rate
  • AT Directed Packets
  • Max Forward Packet Duration
  • Forward Early Termination State
  • DRC Length
  • Trigger Arm - (See Trigger Arm (Single or Continuous) Description ).
  • Timeout Time/State - Range is 0.1 to 266,667.0 seconds. (See Measurement Timeouts ).

PER Measurement Results

  • Integrity Indicator - (see Integrity Indicator ).
  • PER - ratio, in percent, of the packet error count to the number of packets tested.
  • Confidence Pass/Fail - indicates whether the measurement achieved confidence of failure or success, or whether maximum packet count was reached.
  • FTC Throughput - in kbps, indicates the average throughput of Forward Traffic Channel according to AT's DRC request.
  • Packet Error Count - number of packets received with errors.
  • Packets Tested - number of packets tested.

Intermediate measurement results are periodically available on the front panel display, but are not available programmatically through the GPIB interface.

A typical display is shown below:

PER Input Signal Requirements

Input Signal Requirements

Top of pagetop of page     

+
Privacy Statement | Terms of Use | Webmaster  +  © Agilent 2000-2008
+