20201124-145836\Report_unicast_latency.pdf November 24, 2020 1
Unicast Latency Report - UDPDevice Tested
WLAN Switch Model: WLAN Switch Version: AP Model: FS-AP1167CAP SW Version: V200R106C60B202SP01
Overview
The latency test measures the delay incurred by frames passing through the system under test (SUT). It alsomeasures the amount of jitter, which is the variation in latency over many frames. Latency and jitter are keyperformance metrics that determine how well the SUT can handle traffic, such as voice or real-time video,that is sensitive to the delay between source and destination. This test measures latency and jitter accordingto RFC 2544 and RFC 3550, respectively.
Measured Latency
The following graph summarizes the measured minimum, maximum, and average latency performance of theSUT at the specified frame sizes, accumulated over all trials. Lower values indicate better performance. Also,a smaller difference between maximum and minimum latency indicates a better-performing SUT datapath.
0.06ms0.00ms
0.06ms 0.07ms 0.07ms 0.08ms0.00ms 0.00ms 0.00ms
2.49ms
2.72ms
2.22ms
2.34ms2.28ms
2.41ms
0.00ms 0.00ms 0.00ms
0.13ms 0.13ms 0.13ms 0.13ms 0.14ms 0.15ms
0.00ms 0.00ms 0.00ms
64 bytes @800pkt/sec
88 bytes @700pkt/sec
128 bytes @600pkt/sec
256 bytes @500pkt/sec
512 bytes @400pkt/sec
1024 bytes @300pkt/sec
1528 bytes @200pkt/sec
2048 bytes @100pkt/sec
2340 bytes @50.0pkt/sec
0.00ms
0.50ms
1.00ms
1.50ms
2.00ms
2.50ms
3.00ms
LatencyMINMAXAVG
Normal values for latency range from 1 to 1000 microseconds. Values in excess of 20 milliseconds are causefor concern, as they can pose problems for VoIP traffic.
Latency Histograms
The following latency histograms show the distribution of latency values produced by the SUT. Eachhistogram has 16 buckets (time ranges) into which the measured latency values are placed; the bucketboundaries are determined automatically during the test. Each bucket contains the number of measured
20201124-145836\Report_unicast_latency.pdf November 24, 2020 2
latency values that were less than the upper boundary of the bucket, and greater than the upper boundary ofthe preceding bucket. A separate histogram is plotted for each frame size, offered load and trial.
Ideally, all of the measured latency values should be clustered into as few buckets as possible, indicating aconsistent and uniform delay through the SUT datapath.
0 3
12755
2052
201 48 92 83 65 45 68 90 139 145 190 23
< 43.00
us
< 57.00
us
< 76.00
us
< 101.0
us
< 134.0
us
< 178.0
us
< 236.0
us
< 314.0
us
< 417.0
us
< 554.0
us
< 737.0
us
< 979.0
us
< 1.30
1ms
< 1.72
9ms
< 2.29
8ms
< 4.98
4ms
0
5000
10000
15000
Cou
nt
Latency Histogram Trial:1 Size:64 Rate:800.0 frames/sec
1 3
9255
3630
257 45 78 74 53 46 67 71 117 117 174 12
< 44.00
us
< 58.00
us
< 77.00
us
< 102.0
us
< 135.0
us
< 180.0
us
< 238.0
us
< 316.0
us
< 420.0
us
< 557.0
us
< 739.0
us
< 980.0
us
< 1.30
1ms
< 1.72
6ms
< 2.29
0ms
< 5.43
6ms
0
2000
4000
6000
8000
10000
Cou
nt
Latency Histogram Trial:1 Size:88 Rate:700.0 frames/sec
20201124-145836\Report_unicast_latency.pdf November 24, 2020 3
0 0
9305
1869
156 21 28 57 66 40 48 75 78 95 127 35
< 44.00
us
< 58.00
us
< 77.00
us
< 101.0
us
< 134.0
us
< 177.0
us
< 234.0
us
< 310.0
us
< 410.0
us
< 542.0
us
< 717.0
us
< 948.0
us
< 1.25
3ms
< 1.65
7ms
< 2.19
1ms
< 4.43
4ms
0
2000
4000
6000
8000
10000
Cou
ntLatency Histogram Trial:1 Size:128 Rate:600.0 frames/sec
0 9
8597
785
69 5 41 61 56 35 28 52 65 78 118 1
< 50.00
us
< 66.00
us
< 87.00
us
< 114.0
us
< 150.0
us
< 196.0
us
< 258.0
us
< 339.0
us
< 445.0
us
< 585.0
us
< 768.0
us
< 1.00
8ms
< 1.32
5ms
< 1.74
0ms
< 2.28
6ms
< 4.68
4ms
0
2000
4000
6000
8000
10000
Cou
nt
Latency Histogram Trial:1 Size:256 Rate:500.0 frames/sec
20201124-145836\Report_unicast_latency.pdf November 24, 2020 4
0 0
6647
820
83 19 23 71 26 26 29 33 49 68 93 13
< 54.00
us
< 70.00
us
< 91.00
us
< 119.0
us
< 156.0
us
< 204.0
us
< 266.0
us
< 348.0
us
< 455.0
us
< 594.0
us
< 776.0
us
< 1.01
3ms
< 1.32
4ms
< 1.73
0ms
< 2.26
0ms
< 4.55
8ms
0
2000
4000
6000
8000
Cou
ntLatency Histogram Trial:1 Size:512 Rate:400.0 frames/sec
0 1
4668
814
92 34 88 58 20 16 18 33 38 52 67 1
< 61.00
us
< 79.00
us
< 103.0
us
< 133.0
us
< 173.0
us
< 224.0
us
< 291.0
us
< 378.0
us
< 491.0
us
< 637.0
us
< 827.0
us
< 1.07
3ms
< 1.39
3ms
< 1.80
8ms
< 2.34
7ms
< 4.81
4ms
0
1000
2000
3000
4000
5000
Cou
nt
Latency Histogram Trial:1 Size:1024 Rate:300.0 frames/sec
20201124-145836\Report_unicast_latency.pdf November 24, 2020 5
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
< 1.00
0us
< 2.00
0us
< 3.00
0us
< 4.00
0us
< 5.00
0us
< 6.00
0us
< 7.00
0us
< 8.00
0us
< 10.00
us
< 11.00
us
< 12.00
us
< 13.00
us
< 14.00
us
< 15.00
us
< 16.00
us
< 32.00
us0.0
0.001
Cou
ntLatency Histogram Trial:1 Size:1528 Rate:200.0 frames/sec
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
< 1.00
0us
< 2.00
0us
< 3.00
0us
< 4.00
0us
< 5.00
0us
< 6.00
0us
< 7.00
0us
< 8.00
0us
< 10.00
us
< 11.00
us
< 12.00
us
< 13.00
us
< 14.00
us
< 15.00
us
< 16.00
us
< 32.00
us0.0
0.001
Cou
nt
Latency Histogram Trial:1 Size:2048 Rate:100.0 frames/sec
20201124-145836\Report_unicast_latency.pdf November 24, 2020 6
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
< 1.00
0us
< 2.00
0us
< 3.00
0us
< 4.00
0us
< 5.00
0us
< 6.00
0us
< 7.00
0us
< 8.00
0us
< 10.00
us
< 11.00
us
< 12.00
us
< 13.00
us
< 14.00
us
< 15.00
us
< 16.00
us
< 32.00
us0.0
0.001
Cou
ntLatency Histogram Trial:1 Size:2340 Rate:50.0 frames/sec
20201124-145836\Report_unicast_latency.pdf November 24, 2020 7
Test Conditions
Parameter Value Description
Frame Sizes [64, 88, 128, 256, 512, 1024, 1528, 2048, 2340] Frame sizes in bytes
ILOAD [800, 700, 600, 500, 400, 300, 200, 100, 50] Traffic load, frames/sec
Test Configuration
Parameter Value Description
Learning Time 2 secTransmission time (seconds) for initial learningpackets, to allow the SUT to set up forwarding tables
Transmit Time 20 sec Trial duration (seconds) - i.e., duration of test traffic
Settle Time 2 sec Idle time after test traffic transmission completes
Number of Trials 1Number of times measurements are repeated foraveraging
Prefer IPv6 addressing False
If this flag is enabled and clients has an IPv6 addressconfigured, then generated traffic will use IPv6addresses. If there is no IPv6 address configured onclients then generated traffic will use IPv4 addresses.
Test Topology
The test topology is shown below. Traffic is transmitted in the direction of the arrows. The test client portidentifiers and IP addresses are indicated in the boxes, together with the security mode and channel ID forWLAN clients.
192.168.1.111_card1_port1AUTO
192.168.1.111_card2_port12400SUT
192.168.1.11 192.168.1.10 No security
A total of 2 ports were used in this test.
20201124-145836\Report_unicast_latency.pdf November 24, 2020 8
Client Configuration
Client Group
RxDataMCS
TxDataMCS
TxMgmt.PHYRate(Mbps)
PHY Rate(Mbps) IPv6 Port
AdoptTxDataMCS
Group_001 15 N/A 24.0 300.0 Disabled 192.168.1.111_card2_port1 N/A
Group_002 N/A N/A N/A 1000 Disabled 192.168.1.111_card1_port1 N/A
Client MAC Configuration
Client GroupPHYType A-MPDU
RXA-MSDU
TX A-MSDU(Max, Target) LDPC
ChannelBandwidth
GuardInterval
ChannelModel
DynamicBandwidth
Group_001 11n On On On (Auto) Off 40 short Bypass N/A
Client MIMO Configuration
Client Group MIMO MU/SU-MIMO
Group_001 2x2 N/A
Client 802.11k Configuration
Client Group 802.11k Measurements Enabled Periodicity Max. Frames
Group_001 Off None N/A N/A
Methodology
The test is performed by associating test clients with the SUT ports, performing any desired learningtransmissions, and then generating test traffic between the test clients. Proprietary timestamps inserted ineach test traffic frame are then used to calculate the minimum, maximum and average latency as per RFC2544, as well as the smoothed interarrival jitter according to RFC 3550. The results are recorded separatelyfor each combination of test conditions, as well as for each trial if multiple trials are run. Results frommultiple trials are averaged into the graphs shown above.
Different intended loads (ILOADs) and frame sizes can be set up, to understand how latency varies withdifferent types of traffic in a real environment. Each combination of ILOAD and frame size is testedseparately. Test traffic may be configured to flow either from Ethernet to wireless, from wireless to Ethernet,or from wireless to wireless. If multiple APs are involved in the test, the ILOAD is divided evenly across theAPs; if multiple clients are associated with an AP, the ILOAD for that AP is divided evenly between theclients.
Latency measurements are made accurately even in the presence of frame loss. However, the ILOAD shouldbe set such that no frame loss occurs; otherwise, buffer occupancy delays can obscure actual SUT datapathdelays. The throughput test may be used to determine this traffic level.
20201124-145836\Report_unicast_latency.pdf November 24, 2020 9
Detailed Results
FrameSize
FrameRate
TrialNumber
MinimumLatency
MaximumLatency
AverageLatency
AverageJitter
64 800.0 1 56.00us 2.492ms 128.0us 98.03us
88 700.0 1 0.000s 2.718ms 129.0us 143.0us
128 600.0 1 58.00us 2.217ms 125.0us 127.5us
256 500.0 1 65.00us 2.342ms 127.0us 129.3us
512 400.0 1 70.00us 2.279ms 135.0us 150.2us
1024 300.0 1 78.00us 2.407ms 149.0us 77.46us
1528 200.0 1 0.000s 0.000s 0.000s 0.000s
2048 100.0 1 0.000s 0.000s 0.000s 0.000s
2340 50.0 1 0.000s 0.000s 0.000s 0.000s
Detailed latency histogram tables are not included in this report, but may be obtained from the file'Results_unicast_latency.csv' in the results directory.
Access Point Information
The following table shows the SUT details. The received signal strength indication (RSSI) from the SUT issampled on each port at the start of each trial and averaged over all of the trials.
Port Name Type RxAtt* Chan BSSID SSIDRSSI (dBm)A,B,C,D
192.168.1.111_card2_port1 80211ac off 1 7C:DD:76:00:DC:86 A2-TN/A, N/A, -14,N/A
192.168.1.111_card2_port1 80211ac off 1 7C:DD:76:00:DC:86 A2-TN/A, N/A, -14,N/A
192.168.1.111_card2_port1 80211ac off 1 7C:DD:76:00:DC:86 A2-TN/A, N/A, -14,N/A
192.168.1.111_card2_port1 80211ac off 1 7C:DD:76:00:DC:86 A2-TN/A, -19, N/A,N/A
192.168.1.111_card2_port1 80211ac off 1 7C:DD:76:00:DC:86 A2-TN/A, -19, N/A,N/A
192.168.1.111_card2_port1 80211ac off 1 7C:DD:76:00:DC:86 A2-TN/A, N/A, -14,N/A
192.168.1.111_card2_port1 80211ac off 1 7C:DD:76:00:DC:86 A2-TN/A, -19, N/A,N/A
192.168.1.111_card2_port1 80211ac off 1 7C:DD:76:00:DC:86 A2-TN/A, -19, N/A,N/A
192.168.1.111_card2_port1 80211ac off 1 7C:DD:76:00:DC:86 A2-TN/A, -19, N/A,N/A
192.168.1.111_card2_port1 80211ac off 1 7C:DD:76:00:DC:86 A2-TN/A, N/A, -14,N/A
20201124-145836\Report_unicast_latency.pdf November 24, 2020 10
Port Name Type RxAtt* Chan BSSID SSIDRSSI (dBm)A,B,C,D
192.168.1.111_card2_port1 80211ac off 1 7C:DD:76:00:DC:86 A2-TN/A, -19, N/A,N/A
192.168.1.111_card2_port1 80211ac off 1 7C:DD:76:00:DC:86 A2-TN/A, -19, N/A,N/A
192.168.1.111_card2_port1 80211ac off 1 7C:DD:76:00:DC:86 A2-TN/A, N/A, -15,N/A
192.168.1.111_card2_port1 80211ac off 1 7C:DD:76:00:DC:86 A2-TN/A, -19, N/A,N/A
192.168.1.111_card2_port1 80211ac off 1 7C:DD:76:00:DC:86 A2-TN/A, -19, N/A,N/A
192.168.1.111_card2_port1 80211ac off 1 7C:DD:76:00:DC:86 A2-TN/A, -19, N/A,N/A
192.168.1.111_card2_port1 80211ac off 1 7C:DD:76:00:DC:86 A2-TN/A, -19, N/A,N/A
192.168.1.111_card2_port1 80211ac off 1 7C:DD:76:00:DC:86 A2-TN/A, -19, N/A,N/A
The RSSI is measured at the WaveBlade SMA connector. RSSI values should be between -25 dBm and -35dBm for port types of 80211 and 80211n ports when the RX attenuation (RxAtt*) option is 'off'. For 80211nport types with attenuation 'on' the RSSI values at the port should be between -5 dBm and -15 dBm. If theRSSI is not in this range, modify the external attenuation to bring it into this range.
Port Configuration
The following table shows the port configuration details like Bandwidth, Channel, Band, CenterFrequency.
PortNamePortType Channel Band
ChannelBandwidth
CenterFrequency
192.168.1.111_card2_port1 80211ac 12.4GHz 40 MHz N/A MHz
Other Information
Results Directory
WaveApps Version
WaveTest Version
C:\Users\Dell\VeriWave\WaveApps\Results\20201124-145836
7.6, 2019.04.02.18-ixia
7.6-124-ixi, 2019.04.02.17
Ixia IxVeriWave8770 SW Nimbus Ave Beaverton, OR 97008(800) 457-5915 International: (503) 473-8350http://www.ixiacom.com/products/ixveriwave
Copyright 2019, Ixia, Inc. The Ixia logo, WaveInsite, WaveTest, WaveBlade,
WaveManager, and VCL are trademarks of Ixia, Inc. All other products
and services mentioned are trademarks of their respective companies.