Top Banner
White Paper 6 Best Practices for Launching VoWiFi, VoLTE and EVS Launching Next-Gen Voice Services Powered by new wireless and IP technology, we live in an increasingly connected world. Now we expect to be able to access services anytime and anywhere and seamlessly move between locations. Oh, and the quality needs to be great— everywhere! With the advent of Voice over Wi-Fi (VoWiFi) and Voice over LTE (VoLTE) and a range of supporting technologies including the Enhanced Voice Services (EVS) codec, Internet Multimedia Subsystem (IMS) and new “carrier- grade” Wi-Fi standards, providers are now able to deliver next-generation voice services with unprecedented quality and accessibility. The latest development in the next generation of voice services is the introduction of the EVS codec. EVS encodes input audio signals with a bandwidth of up to 20kHz, the full bandwidth of audio perceptible to the human ear. EVS- encoded speech is more faithfully reproduced than previous generation codecs such as AMR-WB and AMR–NB. In short, audio over EVS sounds more like the real thing. On the technical side the EVS codec also holds a lot of promise—it has been touted to perform better than previous generation codecs when signal levels are poor and to improve call quality for connections to lower bandwidth codecs.
13

6 Best Practices for Launching VoWiFi, VoLTE and EVS

Feb 22, 2023

Download

Documents

Khang Minh
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: 6 Best Practices for Launching VoWiFi, VoLTE and EVS

White Paper

6 Best Practices for Launching VoWiFi, VoLTE and EVSLaunching Next-Gen Voice ServicesPowered by new wireless and IP technology, we live in an increasingly connected world. Now we expect to be able to access services anytime and anywhere and seamlessly move between locations. Oh, and the quality needs to be great—everywhere! With the advent of Voice over Wi-Fi (VoWiFi) and Voice over LTE (VoLTE) and a range of supporting technologies including the Enhanced Voice Services (EVS) codec, Internet Multimedia Subsystem (IMS) and new “carrier-grade” Wi-Fi standards, providers are now able to deliver next-generation voice services with unprecedented quality and accessibility.

The latest development in the next generation of voice services is the introduction of the EVS codec. EVS encodes input audio signals with a bandwidth of up to 20kHz, the full bandwidth of audio perceptible to the human ear. EVS-encoded speech is more faithfully reproduced than previous generation codecs such as AMR-WB and AMR–NB. In short, audio over EVS sounds more like the real thing. On the technical side the EVS codec also holds a lot of promise—it has been touted to perform better than previous generation codecs when signal levels are poor and to improve call quality for connections to lower bandwidth codecs.

Page 2: 6 Best Practices for Launching VoWiFi, VoLTE and EVS

www.spirent.com

6 Best Practices for Launching VoWiFi, VoLTE and EVS

Up to 20 kHz

VoLTE EVS 3G AMR-WB 3G AMR-NB

Up to 7 kHz Up to 3.4 kHz

Figure 1: The EVS codec encodes audio inputs up 20kHz whereas previous generation codecs supported narrower bandwidths of 3.4 to 7kHz.

Over the past few years, Spirent has worked closely with industry’s first round of providers rolling out next-generation voice services. Based on that experience, Spirent has developed best practices for assuring the successful launch of next-generation voice services with a focus on VoLTE, VoWiFi and EVS-enabled voice services. This white paper details the key challenges providers face as they launch these services and shares Spirent’s recommended best practices and lessons learned. The whitepaper also illustrates key principles with test results of next-gen voice services from actual operational networks.

Assurance Challenges for Launching Next-Gen Voice ServicesEvaluating Inter-service, Inter-codec Calling (VoLTE & VoWiFi)

The first challenge we will examine is how to assess the performance of EVS voice services within a VoLTE network. To do this comprehensively it’s important to test how well the service works only when placing calls between EVS capable devices, but also when placing calls to the previous generation of devices that use narrow and wideband codecs including phones outside the mobile network via the PSTN (public switched telephone network).

In each of these scenarios different parts of the mobile network are exercised and different transcoding is required. It’s possible (and common) that the voice service may be working well in one scenario but exhibit an issue in another scenario. That means it’s critical to test VoLTE to VoLTE calls, VoLTE to 3G calls and VoLTE to Landline calls as well as making sure all the combinations of narrow and wideband codecs are exercised. An identical challenge exists for evaluating EVS services over VoWiFi: it’s critical to test VoWiFi to VoWiFi calls, Wi-Fi to 3G, VoWiFi to 3G, VoWiFi to Landline calls and again to make sure all combinations of codecs are exercised.

To illustrate the importance of testing these various combinations, we’ll share results from some mobile-to-mobile tests we performed on a live operational network enabled with EVS and AMR-WB codecs. Our tests focused on comparing EVS and AMR-WB speech quality in a variety of typical user locations. The chart on the left of Figure 2 shows that EVS to EVS calling was superior in terms of speech quality when compared to AMR-WB to AMR-WB calling. The chart on the right side of Figure 2 shows cross-codec test results for AMR-WB to EVS codec calling (and vice versa). During this particular test, we observed a problem with transcoding which led to degraded speech quality for cross-codec connectivity only. After this problem was identified, the carrier was able to isolate the root cause to an IMS firmware issue which was quickly fixed.

Page 3: 6 Best Practices for Launching VoWiFi, VoLTE and EVS

3

3.84.14.1

4.4

Mean Max

Speech quality (POLQA MOS)

AMR-WB EVS

3.0

Good speech quality

3.43.8

3.03.4

Mean Max

Speech quality (EVS to AMR)

AMR-WB party EVS party

Lower quality than AMR-AMR

Figure 2: Test results for EVS, AMR-WB and cross-codec EVS / AMR-WB calling in an operational network.

Troubleshooting Mobile-to-Mobile Issues (VoLTE & VoWiFi)

The next challenge we’ll look at is how to troubleshoot mobile-to-mobile voice service issues in a VoLTE network. Tests based on mobile to mobile calling are often the only way to “exercise” a new service or technology due to the fact that the service only works on a subset of devices or for certain network infrastructure. If the test results reveal a voice services problem exists it can be challenging to pinpoint and troubleshoot exactly what could be causing it:

• Is it a device issue?

• Is it an LTE Access network issue?

• Is it a core/IMS network issue?

• Is the problem is occurring in the uplink of one device or

the downlink of the other?

These challenges also apply to VoWiFi networks—is it a device issue, a network issue or perhaps a problem with the venue Wi-Fi network or its backhaul to the core network?

The nature of Wi-Fi means congestion, interference and latency are additional possible causes of the problem and often outside the direct control of the mobile operator.

Following is a practical example of a set of mobile-to-mobile calling issues which could have multiple root causes and are therefore extremely challenging to troubleshoot. We performed over 1,500 mobile-to-mobile calls in multiple locations served by operational Wi-Fi and 3G (UMTS) networks. We used two pairs of mobile phones, where one pair was making Wi-Fi to Wi-Fi calls and the other was making Wi-Fi to 3G calls. In Figure 3, which shows the test results, we can see Wi-Fi to Wi-Fi performance is superior compared to Wi-Fi to 3G in terms of call completion, call setup and speech quality. Because the Wi-Fi to 3G calls in this test campaign include an uplink and downlink on each technology, it is extremely challenging to isolate the root causes of the performance differences observed.

5.1

8.6

0

2

4

6

8

10

WiFi-WiFi WiFi-Mobile

Cal

l Set

up T

ime

(s)

Slower

Faster

Wi-Fi to Wi -Fi had faster call setup vs. Wi-Fi to Mobile

92%

87%

75%

80%

85%

90%

95%

WiFi-WiFi WiFi-Mobile

Cal

l Com

plet

ion

Succ

ess

Rat

e (%

)

More Reliable

Less Reliable

Wi-Fi to Wi -Fi was more reliable vs. Wi-Fi to Mobile

Call Completion Success = Successful Initiation + Successful Retention (No Drop)

Better Quality

Worse Quality

Wi-Fi to Wi -Fi had better speech quality vs. Wi-Fi to Mobile

3.3

2.5

0

1

2

3

4

WiFi-WiFi WiFi-Mobile

Spee

ch Q

ualit

y (P

OLQ

A M

OS)

Figure 3: Over 1,500 VoWiFi calls made in multiple locations in an operational network. Two scenarios: Wi-Fi to Wi-Fi and Wi-Fi to 3G Mobile.

Page 4: 6 Best Practices for Launching VoWiFi, VoLTE and EVS

www.spirent.com

6 Best Practices for Launching VoWiFi, VoLTE and EVS

Evaluating VoLTE-VoWiFi Handoff

The next challenge we’ll examine is evaluating the impact of handoff between VoLTE and VoWiFi on user experience. More and more mobile network operators are using VoWiFi as a way of easing the load on their LTE / 3G networks and expanding their coverage. An important aspect of this policy is to ensure seamless handover of calls from the VoLTE network to the VoWiFi network (and vice versa) without the user perceiving discontinuity in the call or degradation in audio quality.

As the call hands over from one access technology to the other the routing through the backhaul and core IMS network often changes and there are number of places where things can go wrong. Figure 4 and 5 depict measurements of speech quality for VoLTE to VoWiFi handoffs in an operational network.

For these measurements, we evaluated speech quality before, during and after a handoff. The test scenario included the following steps: establish a call on VoLTE, emulate a user walking into Wi-Fi coverage (by adjusting the attenuation of the Wi-Fi signal transmitted by the access point), waiting for the handoff to Wi-Fi, staying on Wi-Fi and collecting multiple speech samples and then emulating walking out of Wi-Fi coverage (again by varying Wi-Fi signal levels) until the call transitions back to VoLTE.

In the chart at the top of Figure 4, MOS values before the handoffs were very strong, averaging a score around 4. However, during the handoff we see scores that fluctuate between 1.5 and 2 with one example just below 3, reflecting a substantial degradation in user experience. The bottom graph in Figure 4 shows the WAV file for the speech during the handoff. During the handoff period from 5.7 seconds to 7.2 seconds, speech was almost non-existent. That’s 1.5 seconds where the user will not hear the other person talking.

0

1

2

3

4

5

1 2 3 4 5

Spee

ch Q

ualit

y (P

OLQ

A M

OS)

Hando� Iterations

DownlinkBefore HODownlinkDuring HO

Speech During HOSpeech Before HO

Figure 4: Speech quality (POLQA MOS) before and during handoff from VoLTE to VoWiFi.

Page 5: 6 Best Practices for Launching VoWiFi, VoLTE and EVS

5

0

0.5

1

1.5

2

1 2 3 4 5 6 7 8 9 10U

plin

kA

udio

Los

t (s)

0

0.5

1

1.5

2

1 2 3 4 5 6 7 8 9 10

Dow

nlin

kA

udio

Los

t (s)

1 2 3 4 5 6 7 8 9 10

1 2 3 4 5 6 7 8 9 10

Iterations Iterations

VoLTE to VoWiFi VoWiFi to VoLTE

Figure 5: Seconds of audio lost during VoLTE to VoWiFi and VoWiFi to VoLTE handoffs by uplink and downlink.

Figure 5 shows the number of seconds of lost audio across ten test iterations for uplink, downlink and various technology transitions. We can see that the VoLTE to VoWiFi transition has the most degraded user experience on both links with the highest average value on the uplink. Handoff performance issues such as these are typical of new technologies and therefore represent an important area of investigation prior to launch.

Assuring VoLTE/VoWiFi Interconnect

The final challenge we will explore is assuring the quality of voice services between mobile operators and MSOs as they start to interconnect their voice services with each other. Historically, wideband VoLTE and VoWiFi calls have only been possible between users within a single mobile operator. As soon as the call interconnected with another operator it would have to go through the PSTN and be transcoded down to the narrower bandwidth used by the PSTN. That meant it was impossible to have end-to-end wideband audio even though the two devices might have been capable of supporting such a call within their own network.

Now mobile operators and MSOs are beginning to interconnect and support end-to-end wideband calling between their services. With this new interconnection comes a host of new dimensions or scenarios that need to be tested in the field to assure quality at launch. Examples of new test dimensions include connections between providers or services and across markets. We now have to make sure that calls between operators work well, for example where one party connects via a mobile operator using VoLTE or VoWiFi and the other party connects on an MSO provider offering VoWiFi service only.

Add to these complexities the need to test different geographical markets and different telephony services and you have a rapidly growing set of scenarios that need to be assessed. An example of an issue that can be problematic in this type of interconnect environment is the challenge of mouth-to-ear audio delay in a VoLTE or VoWiFi call. The human ear is very sensitive to conversational audio delay so mobile networks go to great care to make sure the latency in their network is kept to a minimum. To control latency these packet-based networks use mechanisms such as IP Quality of Service (QoS) to make sure packets which are latency sensitive are prioritized vs. non-latency sensitive traffic. When calls are connected between networks the management of QoS becomes more challenging and the possibility of longer latency times is significantly increased. If that delay becomes too large it can significantly decrease the quality of service that subscribers experience.

Page 6: 6 Best Practices for Launching VoWiFi, VoLTE and EVS

www.spirent.com

6 Best Practices for Launching VoWiFi, VoLTE and EVS

Let’s take a look at measurements of audio delay between two handsets in an operational network. For these measurements, a speech sample is sent from one handset to the other and we measure how long it takes between the moment the speech is inserted in one handset to the time it is received by the second handset and decoded and converted to speech again. The tests were performed for both AMR-WB calls and EVS calls. In Figure 6, AMR-WB has a lower mean and maximum audio delay compared to EVS. In addition, we note that the delay for the EVS call averages above 230 ms with a maximum delay of 300 ms. As we can see from the ITU-T Rec. G.114 chart, delays above 200 ms will become increasingly annoying to the end user.

180

230220

300

Mean Max

Mouth-to-ear audio delay (ms)AMR-WB EVS

200

Increasingly noticeable and

annoying

EVS

max

mean

Figure 1/G.114 – Determination of the e�ects of absolute delay by the E -model ITU-T Rec. G.114

ITU-T Rec. G.114

Figure 6: The impact of mouth-to-ear delay on user satisfaction (ITU-T Rec. G. 114)

and mouth-to-ear audio delay measurements for AMR-WB and EVS in an operational mobile network.

Best Practices for Launching Next-Gen Voice ServicesMeasure What Matters

To address the challenges we’ve described in the previous sections, there is one overriding best practice: measure what matters to subscribers before launching new VoLTE, VoWiFi and EVS-enabled devices and services. That means measuring all the key factors that impact the service experience using unmodified consumer devices in the live network.

Specifically, we recommend assessing the following service experience factors:

1. Speech Quality—The POLQA algorithm is specifically designed to measure the perceived quality of HD voice services including EVS.

2. Call Success Rate—The ability of the device/service to successfully initiate calls.

3. Conversational Audio Delay (aka Mouth-to-Ear Audio Delay)—The time it takes for speech from one user to be heard by the other.

4. Call Setup Time—The time it takes to initiate a call and begin conversation.

5. Lost Audio During Handoff—Degradations in speech quality or periods of silence during handoffs between access technologies or within the same access technology.

All of the measurements we’ve shared in the previous sections are examples of real-world VoLTE, VoWiFi and EVS service issues: these are problems that would have had a significant impact on the customer’s experience of the service if they hadn’t been caught during pre-launch testing. Particularly with complex, immature next-gen voice technologies, we highly recommend testing voice experience before launch to make sure problems are found and fixed before they impact the customer. Following are specific best practices which derive from this key principle.

Page 7: 6 Best Practices for Launching VoWiFi, VoLTE and EVS

7

Practice 1: Test Device Performance in the Lab

The first best practice we recommend is to test the service experience of devices in the lab during both the R&D and device quality assurance phases. In the lab environment, we are able to emulate the network in a controlled way enabling us to perform repeatable tests that isolate device issues. Some examples of the control we can exercise over the emulated network include configuring the IMS core to represent carrier-specific implementations, using multiple IMS cores (with different configurations), introducing network impairments and controlling the access technology offered including LTE, UMTS and Wi-Fi (see Figure 7). There are two typical test modes in the lab: using two actual devices to make mobile-to-mobile calls and using an emulated device in lieu of one end-point. We recommend using both approaches to fully exercise new devices, including testing cross-device model / manufacturer interoperability.

The main advantages of lab testing are:

1. Tests can be performed quickly with minimal human resources, allowing key issues to be identified before moving on to testing in the field / live operational network.

2. Tests which aren’t possible in the live network can be performed due to the ability to control the emulated network or device end-point.

3. Test results are highly repeatable since all aspects of the end-to-end network are controlled, allowing accurate performance benchmarking vs. previous releases.

CORE / IMS

Isolate and Diagnose Device Issues

End-to-End Testing With 2 nd DUT

OR

Simulated Device (Configurable IMS Client and Audio Codec)

Configurable IMS Core

Configurable LTE, UMTS, and WiFiNetwork Emulation

VoLTE / VoWiFi /

EVS-enabled devices

Voice Probe Connected to Device-Under-Test

AMR-NBAMR-WB

EVS SPEECH QUALITY

CALL SUCCESS RATE

AUDIO DELAY

Configurable Network

Impairments

Network Emulation

Figure 7: End-to-end voice service testing in the lab enables rapid isolation of device issues.

The tradeoff of lab testing is that it doesn’t fully replicate the complexity of a live operational network environment including loading, RF interference and other factors. Therefore, we recommend both lab-based and live network testing as a best practice. The following sections describe best practices for live network evaluation of service experience.

Page 8: 6 Best Practices for Launching VoWiFi, VoLTE and EVS

www.spirent.com

6 Best Practices for Launching VoWiFi, VoLTE and EVS

Practice 2: Test Inter-Service, Inter-Codec Service Experience in Live Network

The first live network best practice we recommend is to test service and codec interactions in the operational network using real, unmodified consumer devices. Our experience has shown that some problems only manifest with specific devices, codecs or even geographies (due to different infrastructure providers and network provisioning). For that reason it’s important to exercise all of these permutations in real-world conditions as part of your testing. Figure 8 provides an example of various types of mobile-to-mobile testing that that should be performed for VoLTE and VoWiFi devices.

CORE / IMSPSTNBACKHAUL /

AGGREGATION

Voice Probe Connected to Unmodified Consumer Devices

Wi-Fi: Home, O�ce, Stadium, Metro area, etc.

LTE OR

VoLTE / VoWiFi /

EVS-enabled devices

Voice Probe Connected to Unmodified Consumer Devices

VoLTEEVS

3GAMR-WB

3G AMR-NB

SPEECH QUALITY

CALL SUCCESS RATE

AUDIO DELAY

SPEECH QUALITY

CALL SUCCESS RATE

AUDIO DELAY

X

SPEECH QUALITY

CALL SUCCESS RATE

AUDIO DELAY

X

Figure 8: Measuring service experience for various inter-service, inter-codec permutations

using a voice probe capable of interfacing with unmodified consumer devices.

It’s entirely possible that a device that has good speech quality may have poor audio delay or vice versa, and that those issues may only be present when exercising a specific codec. To make this type of testing practical it’s necessary to use service experience measurement technology that can automatically place hundreds of calls and measure perceived audio quality and audio delay using a repeatable and statistically valid approach. Furthermore, the service experience measurement capability needs to be able to interface with unmodified consumer devices regardless of technology or form factor.

Page 9: 6 Best Practices for Launching VoWiFi, VoLTE and EVS

9

Practice 3: Troubleshoot Issues Using a Test End Point (in IMS or PSTN)

Another practice we recommend is troubleshooting service issues by testing to end-points within the mobile network or PSTN. As previously described, troubleshooting the root cause of mobile-to-mobile tests can be extremely challenging. One of the key challenges is simply isolating problems to one link or the other (particularly with cross-technology calling). To address this challenge we recommend deploying a voice test end point within the mobile network IMS or connected directly to the PSTN. The voice test end point terminates calls to/from a mobile device and enables speech quality testing for the mobile to end point and end point to mobile paths. This enables isolation of problems to either the uplink or downlink path.

In Figure 9, we depict a voice test end point called an HD Voice server which can terminate end-to-end VoLTE/VoWiFi calls and source/record high definition audio to enable speech quality testing on the uplink and downlink.

CORE / IMSPSTNBACKHAUL /

AGGREGATION

HD Voice Server in IMS/Core Enables Link Isolation

Voice Probe Connected to Unmodified Consumer Devices

VoLTE, VoWiFi

VoLTE / VoWiFi /

EVS-enabled devices

SPEECH QUALITY

CALL SUCCESS RATE

AUDIO DELAY

SPEECH QUALITY

CALL SUCCESS RATE

AUDIO DELAY

X

Wi-Fi: Home, O�ce, Stadium, Metro area, etc.

LTE OR

Figure 9: Using an HD voice server test end point to isolate uplink and downlink issues with next-gen voice services.

In Figure 10, we indicate a similar approach for terminating calls via the PSTN. In this case the voice server end point connects to the PSTN via an E1/T1 connection. The PSTN voice server also allows isolation of uplink/downlink issues but is limited to supporting narrow band audio only due to the limitation of the transcoding at the boundary of the PSTN network.

CORE / IMSPSTNBACKHAUL /

AGGREGATION

µ-lawNB

PSTN Voice Server Evaluates Landline Calling Experience / Enables Link Isolation

Voice Probe Connected to Unmodified Consumer Devices

VoLTE / VoWiFi /

EVS-enabled devices

SPEECH QUALITY

CALL SUCCESS RATE

AUDIO DELAY

SPEECH QUALITY

CALL SUCCESS RATE

AUDIO DELAY

X

Figure 10: Using a PSTN voice server test end point to isolate uplink and

downlink issues for calling between next-gen voice services and the PSTN.

Page 10: 6 Best Practices for Launching VoWiFi, VoLTE and EVS

www.spirent.com

6 Best Practices for Launching VoWiFi, VoLTE and EVS

Practice 4: Test Service Experience Before / After / During Handoff

The next best practice we recommend is to test the service experience of VoLTE and VoWiFi before, during and after handovers. We recommend using a voice probe to automate measurement of speech quality, call performance and audio delay during each handover scenario. Furthermore, we recommend creating a controlled handoff between VoWiFi and VoLTE by applying attenuation to the antenna of the Wi-Fi Access Point under test. This provides a repeatable and controllable way to test how different devices perform this transition. Figure 11 shows the setup for handoff testing. Note: in this diagram Wi-Fi and LTE access networks are depicted going through the same backhaul / aggregation network for simplicity.

CORE / IMSPSTNBACKHAUL /

AGGREGATION

Voice Probes Evaluate Calls to Before / During /

After Hando�

Voice Probe Connected to Unmodified Consumer Devices

Wi-Fi

LTE

VoLTE / VoWiFi / EVS-enabled

devices (in hando�)

Voice Probe Connected to Unmodified Consumer Devices

VoLTE/ EVS

SPEECH QUALITY

CALL SUCCESS RATE

AUDIO DELAY

X

ONT During

Hando�

BeforeHando�

AfterHando�

DuringHando�

BeforeHando�

AfterHando�

SPEECH QUALITY

CALL SUCCESS RATE

AUDIO DELAY

SPEECH QUALITY

CALL SUCCESS RATE

AUDIO DELAY

Figure 11: Testing service experience before, during and after VoLTE to VoWiFi handoffs.

Page 11: 6 Best Practices for Launching VoWiFi, VoLTE and EVS

11

Practice 5: Test Inter-Provider / Inter-Service Experience

With the advent of VoLTE roaming, another practice we recommend is to test inter-provider and inter-service experience. As we mentioned before, this means testing the speech quality, call performance, and audio delay of mobile-to-mobile calls when each party is using a different service and potentially a different provider (see Figure 12). Sometimes the two devices in such calls will be collocated which makes testing relatively easy, but sometimes these devices will have to be physically remote from one another which makes testing much more challenging

One of the most complex tests we recommend is measurement of audio delay between providers in remote locations (i.e., across markets). Due to the physical distances and potential use of third party IP exchange (IPX) networks, this scenario brings the greatest likelihood of experience-impacting audio delays due to packet latency. To make inter-market audio delay measurements requires very accurate synchronization of clocks between the locations. We recommend use of a voice probe that makes use of satellite or other high resolution timing sources for synchronization so that highly accurate inter-market audio delay measurements may be performed.

BACKHAUL / AGGREGATION

CORE / IMS

CORE / IMS BACKHAUL / AGGREGATION

CORE / IMS

BACKHAUL / AGGREGATION

CORE / IMS

BACKHAUL / AGGREGATION

IPX / IPInterconnect

ONT

ONT

Mobile Operator

VoLTE

MSO VoWiFi

Mobile Operator

VoLTE

MSOVoWiFi

SPEECH QUALITY

CALL SUCCESS RATE

AUDIO DELAY

SPEECH QUALITY

CALL SUCCESS RATE

AUDIO DELAY

SPEECH QUALITY

CALL SUCCESS RATE

AUDIO DELAY

X

SPEECH QUALITY

CALL SUCCESS RATE

AUDIO DELAY

X

Figure 12: Testing service experience for various types of inter-provider and inter-service

connectivity including VoLTE provider to VoLTE provider, VoLTE provider to VoWiFi provider and more.

Based on our experience, it is possible that services may work great between some partners but certain aspects of a service, such as audio delay, might be problematic in specific scenarios. By evaluating the service experience prior to launch, any issues can be addressed before they impact customers.

Page 12: 6 Best Practices for Launching VoWiFi, VoLTE and EVS

www.spirent.com

6 Best Practices for Launching VoWiFi, VoLTE and EVS

Practice 6: Be Efficient!

The final practice we recommend is to be efficient. The technology behind next generation voice services and the number of permutations of service and codec interactions can be daunting. To be able to cost-effectively evaluate the large number of scenarios required, it is critical to make effective use of testing resources—both people and equipment.

Teams involved in the testing activities tend to be fairly large and geographically disperse. We recommend maintaining a central, cloud-based repository of test plans and test results to ensure that consistent processes are followed across the team (see Figure 13). In our experience, this approach leads to faster, more accurate measurement campaigns with the fewest possible resources.

3. Test multiple devices at once (up to 6)

2. Measure speech quality and call performance at the same time

1. Centralized config, upload & reporting

Figure 13: Best practices for running efficient service experience measurement campaigns.

Secondly, evaluating speech quality and call performance are time consuming activities. Testing them at the same time allows test time to be compressed as compared to testing them individually. Systems which allow call automation via Bluetooth whilst simultaneously measuring voice quality via the device headset connection enable simultaneous testing, saving substantial time and resources.

Finally, we’d like to recommend testing of different scenarios in parallel to save time. Test equipment should allow up to six devices to perform automatized testing in parallel. With this capability it’s possible to test intra network mobile-to-mobile calls (within a single network), inter network mobile-to mobile calls between networks and inter-service scenarios all at the same time.

Page 13: 6 Best Practices for Launching VoWiFi, VoLTE and EVS

About Spirent Communications

Spirent Communications (LSE: SPT) is a global leader with deep expertise and decades of experience in testing, assurance, analytics and security, serving developers, service providers, and enterprise networks.

We help bring clarity to increasingly complex technological and business challenges.

Spirent’s customers have made a promise to their customers to deliver superior performance. Spirent assures that those promises are fulfilled.

For more information, visit: www.spirent.com

Contact Us

For more information, call your Spirent sales representative or visit us on the web at www.spirent.com/ContactSpirent.

www.spirent.com

© 2018 Spirent Communications, Inc. All of the company names and/or brand names and/or product names and/or logos referred to in this document, in particular the name “Spirent” and its logo device, are either registered trademarks or trademarks pending registration in accordance with relevant national laws. All rights reserved. Specifications subject to change without notice.

Americas 1-800-SPIRENT +1-800-774-7368 | [email protected]

US Government & Defense [email protected] | spirentfederal.com

Europe and the Middle East +44 (0) 1293 767979 | [email protected]

Asia and the Pacific +86-10-8518-2539 | [email protected]

Rev B | 08/18

6 Best Practices for Launching VoWiFi, VoLTE and EVS

ConclusionIn this white paper, we’ve presented some of the challenges involved in deploying next generation voice services and recommended best practices to help tackle those challenges. The EVS codec, VoLTE and VoWiFi represent a big step forward for ubiquity and fidelity of voice services but testing of these services will be critical to assure a great user experience at launch. In particular, it is critical to test the interaction of new and legacy services and codecs and their interconnection with other services and providers.

We recommend a comprehensive service experience test campaign in both the lab and live network prior to the launch of any new voice service and supporting devices. This campaign should measure what matters—the end-to-end user experience—and should evaluate experience for all customer use cases.

If you’re interested to learn more about Spirent’s solutions which enable these best practices, please see the Umetrix, Nomad UX and Elevate solutions on our website at www.spirent.com.