The cost savings and reduced complexity from enterprises moving to an all-wireless communications network is a seductive one. However, worries still exist among many enterprise IT managers that Wi-Fi is not up to snuff. Indeed, there are still concerns about scalability, quality, and security issues.
A recent TechZine article by Subramania Vasudevan, Director, Advanced Performance in WCTO, Alcatel-Lucent, All-wireless enterprise with LTE and Wi-Fi, notes that enterprise IT managers have a particular lack of confidence in the quality of the wireless link provided by an all Wi-Fi infrastructure.
“There’s the limited ability of the Wi-Fi network to scale with increasing data rate needs,” Vasudevan noted. “In fact, we’ve seen aggregate capacities barely increase — even as Wi-Fi networks densify.”
LTE small cells can help. Small cells help provide in-building LTE on a cost-effective, as-needed basis.
Many mobile operators are considering unlicensed spectrum to bring greater bandwidth into the enterprise, he added. This can help meet the scalability demands. In fact, operators are looking to aggregate LTE in licensed bands along with LTE in the 5GHz unlicensed bands, which are known together as Licensed Assisted Access (LAA) or LTE unlicensed (LTE-U).
The limitations of Wi-Fi often come from the sharing mechanism between the uplink and the downlink. By using an LTE-based system, enterprises can resolve the problem of contentious uplink by means of scheduled access. This frees up the enterprise’s existing Wi-Fi for downlink, according to Vasudevan.
“By offloading the Wi-Fi uplink to cellular, LTE small cells improve enterprise services,” he wrote. “In addition, in-building enterprise traffic, such as Lync application data, can be shunted across the enterprise LAN (i.e., local breakout is enabled).”
At the same time, the combination relies on pre-existing Wi-Fi APs and user equipment, so the sum total is that the LTE downlink capacity can be aggregated with the Wi-Fi APs downlink capacity. This can lead to users everywhere seeing higher throughput in more locations because they benefit from LTE Wi-Fi aggregation and LTE-only for uplink.
The all-wireless enterprise network might be closer than many enterprise IT managers realize. This is a good thing since so many of us use our smartphones as our primary communications device and a significant number of interactions on those devices originate or terminate in-building where coverage and quality of service are a challenge.
It is to keep enterprise customers on the mobile service provider networks for enhanced services that good in-building wireless solutions are seen as both a powerful business tool and a competitive advantage. This is particularly true when it comes to retaining small-to-medium business customers (SMBs).
This is an observation, driven home well in a recent TechZine posting by Tristan Barraud de Lagerie, Product Marketing Manager, Small Cells, Alcatel-Lucent (ALU), Field insights: Small cells retain enterprise customers. As he points out, ALU research has shown that more than 87 percent of enterprises are likely to switch to operators that guarantee good performance, but Alcatel-Lucent research found. He adds that, “Until recently, very few wireless solutions have been dedicated to meeting small enterprise needs. Not even in France and the U.K., where SMEs make up 99.8 percent of all businesses — and employ more than half of the workforce (51 percent in France, 59.3 percent in the UK.”
Source: Alcatel-Lucent
Small cells to the rescue—a five step approach for success
The good news for mobile operators is that small cells can give them the in-building coverage and quality of service they need to prevent churn and provide quality customer experiences. In fact, Barraud de Lagerie outlines five step approach to achieve success which have been applied successfully at thousands of small enterprises in France and the U.K. The five steps include:
The goal is simple, small cell deployments can be and should be positioned as a win/win for the operator and for enterprise customers. This is especially important for attracting and keeping SMBs customers as in-building gives the operators the ability to quickly and cost-effectively introduce new services like high-quality voice over LTE (VoLTE). And next-generation multi-standard (3G/4G/Wi-Fi) small cells make it easier to migrate to new services, and support always-on and all ways connections for the exploding and diverse population of wireless devices and their increased use as the communications platform of preference for all enterprise communications.
]]>Data and signaling growth are usually good news for network operators, since growth often translates into higher revenues. But when growth is averaged over a month or quarter, the daily highs and lows of network activity are smoothed out. And signaling spikes remain hidden within the averages. These spikes can overwhelm available signaling capacity, which impairs the customer experience, as well as the operator’s reputation.
What happens when a spike occurs? Typically, a CPU Overload alarm appears on various mobile nodes. And the Network Operations Center (NOC) immediately starts praying that the burst is short-lived and doesn’t go over maximum peak-rate capacity. Because when that happens, all consumers are denied service access. Then, the process of identifying the source of the problem begins. This can be arduous, because it often involves applications completely out of NOC control. And the issue can’t be resolved easily without solid network analytics that enables engagement with application and device developers.
That’s the reason signaling information is a crucial part of the Alcatel-Lucent Mobile Apps Rankings report and why LTE World 2014 devotes an entire pre-conference day to the topic. It’s also why this blog offers a closer look at how some real-world disruptive signaling spikes got started — and were finally resolved.
Signaling spikes: The basics
There are three kinds of signaling spikes:
The following signaling spikes were observed with the Alcatel-Lucent Wireless Network Guardian (WNG) on multi-vendor networks. These examples demonstrate the impact and resolution of signaling jumps ranging from 36% to 92%.
Microbursts 1: Samsung, Google and a pre-loaded app
As shown in Figure 1, all the serving gateways (SGW) in this example experienced short signaling spikes, six times per day (at 00:00, 1:30 a.m., 6:30 a.m., 8:00 a.m., 12:30 p.m. and 6:30 p.m.). The spikes started with a barely noticeable 8% jump in signaling and grew steadily to the 44% jump that’s clearly visible one year later.At last, one spike became too much for the SGW. Some of its blades were brought down by the overload, which caused a signaling disruption and partial service outage. Traffic was diverted to a higher-capacity backup SGW, until the issue could be isolated and resolved.
Using the Alcatel-Lucent WNG for analysis, the problem was isolated to Samsung S4 devices with Android versions 4.2 and 4.3 — when traffic originated from these devices and tried to reach Google.com. Once Samsung had that information, they were able to determine that a pre-loaded app on that device generated the spikes. This app connected with a Google API to determine the user’s location, so local news could be delivered to the consumer.
In Android version 4.4, the app had already been removed. For versions 4.2 and 4.3, the operator initially believed they could simply remove the offending app as a way to address the signaling spikes. However, pre-loaded apps are difficult to remove. So instead, multiple updates have been pushed to test devices, in attempts to iteratively eliminate the problem.
Extended bursts: Viber outage
One day in April, CPU overload alarms reported that Radio Network Controllers (RNCs) were flooded with requests. The signaling spike pattern shown in Figure 3 was matched to one app: Viber. Further investigation into Viber’s flows showed that Viber servers were no longer responding. But here’s what wasn’t clear: Why would this app outage generate such a load on signaling resources? The answer lies in how Viber handled the call failures. The app loaded on mobile devices tried repeatedly to connect with the Viber server, and that created a growing wave of signaling as more and more users kept trying to connect.
The impact of this outage varied across networks. Where only a few consumers used the Viber app, operators might not have noticed. Where there was a high proportion of Viber users, they experienced a spike. Each network’s ability to tolerate the spike depended on whether it had enough peak hour signaling capacity. The timing of the Viber outage also produced different results in different geographical locations. That is, an outage during the network’s peak time — and the heaviest Viber usage — had a heavier impact.
Microbursts 2: Microsoft Exchange and iOS
This case illustrates another short-term outage where a signaling spike exceeded signaling capacity on a daily basis. As shown in Figure 4, a 36% signaling jump occurred every day at midnight, but the reason for the spike remained mysterious. The Alcatel-Lucent WNG narrowed down the issue: The signaling was initiated when devices tried to reach the Microsoft Exchange server. Lasting less than 1 second, this exchange only involved iPhone devices, and it occurred most often with iOS version 6.1. After obtaining this information, network operators were able to contact Apple and identify the root cause. Then a fix was implemented in a later iOS version update.
These real-world examples of signaling spikes clearly demonstrate that signaling design is an important aspect of the customer experience. More specifically, they bring home three important points: First, a robust and well-dimensioned signaling plane that can absorb sudden spikes is essential. In addition, our device and app ecosystem needs to consider how network signaling interactions can be optimized, when designing products. And finally, a strong network analytics solution is also a necessity for tracking the signaling of each application, detecting signaling anomalies and identifying root causes as quickly as possible.
The Alcatel-Lucent Analytics Beat studies examine a representative cross-section of mobile data customers using the Alcatel-Lucent Wireless Network Guardian, and they are made possible by the voluntary participation of our customers. Collectively, these customers provide mobile service to millions of subscribers worldwide.
Follow Josee on Twitter @joseeloudiadis
What do consumers know about the effects of signaling? We pay attention to our data plan costs — such as how many bytes of data we use for Skype calls, sending photos or watching YouTube videos. But signaling remains mysterious, because we don’t pay for it in any obvious way. And because it operates transparently, we are not aware of its impact.
The Alcatel-Lucent Mobile App Rankings report wants to change that, by including signaling information in the “application cost” section of this report, right along with “data volume costs.” This data can help consumers and mobile app developers gain a better understanding of these nearly invisible events that impact batteries all day long – even when not actively using the mobile phone.
Signaling depletes batteries
The bottom line is this: Mobile applications that signal more often will deplete your device batteries faster. Of course, there are other factors affecting battery life. Among the worst drains are very bright phone screens, Wi-Fi constantly searching for a signal and apps checking regularly for new software versions. But the signaling that an application generates is another significant drain, sapping your battery’s power. Our Mobile Apps Rankings study found that an average user’s mobile device signaled 250 times every day – every 6 minutes on average!
But it’s important to recognize that applications vary widely in the amount of signaling they generate. Figure 1 makes this very clear. Apps appearing near the top of the graph generate the most signaling. And it makes sense that they’re primarily social media apps, since these app users often send short text messages and receive lots of notifications telling them that someone wants to connect or someone has retweeted a message.
Figure 2 examines daily signaling events more closely to discover how many events (on average) can be attributed to a given app installed on a mobile device. Yahoo! Messenger, for example, generates 76% of an average user’s daily signaling events. (That’s 190 out of 250 events every day.)
To help reduce battery drain, these kinds of high-signaling applications would obviously benefit from signaling optimization. But to understand what steps can be taken, we first need to take some of the mystery out of signaling.
What is signaling?
Signals are the messages sent between your device and the network to set up whatever is needed for data to start flowing. The signaling activity falls within four main categories of events:
Ways to reduce signaling
How can battery drain be reduced? Strategies usually revolve around the combination of two approaches: They focus on reducing the number of notifications that wake up the phone. Or they reduce the number of times a phone must grab or release a channel during active use.
Strategy 1: Reducing notifications. Notifications can’t be turned off completely, because they are a necessary part of messaging. But app developers can promote reductions by offering configuration options for different types of notifications. Then consumers can play an important role by looking for apps that provide this flexibility — and carefully choosing just how much notification they want to receive. Here are some examples of these options and how they work:
Social media apps. These apps could improve their flexibility by offering different options for how frequently notifications are sent. For example, retweet notifications could be sent immediately, while new followers could be grouped into one notification per day or week.
Strategy 2: Reducing channel allocation “grab & release” cycle. This type of optimization is in the hands of app developers. But here are a few key points that everyone should know about:
The Alcatel-Lucent Mobile App Rankings report helps shed some light on the impacts of signaling, which are not widely understood, so far. By learning more about signaling, notification and channel allocation, consumers can gain more control over their battery consumption. And developers can design better options for reducing the load on networks and batteries. The collaboration between these three stakeholders ultimately enhances consumers’ experience with their mobile devices.
Alcatel-Lucent Analytics Beat studies examine a representative cross-section of mobile data customers using the Alcatel-Lucent Wireless Network Guardian, and they are made possible by the voluntary participation of our customers. Collectively, these customers provide mobile service to millions of subscribers worldwide.
Application signaling traffic
At this year’s Mobile World Congress, Facebook founder Mark Zuckerberg urged mobile operators to zero-rate Facebook, as part of an effort to bring the Internet to 2 billion new users, while simultaneously increasing operator profits. But, according to Analytics Beat, Facebook increased overall signaling load by 5-10% overnight, when new software was released. With this kind of lightning-fast impact, it’s easy to see why operators are reluctant to eliminate all charges for an app that could quickly become too costly to support.
And high signaling impacts consumers too, by increasing the battery drain on their devices. That means Facebook had another important reasons to resolve their enormous spike in signaling traffic — since revenues from mobile use are increasingly important to its growth. As a result, only 5 months later, a new, improved Android version was released. This version did more than “undo” the signaling increase — it actually improved signaling performance.
But Facebook had to learn about service providers’ predicament before choosing to optimize. And the optimization process also needed data that could pinpoint the cause of the signaling increase. Clearly, visibility into an app’s impact on mobile networks is key to reducing its cost of delivery — and to understanding how the app could be packaged in ways that benefit all parties.
Battery and bandwidth consumption
Another blind spot is the lack of comparative information available to consumers. For example, WhatsApp announced at Mobile World Congress 2014 that it would be introducing a voice capability in the second quarter of 2014, which will “use the least amount of bandwidth.” But how many consumers know the precise amount of bandwidth and battery their current version of WhatApp consumes for texting? Very few. And even fewer know how that consumption compares with competing apps like Yahoo! Messenger or Kik. So how will anyone know whether WhatsApp uses the least data — and whether it will be less costly than Skype, Tango or their current voice plan?
The Alcatel-Lucent Mobile App Rankings report provides some answers for the data consumption of the current data functionality:
If you’re wondering what adding voice support to WhatsApp will do to your battery and data plan, check back with us. We’ll be watching.
As you can see, this impact report includes information that can spur collaboration between operators and app developers and support app optimization and packaging. Service providers can find out which apps impact their networks the most today — and which are on the Watch List (which shows the potential for network impact, if the app’s popularity grows). And consumers can benefit too, by learning about the average cost of popular apps on their data plans and the effects on battery life.
Alcatel-Lucent Analytics Beat studies examine a representative cross-section of mobile data customers using the Alcatel-Lucent Wireless Network Guardian and are made possible by the voluntary participation of our customers. Collectively, these customers provide mobile service to millions of subscribers worldwide.
]]>
“Multi-device Shared Plans are starting to gain traction in the market as they now represent 8.3% of tariff plans, growing 38% QoQ.” |
It’s not surprising to find a high level of interest in shared data plans among US consumers. Major US operators are leading the push for shared data plans, and Verizon Wireless launched its Share Everything plan nearly 2 years ago.
But shared data doesn’t just appeal to US mobile consumers. A sizable segment of the market in all of the surveyed countries is interested in sharing data. More importantly for mobile operators, this segment is willing to act to get its data plan of choice.
For example, respondents who express an interest in a shared data plan are highly likely to purchase one (Figure 1). In Brazil, 97% of these respondents say they are likely or very likely to buy a shared data plan. Likelihood to purchase is also high among interested respondents in the US (85%), UK (80%) and France (80%).
Respondents in Japan are the least inclined take the shared data plunge. Even so, 56% of those interested in sharing data say they would likely or very likely purchase a shared data plan. The very high probability of purchase in Brazil could be influenced by the fact that the survey reached a larger-than-average percentage of households with either 3 or more members (69%) or 3 or more devices (78%).
Those interested in sharing data are also willing to switch mobile operators to get this capability. Respondents in Brazil show the strongest willingness to change operators, with 97% of those interested in shared data indicating that they would definitely or likely make a switch to get a shared data plan. Interested respondents in the UK are also very open to change: 73% would change operators to get shared data. Figure 2 breaks down the churn impact of shared data plans across the 5 surveyed countries.
Shared data plans can play a strong role in subscriber retention, too. Results from all countries indicate that those interested in shared data are willing to stay with their current operator to keep it. The retention impact is highest in the US and Brazil, where 91% of those interested in shared data say that it plays a strong or very strong role in keeping them on board. Japanese consumers are less likely to stay with their current operators to keep shared data, but its retention impact (65% strong or very strong) is still formidable. Figure 3 summarizes the retention impact of shared data plans across the 5 surveyed countries.
What do all these interest, purchase likelihood, churn and retention numbers tell us? That this segment of the market knows what it wants and is willing to take action to get it!
And what’s behind the numbers? Respondents cite anticipated cost savings and efficient use of data across mobile devices as their major reasons for favoring share plans. Many also see shared data plans as a means to minimize “slippage” – data that is paid for each month but that goes unused. Those not interested in shared data plans primarily want to keep an existing unlimited data plan or have few mobile devices or users with which to share the data. They also anticipate that shared data plans will come with a cost increase.
Respondents interested in data sharing generally have a higher monthly mobile spend and a larger data allowance. Those who have purchased a Wi-Fi tethering plan from their mobile operator are also favorably inclined toward shared data plans. Finally, the number of people and mobile devices in the household influences interest in and willingness to adopt shared data plans. A minimum of 3 people or devices is the threshold at which consumers become interested in sharing data.
The Alcatel-Lucent research confirms that there is broad consumer interest in shared data plans. It also confirms that those interested in shared data are likely to take action to get it. Shared data is a tool that every mobile network can use to attract and retain customers.
Connect with the author, Rich Crowe, on Twitter: @rhcrowe.
Download the Six Degrees of Mobile Data Plan Innovation e-book. Please follow us on TMCnet, and click here to subscribe for more information and updates.
Past blogs in this series:
[1] Subscriber base size from “Service Provider Capex, Opex, Revenue, and Subscribers Database Quarterly Worldwide and Regional Database.” Infonetics Research, updated April 7, 2014.
The first step in resolving any problem is to make sure you understand the core issues. So here’s the crucial question for shadow IT: What is the biggest challenge it presents for your IT department?
Holding back the flood?
Today’s flood of mobile devices and cloud services is making shadow IT a bigger headache than ever before. But it’s nothing new. It started with the first enterprise employee who ever put an application in place without the knowledge or approval of IT staff.
Five years ago, the fundamental reasons for bypassing the IT department were the same as they are today: Individuals and departments want the best tools available for doing their jobs efficiently, so they can meet their key performance indicators (KPIs). This is especially true for today’s highly mobile workforce where the boundaries of the where and when you work are broad. A recent New York Times article noted that more than 3.2 million enterprise employees in America telecommute. This highly mobile and adaptable workforce wants to avoid cumbersome IT procedures that limit their choices, add unnecessary effort — and force them to wait around for delivery of their new applications. In other words, if something slows down productivity, these departments will go around it. This impulse is actually good for the business, because it helps employees wring every last drop of efficiency out of their processes.
But you know the consequences. When anyone circumvents standard procedures, their applications fall outside the controls and safeguards that your IT department has put in place. So is your biggest challenge to hold back today’s flood of nonstandard devices and applications, as a way to maintain control? Do you simply need to implement clearer rules, with tougher enforcement? Or is there another possibility?
Facing disruptive change
Some analysts believe that today’s pressure to be productive is driving growth in shadow IT. But a very different market trend may play a more powerful role: It’s now remarkably fast, easy and economical to acquire applications. Anyone with a credit card and a browser can have a new application running in virtually no time — instead of going through the lengthy process of IT procurement.
This application-on-demand method began in the consumer marketplace. But, like the bring-your-own-device (BYOD) trend, it has moved into the workplace. Even key decision makers in non-IT departments are using their budgets to quickly purchase what they need. As early as January 2012, a Gartner analyst predicted that, within 5 years, CMOs would spend more on IT than CIOs.
Now, as a result of this growing trend, 80 percent of respondents to a recent Frost & Sullivan survey said they use nonapproved software as a service (SaaS) applications in their jobs. In other words, 4 out of 5 employees are using shadow IT. The applications they rely on make up more than one third of the 20 SaaS applications used at the average company. And they extend across all application types. http://www.zdnet.com/6-reasons-why-shadow-it-is-emerging-from-the-shadows-7000024854/
Here’s an especially interesting point about this trend: The same study found that IT employees use a higher number of nonapproved applications than other employees do. The analyst who reported on these findings suggests that the high occurrence of nonapproved SaaS applications may indicate that they’re “no longer in the shadows.”
So perhaps the biggest challenge facing your IT department is to recognize the consequences of disruptive change. Shadow IT is the direct and simple reaction of employees who feel their needs are not being met through traditional IT channels. And today, they have easy, nearly instantaneous alternatives for getting what they want. These user-friendly options are inundating the marketplace. What will the consequences be if you focus on trying to control and contain them?
The next step
The second step in killing shadow IT is to embrace it and render it powerless! When IT organizations embrace shadow IT and implement strategies to enable employees to safely and easily access the innovative new applications they need everyone wins. That’s what we’ll discuss in the next blog in this series.
Bryan Davies leads Enterprise Communications Marketing for Alcatel-Lucent. To continue the discussion, follow Bryan on Twitter @brdavies.