Mobile network operators need to take steps in order to prepare their networks and business operations for the oncoming challenges of 5G and IoT.
Editor’s Note: In an attempt to broaden our interaction with our readers we have created this Reader Forum for those with something meaningful to say to the wireless industry. We want to keep this as open as possible, but we maintain some editorial control to keep it free of commercials or attacks. Please send along submissions for this section to our editors at: [email protected].
We know the mobile devices we carry are becoming smarter. High and ultra high-definition video are being woven into the fabric of our lives. Users demand mobile carriers deliver faster mobile broadband networks and better experiences. There is a gigantic market opportunity to connect the expected 30 billion devices to the internet of things by 2020, but only if we can design a cost-effective and significantly lower latency network than the 100 milliseconds tolerated by voice and mobile data today.
Emerging applications driving this need include real-time remote control, autonomous vehicles and live video from large sporting events. All of this is good news for mobile operators. Not so good news – subscriber acquisition is no longer a significant growth driver for the industry in developed markets, and many developing markets will reach a similar point in the next few years.
According to a recent GSMA report, the IP messaging application user base is growing exponentially with a global user base approaching 3 billion. Operators globally are increasingly investing in the development of exclusive mobile-only or at least mobile-first content to drive data traffic and subscriber loyalty.
Mobile network operators can’t just build more bandwidth in their networks proportional to the traffic because the average revenue per unit is fixed and the number of users is not growing at the rate it used to. The net promoter score is always hanging like a sword on their heads so MNOs can’t afford to compromise on quality of service because subscriber loyalty is nonexistent and they can’t take the current ones for granted.
There are two types of applications mobile networks have to contend with. One is service provider-managed apps like text messaging and voice over LTE, and the other is over-the-top content from apps like Skype, WhatsApp and Netflix. According to GSMA, revenue from mobile services worldwide is forecast to grow by around 2% annually through 2020. Organic revenue growth is slowing in line with slowing subscriber growth, but is being offset by new revenue opportunities resulting from rising mobile internet adoption and the move to higher-speed networks. It is projected that annual revenue from voice, text messaging and mobile data combined will surpass $1.5 trillion by 2020. However, as a proportion of overall mobile ecosystem revenue, the contribution from these services will fall from 41% today to 38% by 2025. By contrast, revenue from content services – services such as Netflix and Spotify – will increase from three percent of ecosystem revenue today to 17% by 2025.
To deal with the data tsunami challenge, MNOs need cost-effective visibility so that they can separate OTT apps versus service provider managed apps, and apply granular and context-based policies to each of these application categories.
Today’s 4G networks have inline bump in the wire bandwidth management solutions that are unaffordable in terms of cost and latency. Deep packet inspection technology enables mobile operators to examine the apps transmitted over the network. The DPI tools typically work on IP and the service assurance tools typically work in terms of Third Generation Partnership Project standards with unique subscriber number, or the international mobile subscriber identity and the IP-based protocols such as the GPRS tunneling protocols. Correlating between these two worlds is what increases the cost and latency.
MNOs are spending in excess of 50 cents per subscriber for the current inline DPI and service assurance solutions. This strategy will not scale as the number of subscribers (human and sensors) increases. For example, if a mobile operator wanted to gain visibility and insight on 5 million users connected to their network, this can cost in excess of $2.5 million per year. In addition, today’s architecture introduces latency in excess of 100 milliseconds, which is not an option for the internet of things or connected devices such as a driverless car.
In order to meet the demands of tomorrow’s network, mobile operators need three capabilities to achieve better visibility.
1. The solution needs to be offline, or out of band, to address the cost and latency challenges for the future mobile network. This can be done in the form of a new application programming interface inspired by MapReduce, a well-known big data paradigm from Google. MapReduce is a framework for processing large datasets using a large number of computers (nodes), collectively referred to as a cluster or a grid.
The MapReduce concept can be applied to a mobile network with an API that enables the analytics tools to perform the “Map” function. The API then enables the “Reduce” function, which sorts through and filters out unnecessary data such as duplicate packets. The overall system orchestrates the processing by marshaling the distributed servers, running the various tasks in parallel, managing all communications and data transfers between the various parts of the system, and providing for redundancy and fault tolerance.
2. In order for this to work, application usage needs to be monitored on mobile service provider, fixed broadband or enterprise networks. Real-time notifications should be leveraged to block, throttle or prioritize network traffic by application and subscriber. On today’s network there are many different types of applications that should be monitored, including Skype, WhatsApp, Viber, YouTube, Netflix, Hulu, Gmail and Yahoo Mail.
3. It is important for MNOs to have visibility into the correlation of user, transport and control plane traffic, enabling predictive analysis for detecting issues before they impact customers.
Implementing the above architecture will result in reducing the data set to a manageable size and can significantly decrease the cost per subscriber, while providing full visibility into customer behavior and experience. This will also significantly reduce latency from hundreds of milliseconds to just a few milliseconds making the solution suitable for “5G” and IoT applications.
Thus mobile network operators need a solution that does not have to grow linearly with subscriber and application scale, and makes the network visibility solution independent of the traffic growth or technology upgrades so MNOs can focus on more important issues of driving business and profits.
Sanjay Munshi is a senior director at Brocade for the Network Visibility and Analytics business unit. He is responsible for defining the vision, strategy and execution plans that drive the company’s business results for big data analytics software solutions in the data centers of mobile and cloud operators. His product development experience includes leadership roles at Extreme Networks, Nortel and Bay Networks.
