 | Issue: | 2010 |
| Article no.: | 7 | |
| Topic: | The mobile video traffic explosion | |
| Author: | Avichai Levy | |
| Title: | Executive Vice President of Marketing | |
| Organisation: | Mobixell | |
| PDF size: | 353KB |
About author
Avichai Levy is the Executive Vice President of Marketing at Mobixell. Previously, Avichai Levy was CEO of Targetize, a start-up providing mobile video and audio search and delivery solutions. Prior to that, Mr Levy was Managing Partner of StarVision, a strategy and marketing consulting company. Mr Levy was also a Senior VP Marketing and Business Development and one of the founders of TTI Telecom, a NASDAQ traded company that provides overall management solutions for telecommunication networks. Avichai Levy earned a BSc in math and computer science and an MBA from Tel Aviv University.
Article abstract
Video and Internet service usage is growing tremendously and straining mobile networks. It is vital that the methods and systems operators employ to cope with these challenges are capable of making intelligent, fundamental decisions, including sophisticated handling of video. Solutions are available in the form of intelligent video optimization platforms that can make smart decisions based on a wide range of network, service and user information to reduce the costs of video delivery while maintaining and often improving video viewing quality.
Full Article
In today’s ever-evolving mobile industry, operators find themselves in a paradoxical situation. Smartphone ownership and mobile broadband to the laptop is booming at the same time that the growth of the average revenue per user (ARPU) is flat or even moderately declining. On the one hand, the overwhelming growth of mobile Internet usage is an important new revenue stream. Some say mobile Internet is the next ‘killer-app’ after voice and short message service (SMS) that we’ve all be waiting for – this is the good news. On the other hand, the strain on the network and potential cost to build up those networks to meet the enormous data demands is a scary proposition – this is the bad news. Laptops, netbooks and smartphones generate the vast majority of data traffic. These high-end devices account for more than 90 per cent of overall mobile broadband traffic. The difficult issue for carriers to cope with here is the 108 per cent compound annual growth rate of mobile data traffic expected all the way through to 2014 (Cisco Visual Networking Index 2009-2014). With the exponential growth in video content due to the growing popularity of video-sharing sites such as YouTube (which in some networks is responsible for 50 per cent of all video traffic), social networks, and streaming content from leading Internet content providers, the portion of video traffic in overall mobile broadband traffic has grown significantly to more than 50 per cent and is expected to reach 66 per cent of overall wireless data traffic by 2014 (Cisco Visual Networking Index 2009-2014). So how will wireless carriers create revenues from the mobile broadband and mobile Internet traffic in order to justify the growing investment? Some assume that operators will compete with companies such as Apple and Google and provide content, Internet and application services. Others think that operators may create revenues from providing information to third parties, such as location information or user profile information. Today’s highly competitive market is driving wireless operators to offer flat rate or multi-tier data plans, which stimulate user consumption, but reduce the total revenue per bit. Is the ‘all you can eat’ model going away? Many operators are now offering differentiated plans, pricing and packages for consumers based on data usage. Some operators are trying to create different classes of service that charge differently for each application for video, social media usage and more; this helps operators efficiently segment their market and provides each user with a data package optimised according to the user’s needs. As wireless networks become more and more burdened by the increasing demand for data applications such as video, negative effects such as higher latency, buffering and denial of service are growing more common. Such phenomena can have a significant impact on the user’s experience. This is particularly true for video applications, which are especially sensitive to real-time variations in wireless network speed. Moreover, due to the bandwidth-hungry nature of video applications, a few wireless network users downloading video applications may have a greater negative impact on bandwidth capacity than tens or hundreds of web browsing users. Coping with the increasing demand for bandwidth-hungry data applications over their wireless networks is one of the most challenging issues that operators must address at the moment, as they aim to find the right balance among coverage, quality of service and profitability. Upgrading wireless networks to new 4G-access technologies such as LTE could probably alleviate the problem, but this is a costly and lengthy process. Besides, increasing capacity by itself does not guarantee a proper long-term solution, as data applications will also increase in terms of complexity and bandwidth consumption, and the user demand for these applications will also increase. AT&T mobility CEO Ralph de la Vega (CTIA Wireless opening keynotes, March 2010): “Spectrum and new 4G technologies will only go so far, mobile apps need to be optimized for the mobile network…Something as simple as sending optimized video for a phone’s media player before it hits the radio access network or developing more efficient mobile browsing technologies which resize content for a small-screen form factor can cut down on massive volumes of traffic currently flooding U.S. wireless networks…As the Internet and mobile worlds converge, the distinction between the types of applications that run on either is beginning to blur as well. Applications like video that required no optimization for the wide-open pipes of the wireline Internet are taxing the far more limited resources of the mobile broadband network.” A costly proposition Let’s not ignore the costs involved, either. Data traffic is growing faster than data revenue, while the cost to deliver these services remains flat. It is plain to see that operators face declining marginal revenue per bit. Put more simply, operator costs to deliver data and video traffic are growing at a greater rate than the revenues from those services. Unfortunately, the mobile operator network for delivery of mobile Internet and mobile broadband is built on a costly and fragmented collection of platforms and systems that are difficult and costly to manage, not always reliable, inefficient and unable to scale to handle the growing demands of customers. This is a poor starting point from which to take on the challenge of mobile Internet. What’s a mobile operator to do? Wireless carriers need to reduce the complexity and cost of their data services delivery network as well as facilitate the mining of value at the service level for their subscribers and business partners. Operators must build a single comprehensive environment that can manage and enhance the traffic – and it better be able to handle video both in terms of efficiency and user experience. But this is not enough. To really excel, the mobile Internet and data platform must be intelligent. It must be able to make decisions based on an endless number of factors such as user data, the data service in use, the handset or end-user device being used, network congestion, the time of day, etc. The systems that operators will deploy must be flexible and robust, providing the ability to make decisions about traffic as it flows through the network, including knowing when and how to touch the traffic, and just as importantly when to leave the traffic alone. And since video is such dominant component of data traffic, two-thirds of the traffic to be precise, operators would do best to begin here employing the 80-20 rule. Designing an effective video optimization solution is not trivial. Brut force versus intelligent video optimization When we take a detailed look at the video traffic, we learn, for example, that approximately half the videos watched are viewed only once while the remaining half are viewed several times. This is a service level statistic of video traffic, but we can also take a network view. When we analyze network traffic, we see that cell congestion is often limited to five to ten per cent of cells at any given time. Taking such traffic parameters into account together with additional network, service and user-related factors is the smart, more precise and cost-effective way to optimize video traffic. It’s also greener. The brute force method, whereby all video traffic is optimized regardless of the many factors already mentioned in this article, is extremely costly. It will work, but it will undoubtedly require enormous investment in data centres and processing power. When dealing with declining marginal revenues per bit, as wireless carriers are, the brute force method is less than effective to say the least. User experience Whichever method is used, it must take into consideration the impact on user experience. When performing optimization, operators must strive to provide continuous uninterrupted video play, short latency of start play, smoothness of video play, and high quality of sound playback. Ideally, the user experience will be improved by the video optimization and not be degraded. Viewers will only accept video optimization if the perceived experience is not affected. So the optimization method and platform need to substantially reduce the number of bits to reduce operator cost without perceptibly degrading the video quality and user experience. Intelligent, high-quality video optimization will solve network congestion during peak traffic while maintaining user experience and saving a significant portion of the high cost of network upgrading. Intelligent video optimization should be able to select the optimum combination of codecs, and bit rate and frame rate adjustments to substantially reduce the number of bits with no perceived degradation in video quality. In short, video and Internet service usage is growing at a tremendous rate, putting a great strain on mobile networks. Today’s mobile data infrastructures are still severely lacking and quite costly. It is vital that the methods and systems operators employ to cope with these challenges are capable of making intelligent, fundamental decisions, including sophisticated handling of video. Solutions are available in the form of intelligent video optimization platforms that can make smart decisions based on a wide range of network, service and user information to reduce the costs of video delivery while maintaining and often improving video viewing quality.
