Mobile Video

Last month, the article “Mobile Backhaul Evolution” focused on the evolution of mobile wireless backhaul networks from TDM-based implementations to circuit-over-packet networks to packetized transport networks.

Continuing in the series of articles related to mobility, this final article will focus on one of the biggest drivers of mobile data growth—streaming video—and how mobile video is being deployed and optimized for delivery over 3G and WiMAX networks, and future LTE networks.

Mobile Video Drivers

To fully understand the drivers for mobile video, it is important to look at three trends that are creating a “perfect storm” of explosive data growth on mobile networks:

1. The growth of the Intelligent terminal: Consumer electronics manufacturers and feature phone vendors alike recognize the growing trend towards intelligent handheld devices that deliver a media-rich, interactive, and personalized experience to the consumer.

2. The rise of the “flat rate” data plan: Mobile data charging models have historically been difficult for the average consumer to understand. Users did not understand the concept of a “megabyte,” and data usage could not be equated directly into a time-based pricing model. Flat Rate data plans are allowing users to receive predictable charges for their mobile data usage.

3. Over-the-Top video applications (YouTube, Slingbox, Netflix, etc.) are delivering video experiences that are no longer tied directly to the home theater or Set-top box. Many of these applications are available on the mobile terminal itself.

4. The advent of “mobile broadband”: With messaging, regarding the Mobile Internet and broadband speeds over wireless networks, consumers are beginning to expect that a service is available to them anytime, anywhere—including when they are connected to a mobile service provider.

With these trends, it is no wonder that Cisco, in its Visual Networking Index, has forecasted mobile data traffic to grow at a Cumulative Average Growth Rate (CAGR) of 131% between 2008 and 2013. Mobile video specifically will account for 63% of all mobile data traffic by 2013, and is forecasted to grow faster (150% CAGR) than any other application category that Cisco measures. In support of Cisco’s forecasts, Infonetics Research expects mobile video subscribers to grow from 40M in 2008 to nearly 400M in 2013—a tenfold increase.

Streaming Media

Streaming media—live or on-demand—accounts for a majority of the mobile video traffic seen today. Service providers have partnered with aggregators or directly with content providers to deliver content to their mobile customers. The main driver of this streaming media surge is access to live sporting events and social/viral videos. The architectures for delivering mobile video vary significantly amongst operators, and can be broadly categorized into three classes:

  • Unicast video delivery networks: Unicast video delivery networks rely on RTSP, HTTP, or Adobe Flash-based streaming servers to deliver a unique video stream to each subscriber. Unicast video networks allow a service provider to leverage existing infrastructure (both transport and control), but are not highly scalable. Unicast video also allows a mobile operator to deliver an individual experience to the subscriber based not just on technical considerations (available bandwidth in cell tower, optimized video codec, etc.), but also subscriber preference (targeted advertising, for instance). Figure 1 illustrates how a unicast video delivery network would work.

Figure 1

Figure 1: Unicast Video Delivery Network

  • Multicast video delivery networks: Although standardized in the mobile industry (3GPP, 3GPP2, WiMAX Forum), multicast services have not been widely adopted by mobile operators. Multicast solutions allow a mobile operator to achieve greater reach for a mobile video stream by reducing the amount of data traffic, but still consume a great deal of operator spectrum to deploy. Figure 2 provides an illustration of Broadcast Multicast Services (BCMCS), a multicast network for CDMA.

Figure 2

Figure 2: Broadcast Multicast Services (BCMCS)

  • Broadcast video delivery networks: Broadcast video delivery networks rely on overlay networks and separate spectrum for the delivery of video services to devices. Since these networks are completely isolated, they require devices that have a separate radio for receiving mobile video signals. However, with such a large number of Points of Presence (PoPs, i.e., cell towers) in the mobile network, and a large number of channels offered, the likelihood of having multiple users in the same PoP watching the same content at the same time is actually very low. The most common examples of broadcast architectures for mobile video delivery are Digital Video Broadcasting – Handheld (DVB-H, Figure 3) and Qualcomm MediaFLO.

Figure 3

Figure 3: DVB-H

Quality of Service

Quality of Service (QoS) for mobile video can be achieved only for unicast video traffic. For multicast and broadcast, the mobile service provider must deliver video to the “least common denominator.” This ensures that every device that is part of the multicast or broadcast group receives a consistent experience. Offering this level of consistent experience over a unicast network is much more difficult due to the unpredictable nature of RF propagation and consumption trends.

Unlike today’s IPTV-based solutions for implementing QoS for video, such as Call Admission Control (CAC), implementing QoS on the mobile network requires a different approach. The challenge with offering QoS and guaranteed experience over a mobile network is that network conditions change so rapidly and so drastically, even when a subscriber is stationary, relying on tools that assume some baseline level of network quality is not possible.

Many service providers have implemented various methods for providing what is becoming known as “bitrate optimization.” This method relies on either an inband or out-of-band signaling channel between the client and a network element. The network element learns about conditions on the mobile device, including lost frames, and determines how to best handle the next video frame—such as deliberately dropping a P-Frame if the correlated I- or P-Frame has been lost in transit. “Bitrate optimization” has been effectively deployed by a number of operators worldwide.

Another solution for providing optimized video delivery mobile device is content transformation. A content transformation network element provides on-the-fly transcoding/transrating of video content to a format that can be processed by a mobile device and can be transported efficiently with the available capacity within a cell. While this method provides a significant advantage to a mobile operator, the amount of processing required to transcode video traffic in real-time may not make the solution economically viable.

Summary

This article offered some insight into the mobile video trends and topologies, and it also concludes the mobile Chalk Talk series. These articles were intended to provide the readers with a high-level view of the technologies, services, and standards that are shaping the future of the Mobile Internet, and to help develop an awareness of the intricacies and nuances in providing mobile data services. The mobile experience continues to evolve—from the days of voice-only networks, to an evolution of data services, to the deployment of high-capacity, interactive, personalized, and rich content delivered to any device, anywhere.

If you have any questions regarding the topics discussed in this article, you can email questions to Kevin from now through October 1, 2009.

About the Author:

Kevin Shatzkamer is a Customer Solutions Architect at Cisco Systems with responsibility for long-term strategy and architectural evolution of mobile wireless networks. He has worked at Cisco and with the mobile wireless industry for nine years, focusing on various technologies ranging from GSM/UMTS to CDMA networks, packet gateway, network-based services and security, video distribution, Quality of Service, and end-to-end design theory. Kevin has 16 pending patents related to all areas of work. Kevin holds a Bachelor of Engineering (BE) degree from the University of Florida and a Master of Business Administration (MBA) degree from Indiana University. He is also an author of IP Design for Mobile Networks, a Cisco Press book that details IP’s role in current and future mobile networks.

Kevin Shatzkamer

IP Design for Mobile Networks

IP Design for Mobile Networks
Mark Grayson, Kevin Shatzkamer, Scott Wainner
ISBN: 158705826X
Pub Date: 6/15/2009
US SRP $60.00
Publisher: Cisco Press