4G Ecosystem Delivers New Capabilities, Devices & Participants

In this second article based on our series of interviews, Jose Puthenkulam, Director of WiMAX Standards at Intel provides his perspective on the evolving 4G ecosystem including user expectations, devices, mobile operators, spectrum, network infrastructure, business models, and new applications.

Mobile WiMAX & LTE: 3G networks evolving to 4G

To better understand the evolution of 3G to 4G networks, it is helpful to first examine them from a standards perspective.  It may not be widely known that ITU-R M.1457 recommendation (AKA 3G -IMT 2000 Release 1) - includes both IEEE 802.16e-2005 ("Mobile WiMAX") and the 3GPP LTE Release 8.  In fact, many of the core 4G ingredients are already in those standards: OFDMA, flat- all IP- network, fixed or mobile operation, MIMO, hybrid ARQ (automatic repeat request) at the PHY layer, multi-megabit speeds delivered to users, advanced FEC, etc.  Incremental improvements are also being made to the evolving IEEE 802.16m and LTE Advanced standards in order to align them with the 4G requirements specified in the ITU-R IMT-Advanced recommendation (which has yet to be finalized and numbered). 

The upshot of all this is that both Mobile WiMAX and LTE are officially designated 3G technologies, which are evolving to 4G in their respective standards bodies (IEEE 802.16 and 3GPP).  The race to 4G has been accelerated by a number of factors; the early deployment of Mobile WiMAX is one of the most important.

For more information please refer to:

4G Myth Busting: Intel's March 25th talk on Advanced WiMax

ITU-R Recommendations on IMT-2000


4G User Expectations and Network Capabilities

Users want a "seamless connectivity" experience for mobile Internet access.  They want to be able to easily access well-known Internet brands (e.g.  Google, Yahoo, MSN, Amazon, Facebook, etc) from whatever notebook PC or device they are using at the time.  That device could be a netbook, smart phone, MID, or other gadget.  Users expect seamless hand-over between base stations, roaming between carriers, and no dropped calls or Internet session.  And they want speed too- at least 2 to 5 M bit/sec average data rate over a mobile broadband network. 

4G will mark a huge transition from circuit switched voice and TDM wireless transmission to an all IP packet switched network and VoIP.  To a very large extent, that has already happened in fixed line networks (DSL, cable, or wireless), where VoIP is available over a broadband network from many different service providers.  Puthenkulam expects VoIP over WiMAX handsets to be available from Clearwire in 2010 and VoIP over LTE to be available by 2012 (presumably from VZW and Vodafone). 

4G mobile network technical capabilities must be in place to support user expectations and requirements.  These include: a flat (non-hierarchical) all IP network, MIMO on uplink and downlink, spatial multiplexing with transmit diversity, schemes to minimize packet loss (e.g.  advanced FEC and ARQ), robustness (high availability), reliability, and excellent security.


4G Devices and Network "Stress Testing"

While smart phones are arguably viewed as the driver for 4G networks, it will actually be notebooks (and possibly netbooks) that will stress test those networks.  Smart phones will work much better on a 4G network (higher speeds and lower latency) - they don't have enough storage capacity to send and receive large amounts of data or video so they won't sufficiently impact network capacity.  Imagine emails exchanged with tens of Megabytes of attachments.  While this is routine using a notebook PC, it would be painfully slow on a smart phone.

Within any given geographical area, the 4G network must serve many users at high speeds, high throughput, and low latency.  The network must not crash if all potential mobile data users were to access the network at the same time (Note: this often happens with WiFi used in conference rooms or at conventions).  Network capacity and planning for 4G must consider clusters of "heavy duty users" sending large volumes of data.  Multiple concurrent PC users -downloading or uploading large video files or presentations - could potentially break the network.  So 4G network capacity must be matched and scaled against this multi-megabit file transfer scenario, with several interleaved real time applications (e.g.  voice or video conferencing) that are tested for low latency.

In the future, Puthenkulam believes that the mobile Internet will be the prime application of smart phones, rather than voice.  He believes one of the reasons for this is the richness of alternative communication media like Email, Twitter, Facebook etc.  Music streaming will also become more popular in his view.  Internet radio will overtake satellite radio in a 4G world (this author strongly disagrees).  Personal casting¹ via media servers and social networks (like Twitter or You Tube) will be an important capability of 4G smart phones and all-in-one devices. 

Vehicle entertainment systems and devices have a lot of potential in Puthenkulam's opinion.  In one scenario, you could use a handheld device credential to authenticate a "4G terminal" in a car.  Once authenticated, that 4G terminal could provide a variety of services and applications - from emergency road service with automatic location ID to music streaming to portable on-line gaming applications (presumably for kids sitting in the back seat of the car, rather than the driver).

Smart sensors and meters for infrastructure applications, such as smart grids, could be very effective in a 4G network.  They could monitor how much energy was used in the home or office and optimize energy use by controlling a thermostat or other appliances.  Home solar cells could also be connected through a 4G enabled smart grid.  Other innovative machine-to-machine (M2M) devices and applications are possible using a 4G network.  An automated parking attendant that can keep tabs on empty parking spaces and time stamp those that are occupied is one example.  Child monitoring systems in a day care center is another.  Various forms of web server-to- web server communications capabilities for mobile commerce are also a possibility

Editor's Note:  Mobile e-business was envisioned to be a big application for web services in 2003, but six years later, it still hasn't happened yet.


The Evolving 4G Operator Ecosystem, Spectrum, and Network Infrastructure

Both Mobile WiMAX and LTE have similar capabilities from an operator service and end user standpoint.  To upgrade to either (from 2.5G or 3G) will require a new Radio Access Network (RAN), which will be about 90% -to-95% of CAPEX.  The other 5%-to-10% of CAPEX will be in Core IP Network Equipment.  The transition between 2.5G or 3G to an LTE or Mobile WiMAX 4G network is depicted in the presentation by Siavash Alamouti, Intel Fellow and Chief Technology Officer of the Mobile Wireless Group.

Editor's Note:  The incumbent cellcos are almost all planning for LTE, while the majority of Greenfield operators have chosen Mobile WiMAX.

A key question is how expensive will the 4G network equipment and devices be? Mobile WiMAX has already gone through two cycles of cost reduction in terms of chip sets and base stations.  LTE has not yet been deployed, so is already behind Mobile WiMAX in the cost reduction cycle.  Puthenkulam believes that this gives WiMAX a big advantage over LTE.

The duplexing method (for simultaneous transmission between the base station and subscriber unit) is a very important issue when considering spectrum utilization.  The "cellular industry at large" has to support a set of FDD legacy networks, including GPRS, EDGE, GSM, CDMA, and CDM2000.  So when cellular operators turn on LTE service, it will need to inter-work and be compatible with those networks, implying that the initial LTE deployments will be based on FDD.  China Mobile is an exception, having already deployed a TDD network (TDS-CDMA), they are planning on a TDD version of LTE.

Most Mobile WiMAX deployments use TDD, which is much more bandwidth efficient than FDD.  It uses half of the FDD spectrum, which requires separate frequency channels for upstream and downstream directions of transmission and is typically offered as a pair of channels, as depicted in the Figure below:
 



FDD can be quite inefficient (waste bandwidth) in a given direction of transmission, usually upstream, when carrying asymmetric data services.  This is because the actual data traffic may occupy only a small portion of the upstream channel bandwidth at any given time.  Mobile WiMAX traffic will be dominated by asymmetric data (e.g.  much higher downlink than uplink traffic), so TDD is a better choice. 

TDD uses only one channel for transmitting downlink and uplink sub-frames via two distinct time slots within a single frequency channel.  TDD therefore has higher spectral efficiency than FDD.  Moreover, using TDD downlink to uplink (DL/UL) ratio can be adjusted dynamically.




Because most Mobile WiMAX operators are not cellcos, they don't have to be backward compatible with a 2G or 3G FDD network (Sprint as a WiMAX MVNO is an exception).  So they instantly realize a savings in licensed spectrum, which is a very scarce and expensive resource.  As a result, they are able to provide the most spectrally efficient technology that is available on the market now.

For more on this topic, please see:
Settling the confusion of WiMAX duplexing method: TDD or FDD or both?


Fixed line Operators will move to deploy 4G networks

Let's look at fixed broadband (DSL or cable) network operators that want to be 4G network providers.  BT, Comcast, Tata Communications, and BSNL (India government operator) come to mind.  To offer a quad play service bundle, those operators need a broadband wireless/mobile network.  They can either build it themselves or lease capacity and become a MVNO. 

Puthenkulam firmly believes that Mobile WiMAX is a better choice than LTE for those fixed broadband operators, because it is a "native broadband wireless technology." By deploying a single infrastructure, they are able to provide mobile service as well as a variety of fixed line services, e.g.  VoIP and entertainment video. 

On the other hand, LTE (for most operators) needs to be backward compatible with the TDM based 3G network infrastructure, which is not well suited to carrying broadband data.  It's not just the Radio Access Network (RAN), according to Puthenkulam.  The core infrastructure must be substantially upgraded, with higher capacity routers.  So a fixed broadband operator choosing Mobile WiMAX is "one-up" on the cellular operator choosing LTE.  Comcast, Time Warner and Brighthouse investment and partnership with Clearwire is a perfect example of this strategy.  Comcast has just announced it will provide mobile WiMAX service and laptop cards in Portland, OR, as a Clearwire MVNO.  More cities are scheduled for roll-out later this year, tracking CLEAR deployments.

While it hasn't happened yet, Puthenkulam believes that these same fixed broadband operators can migrate from the flat rate VoIP calling plans they have now (e.g.  Comcast) to mobile VoIP, which he thinks will be a lot cheaper than traditional mobile voice service.  Puthenkulam makes the following bold prediction: "Network operators that know how to deliver mobile VoIP - in the most efficient way and with roaming- will thrive and attract a lot of subscribers.  Mobile VoIP will rewrite voice calling as we know it."

Puthenkulam observes that many cellular operators have not completed their 3G roll-outs yet.  This is especially true in emerging markets, where he thinks operators should skip 3G entirely (it's too expensive and inefficient of bandwidth) and go directly to 4G.  Mobile data will be the biggest growth engine for the next three to five years and 3G networks won't be able to support the explosion in mobile data traffic that we are already observing. 

Editor's Note:  This has already been a well-documented problem for AT&T in supporting iPhone traffic over its HSPA based 3G network.  For example, AT&T does not allow iPhone tethering (use as an external 3G modem for notebooks) and completely blocks Sling Media video traffic on its mobile network.

Mobile WiMAX doesn't have the constraints of 3G, which is essentially a packet overlay (HSPA or EVDO) to a TDM network.  Mobile WiMAX is more spectrum efficient (with OFDMA, MIMO and TDD) than 3G networks.  And through the efforts of the Open Patent Alliance (OPA), the intellectual property licensing costs for Mobile WiMAX will be significantly less than for 3G technologies (especially CDMA based). 

The upshot is that network operators moving directly from 2G or 2.5G to Mobile WiMAX (802.16e now; 802.16m later) makes a lot of sense to Puthenkulam.  Nonetheless, Intel will soon be able to provide 3G-HSPA modem technology for mobile computing devices, based on a licensing deal with Nokia announced June 23rd by Intel's Ultra mobility Group (which is responsible for MIDs and Smart Phones).  It will enable Intel to have a more complete wireless technology solution portfolio, according to Puthenkulam. 


Managing Network Capacity: Wireless Backhaul and Reduced OPEX

Mobile WiMAX is well positioned to use wireless backhaul to reduce OPEX costs over wire-line backhaul.  We've seen Dragonwave capitalize on this trend by using micro-wave radio to backhaul WiMAX network traffic.

In the end to end reference architecture defined by the WiMAX Forum NWG, network traffic from several base stations are aggregated to a single backhaul point by the Access Service Network (ASN) Gateway.  The ASN Gateway backhauls the aggregated traffic to the broadband service provider, ISP, or MVNO core IP network (depending on who owns the backbone network). 

By aggregating capacity at a smaller number of backhaul points, the business model can be more flexible, including the offering of un-metered, flat rate billing plans.  Puthenkulam believes that metering traffic will only impact very high bandwidth WiMAX users.

In particular, the WiMAX Forum end-to-end reference architecture supports sharing of the network in a variety of ways and business models:
- Network Access Provider (NAP) owns and operates the network.
- Network Service Provider (NSP) owns the subscriber and provides service. 
- NSP shares the NAP or a NSP uses multiple NAPs.
- Application Service Provider (ASP) provides application services.

According to Puthenkulam, the WiMAX Core IP network² consists of "off the shelf" IP network building blocks with specifications based on IETF RFC's, e.g.  IP routers, Authentication Servers, OSS, BSS, mobile IP home agents, etc.  While WiMAX does not require any specific core IP network, 3G and LTE require specific core network building blocks that are specified by the 3GPP.  In particular, LTE uses an "Evolved Packet Core," for managing network traffic.  Dedicated 3GPP defined Network Elements are used here, rather than generic IP network equipment.


Media Ecosystem is Evolving

As people grow tired of channel surfing their TVs to watch programs and movies, technology is providing new opportunities on when, where and how people view content.  Consumers are now taking advantage of You Tube™, Hulu.com and other sites to watch web based streaming video and audio.  In addition, there's more use of DVRs and VoD to watch favorite TV shows, news and movies. 

By now, we're all familiar with this time shifting theme.  But now it's being combined with place shifting (e.g.  watching video on notebook PCs or smart phones via Sling box).  We are now entering the era of "media personalization," where you can watch whatever you want when you want it and wherever you are.  The key driver that will make this happen is mobile broadband access, especially 4G networks.  Of course, we'll need roaming and hand-offs between mobile networks to make it a reality, but it's coming.  In this vision of the future, almost every type of media will be delivered over IP.  There will be very few exceptions.  There's a huge demographic change that favors web content, delivered over 4G networks, to notebooks, netbooks, and mobile computing devices.  Puthenkulam says that all user devices will have 4G-network access and be able to watch good quality video. 

Editors Note:  There is a difference of opinion as to how mobile video should be delivered to cell phones and other hand held devices.  Some favor using a separate, dedicated network for broadcast video- like Qualcomm's Media Flow.  Others favor using the existing cellular network to watch video on devices, like the software download from MobiTV.  AT&T has currently chosen to block Sling box transmitted video over its 3G cellular network.  This may all change when 4G networks get deployed.  Let's see.


4G Business Models Favor More Services Delivered at Lower Cost (than 3G)

With a flat- all IP network, all information flows over the same Network layer protocol (IP).  Therefore, the network operator will be able to offer the user a "pay for what you use" pricing model that will be a powerful motivator to develop and sell new services.  In this scenario, broadband Internet access will be standard and there will be small, incremental prices for additional services, like VoIP or mobile real time video.  This will be a powerful motivator for innovation.  It will very likely stimulate application software developers to create new and useful applications, which will broaden the market even more.  This will attract new users and drive costs down due to economies of scale.  Some of the new services and applications include: home security and power monitoring, child-care center monitoring, telemedicine and health check-ups.  Certainly, there will be many more that we haven't thought of yet.

There will also be radio technology innovation that far outpaces 3G.  This is clearly evident in MIMO and beam forming that has been included in Mobile WiMAX Wave 2 products (see the first article in this series for an explanation).


Closing Thoughts from a 4G Visionary

- IEEE 802.16 and Mobile WiMAX technologies have all the capabilities to serve the broadband mobile Internet operators and users.

- The fusion of different services that will be delivered over "all IP" 4G networks will enable operators and application developers to custom tailor what you get to the device that you have. 

- End users will benefit immensely from the applications that are spawned by the availability of 4G mobile networks.  Global change will be brought about through mobile Internet based communications.

- Jose Puthenkulam firmly believes that 4G networks will be affordable and will empower the people that use it.  We certainly hope he is right.


Disclaimer:  The views and opinions in this article are those of Jose Puthenkulam and not those of his employer - Intel Corp.  Please refer to Intel's web site or contact their PR department for the company position on the topics discussed.  In addition, please see the first article in this series for the genesis of WiMAX standards. 

________________________________


(1) - Sending small messages, which are broadcast to the people following the sender.  This is how Twitter and other social networking tracking sites work.

(2) - This author believes Cisco's primary WiMAX initiative is to sell "Greenfield" network operators a complete "Core IP NGN" system, which may or may not include WiMAX RAN equipment.





 

Wednesday, July 01, 2009 in Business  | Permalink |  Comments (16)