Is 4G Worth It?

[krosenth]

A recent report by In-Stat made the rounds today and came to land in my Inbox.  The report was titled, "Big Pipes—The Global Market for Cellular/WiMAX Backhaul", but it was the title of the e-mail that caught my eye: "Growth in Mobile Data Triples Backhaul Capacity Demands".   I knew from having read Cisco's own VNI Forecast that mobile traffic was growing extremely rapidly, but "triple" is one of those words that you just don't hear that much around here anymore, at least not since the "bubble" days.

In the report, In-Stat notes that "operators are deploying EV-DO 2000, HSPA/HSPA+, WiMAX, and LTE to meet the growing demand for high speed mobile data."

And Frank Dickson, In-Stat’s VP of Mobile Internet rightly observes, “It does an operator no good to install a base station with 7.2 Mbps capacity if the backhaul is limited to 4.5 Mbps.” 

Obviously such a rapid and extreme build out holds great challenges for mobile operators in terms of resources, logistics, and costs.  And I was thinking of this monumental challenge in light of another article I read citing an IDC study that showed how, with voice revenues declining, operators are relying more and more on data revenues to keep ARPU up.  However, there's been no correlation between higher data speeds and greater ARPU.

So what is the motivation for 3.5 and 4G networks?  How can operators keep up with demand and minimize customer churn without running themselves right out of business?  Is the answer in the billing plans, the services, the network, or all of the above?

[MikeDemler]

In my opinion, there has been too much focus on bit rate in discussions of 4G technology.  Yes, faster is better but a more important issue right now is capacity. Clearwire had a very valid point in their presentation at the CTIA Wireless show in April, regarding total spectrum capacity.

Data ARPU will be growing as more consumers want to buy a smartphone as their next handset.  (A recent survey said 40% will choose a smartphone, and there is a fair amount of pent-up iPhone envy). But what will happen if consumers can't get access to mobile data applications because the network bandwidth has been consumed? We need to look at how many users the 3G systems can support.   Motorola has presented capacity comparisons of LTE and HSPA, which clearly show why 3.5 G solutions are inadequate.

Excuse the plug, but I have just published a report that studies these issues in some detail.  A free excerpt is available at http://www.digdia.com/4g/dc4g09_main.htm

-MIke Demler

Twitter: MikeDemler

[mgrayson]

So I think you raise an interesting question about the merits of continuing to throw more and more complexity at the "RF problem", or whether we are closing in on the diminishing returns as we approach the Shannon bound.

Agilent's Moray Rumney penned an excellent piece looking at the marketing of peak rates versus what is achievable by an average user or one at the edge of the cell - http://cp.literature.agilent.com/litweb/pdf/5989-9793EN.pdf.

"4G is aiming to be a low-cost solution for mobile broadband, but high effi ciency and high data rates mitigate against this. Everhigher peak rates drive up the cost of infrastructure and terminals, and high-effi ciency systems are inherently complex and therefore also drive up costs."

As we look at how best to deliver the traffic estimated to be required by the latest VNI study, we need to understand how to deliver the required Mbps/km^2 with the lowest Total Cost of Ownership - that means examining the optimum RF technology and access and aggregation backhaul systems.

[MikeDemler]

Mark,

Thanks for the pointer to the Agilent paper.  There is a lot of technical detail there, so I just did a quick read-through, but I did see some things I immediately differed with.  It may be a good thing that I am not a network engineer, because I look at things much more simply.

1. It serves as a reminder of how fast markets and hi-tech can move.  I gather that this was written in 2008, given the reference to the Olympics. At the time the author wrote: "3G in mobile phones has not lived up to expectations. 3G subscriptions are just over 7 percent".  Penetration of 3G is now more than 10%, with 40% of cell phone subscribers looking to upgrade to a smartphone as their next handset. But more important is that Data ARPU is approaching ~30-35% of wireless subscriber revenue and growing, while basic cell phone ARPU is flat or declining.
2. I think you are saying that we need a holistic view of the technology that makes up the network, and I agree. But I don't mean to suggest that the move to OFDMA RF technology is not significant.  It very much is. The WiMAX and LTE air interfaces, and their associated wider channel bandwidths and spectra, increase capacity and data rate simultaneously.  That is critical.
3. This problem is not unique to the Telecom industry, as I experienced this in the Electronic Design Automation industry as well - and that problem is that standards bodies don't make a standard a standard.. industry & market ecosystems and their networked (in a business sense) components do. 
   

In the end, 4G is not about what ITU or 3GPP or anybody else says it is.  In my opinion, the 4th generation of wireless telecom is already here, and it is marked by the move to flat IP based-networks with WiMAX or LTE air interfaces. Those 2 elements are basic, and that change is happening now.

The new 4G networks bring significant changes beyond (but including) faster data rates.  I find the 3.x nomenclature to be a bit pedantic. Setting goals for bit rate, etc. (as in the ITU-Advanced discussions) is fine but that neither defines or controls the change that is happening.  Those are the issues I cover in my report: The Emerging 4G Wireless Landscape in the U.S, Operators, Chip Sets, and Consumer Electronics .

[roth.eddings]

I agree with this post that there are two targets witrh regards to mobitly one being low cost the other being very reliable with the same features as lan based solutions I don't belive perfection is or should be available in the mobility market perhaps there should be a cheap solution, or a great solution but providing both within the same technology is not feasible.

Roth

[knagesh]

I believe it will be 'Worth it" in the mid to long run. In the short run, given the macro-economic conditions, many operators are choosing to extend the capabilities of their existing 3G networks.

To explain my rationale, we need to look BOTH on the technology side as well as business side. However, the industry value chain members must work together to make it worth it for all: consumers, service providers, app developers, and infrastructure vendors.

On the technology side, I see three compelling reasons for 4G.

• With 4G (LTE, WiMAX), for the first time, the radio networks are moving to truly end-to-end IP architectures. This will deliver a framework which is longer-lasting, open, and innovative. This, in turn, will stimulate the development of competitive markets on both radio and non-radio portions of the network. Several of us believe that there will be increased confidence in best-of-breed interoperability that decisions on radio and non-radio components may be increasingly decoupled. In the case of LTE, the LSTI Forum (of which Cisco is a member) is laying the foundation for infrastructure interoperability, including roaming services. An open API on the client/device, simplifying the job of app developers, will be a stretch goal!
• The bit efficiencies providing better spectrum utilization. Translates to lower cost, especially as the operators are scaling their data services.
• Video-based consumer and enterprise applications will dominate the way we communicate.

On the business side:

• Higher levels of subscriber maturity would imply lower churns. Mobile Quality of Experience will be a critical success factor.
• Higher performance network applications will promote innovative business models for consumer, SMB, and Enterprise segments. SMBs may do away with wired networks, if this strategy is executed right.

Similar situation exists with 3G networks too. 4G will magnify the problems and the opportunities. Hence, it is a market transition; not an evolution.

Nagesh