AT&T to Ms. Parks: “Go to the back of the bus!”

You may have noticed this news item about a class action lawsuit against AT&T Mobility and Radio Shack.

So, how did we get to the place where Billie Parks, an honest subscriber, gets surprised with a $5000 bill?

Fundamentally, the problem is that mobile broadband has deployed onto a scarce resource, namely wireless spectrum. So, YES, 3G and 4G technologies are capable of multi-megabit download speeds, and, NO, they cannot provide those speeds to all of the subscribers in a sector at the same time.

The solution to this problem, in AT&T’s view, is to charge an additional fee for each Megabyte of data consumed above 5 Gigabytes during the month.

Huh??? How does that solve the problem???

I guess in AT&T Mobility’s view the answer is obvious: the economic penalty for consuming more than 5GB is so steep that Billie Parks will willingly stop consuming data when her limit is reached. Then, with Ms Parks offline, or in the back of the bus, so to speak, there will be room on the channel for other subscribers to consume their share of those multi-megabit download speeds.

That logic is flawed and has appeared to work for a few years, but only because 5GB is plenty for the road warriors using laptops that need ubiquitous access to email, and because the devices that consumers use are not well suited to consuming lots of entertainment. Why would anyone want to pull 5GB of video down to a cell phone???? They wouldn’t, of course.

But Billie is a consumer, interested in entertainment, and using a laptop.

AT&T Mobility extended the subsidy model beyond cell phones to include ultraportable laptops with built-in HSPA, and the result is a very attractively priced computer.

If you are one of the many computer users that have grown to love watching TV shows via Hulu, then you can imagine why Billie Parks ran over her limit. The logic goes like this:

Billie: “I have a trusty old computer at home and I watch my favorite Hulu TV shows on it. AT&T and Radio Shack jointly offered to sell me a very attractive new laptop for $100. How can I pass that up? Now I can watch my Hulu shows anywhere!”

AT&T’s response: “Yup, you sure can.” (And in fine print): “Oh, but be sure to stop watching when you get to 5 GB.“

And then Billie is supposed to say: “OK, I’ll check my account on AT&T’s website frequently and the 5GB limit will probably be reached at a time that is convenient for me.”

Yeah, right.

The real solution to the problem lies not with economic penalties, but with the integration of billing systems and QOS controls. I’ve mentioned wireless broadband QOS in previous posts (here, here) and I’ll keep bringing it up because it is a topic I’m passionate about.

Billie Parks was offered one choice of service:

• Best Effort Hi Rate Pkt Data (HRPD), with a 5GB cap, for $60 per month
  • Good for email, browsing and file downloads
  • Also good for Hulu if the sector is lightly loaded

But mobile packet data service needs should be segmented into multiple categories:

According to the segmentation in that table, here are the service options she could be offered instead:

• Best Effort Low Rate Pkt Data (LRPD), no cap, $10 per month
  • Good for email, Machine-to-Machine telemetry, GIS tracking, instant messaging
• Streaming LRPD, no cap, $15 per month
  • Good for PTT voice but not full-duplex voice
• Realtime Streaming LRPD, no cap, sold in minute usage bundles
  • Good for voice telephony
• Best Effort HRPD, no cap, $40 per month
  • Good for browsing and file downloads, but you may not be satisfied with the result if you try streaming data – depends on sector loading
• Streaming HRPD, no cap, $60 per month
  • Good for video entertainment like Hulu
• Realtime Streaming HRPD, no cap, sold in minute usage bundles
  • Good for video telephony (a future service)

(My price suggestions are just that, and the market will ultimately determine pricing.)

Note that I proposed no caps on any service. When QOS controls are used it’s not necessary to also use caps and penalties to motivate certain usage patterns. QOS will allocate the available bandwidth in the sector according to the service that the subscriber is authorized to receive. And if too many subscribers are authorized for the available bandwidth, then the next subscriber will be denied service, or offered a lower-grade of service. In this manner, when a sector approaches its capacity limits, users with “best effort” subscriptions will find their connection to be slow and intermittent, while users with “streaming” subscriptions will continue to receive satisfactory service. Wa-la! People get what they paid for, with no caps and no billing surprises.

My friend Dan commented about Billie Parks’ lawsuit at lunch yesterday, saying “she’s the Rosa Parks of wireless.” Hopefully her stand against the archaic usage caps will accelerate the deployment of QOS based service offerings on the part of the mobile operators. I’m definitely pulling for her to win this one.

Wireless Tectonics

There are two large technological land masses that dominate the lithosphere of wireless personal communications. Land Mobile Radio is the stable and slowly evolving workhorse of Public Safety, and many residents of that wireless continent are barely able to perceive any movement or change in the terrain. In sharp contrast, 3G Wireless is the hyper-volcanic and rapidly advancing mainland that is shaking the planet, and occasional meteor strikes re-shape its landscape overnight (think Apple iPhone).

 

And while it may have appeared that these two technologies were peacefully coexisting, LMR dominating Public Safety and 3G dominating the mass markets, we can now see that within the Public Safety arena they are actually colliding. The unfolding of that collision is not widely followed outside of the Public Safety industry, but it is a most interesting story to follow, especially if you are a follower of the strategies of all things wireless.

 

Continuing the metaphor for a moment, I’ll enumerate some of the contrasts that comprise the “fault-line” between these two technologies:

 

LMR networks support Push-to-Talk (PTT), which is half-duplex, group-based, voice communications.	3G networks support group based PTT, as well as 1-to-1 PTT and full-duplex voice
Most LMR radios will not operate outside of their home network	Most 3G devices roam successfully beyond their home network
LMR radios are expensive, often costing a few thousand dollars per unit	Most 3G devices are available for less than a few hundred dollars
The prospects for streaming video to an LMR radio anytime in the near future is not good	There are at least 3 video streaming services for 3G devices available today (MobiTV, VCast, MediaFLO)
LMR radios require auxiliary equipment to provide a GPS location	Most 3G devices have integrated GPS capabilities and also support Assisted GPS so that indoor location fixes are possible

 

3G looks very favorable, but as there are two sides to every coin so there are some very good reasons that a firefighter running into a burning building carries an LMR portable unit rather than an iPhone. Here’s some advantages for LMR:

–          Demonstrated ruggedness – confidence that the radio will operate regardless of how it is treated.

–          Even handhelds have big batteries and 3 watts of power

–          Spectrum bands that propagate farther and penetrate better than most cellular bands, especially PCS

–          Base stations are equipped with backup power

–          Simplicity of design, at least for conventional LMR: when you press the Talk button you get to talk, period. No arbitration by a computer about who gets to talk. If two people are talking at the same time then the one yelling the loudest will be heard the best. That’s what I would want if I were suddenly being fired at by the driver I just pulled over for running a stop sign.

 

In short, an LMR network is designed to be a life-critical communications tool that is exclusive to the government or enterprise organization that owns it, while a 3G network is designed to provide ubiquitous, profitable advanced services for the mass market.

 

These “continents” don’t converge very nicely. Or do they? I propose that this convergence will proceed incrementally, without major “earthquakes” as 3G enters the Public Safety domain, and with great benefits to Public Safety organizations. Convergence may be the wrong term however, because for the foreseeable future I believe that both network technologies will continue to exist independently. What will be different is the relative roles of LMR and 3G in Public Safety communications.

 

In Public Safety today, LMR dominates while 3G augments LMR and provides some data capabilities at better cost and performance than LMR can. For example many police vehicles are equipped with laptop computers that are wirelessly connected using 3G aircards.

 

But the domination and augmentation roles will gradually reverse. Eventually, 3G networks will become the dominant solution for Public Safety voice, video, data, and location based services, and LMR will augment 3G by providing life-critical communications to specific types of front-line first responders.  Two large steps towards this future are about to occur:

1. The interoperation of PTT voice between commercial 3G networks and private LMR networks. With this capability the Public Safety PTT talkgroups will begin consisting of a hybrid user base – some carrying radios and others carrying cellphones. Regardless of the device type, when a member of the group presses the talk button and speaks all of the group members will hear it. Once this capability achieves moderate deployment the role-reversal will begin in earnest; personnel that once were loathe to part with their radio will find themselves seduced by the capabilities offered by the cellphone.
2. The 700 MHz D Block  auction will complete sometime in 2009. Even though the networks built for this spectrum must give special priority to Public Safety users none of them will be built out with LMR gear. The D Block will be deployed with some mix of 3G and 4G technology, or perhaps 4G exclusively.

 

Both of these advancements were recently tested by the US Department of Homeland Security, in a pilot called Radio Over Wireless-Broadband (ROW-B).

 

As this role reversal proceeds the market is adjusting and beginning to incorporate other LMR-style capabilities into 3G networks, for example:

–        PTT will become more capable and widely deployed on 3G (VzW  & Sprint ).

–        More rugged devices will become available (see Sonim , Casio )

–        Backup power for base stations will arrive sooner rather than later

 

The next several years will see some dramatic shifts for residents of the Public Safety continent as the 3G world encroaches. LMR will continue to provide a steady hand-hold as the convergence proceeds, but the landscape of new capabilities that regularly arrive will remind all that this is a new world.

Just another brick in the Walled Garden

(My apologies to Pink Floyd regarding that title)

  

So, what exactly is the wireless Walled Garden? The trade press has often talked about the Walled Garden as a concept. But for bottom-up thinkers it’s a difficult concept to characterize in pragmatic terms. In this posting I’d like to get underneath the Walled Garden and discuss some of the bricks that comprise the wall.

 

For starters, I’ll take a crack at a definition for “Walled Garden” as a concept:

 

The “Walled Garden” is a broad strategy deployed in varying degrees by Mobile Operators to enable their participation in the value created using their networks. Since nearly all of the value created on top of a network is associated with the content that travels via the network, a Walled Garden is designed to control content.

 

What do you think of that definition? Feel free to chime in with your thoughts.

 

I suppose that the opposite extreme of a Walled Garden is a Dumb Pipe. Most wire-line ISPs are Dumb Pipes, of course, and have only indirect participation in the incredible wealth that has been created via the Internet. The Net Neutrality debate is relevant here, with one side claiming that the ISPs must only accept Dumb Pipe status, and the other side claiming that network investment is discouraged if the ISPs cannot make content based service decisions. That debate is beyond the scope of this entry but will certainly come up in future postings.

 

In Mobile networks, a Walled Garden strategy is implemented by the use of several technical or business mechanisms for controlling content, which I’ll call ‘bricks,” each of which serves to enable the Operator to decide how they want to participate in the value created via the content.

 

The most lucrative content on wireless networks continues to be full-duplex voice, that is, phone calls. The Mobile Network Operators (MNOs) prefer to keep voice revenues completely to themselves, for good reason, and therefore have deployed several bricks intended to thwart other companies that would like to deploy packet voice telephony (VoIP) on their 3G networks.

 

Here’s a quick look at the bricks I find most interesting:

• SS7 control of Circuit Voice connections. The SS7 network is completely locked down, and no one can access circuits on a Mobile Network without an explicit interconnection agreement. Access is allowed though; MNOs have partnered with MVNOs that resell circuit voice minutes on the Mobile Network, and those resale agreements allow the MNOs to retain significant margins (value).
• Quality of Service (QOS) controls. Given the challenges to accessing circuit connections on a wireless network, packet connections are a necessary alternative. But, the most valuable application, full-duplex voice, requires an end to end packet data connection that operates in a narrow range of tolerance for latency, jitter and packet loss. MNOs have reserved this level of QOS for themselves. This brick is very interesting, and I’ll post a separate blog entry on it later.
• Application Programming Interfaces (APIs).  APIs on mobile devices serve to open up cracks in the Walled Garden where third parties can innovate independently of the Mobile Operator. However, those APIs are strategically deficient, or are offered only to third parties that enter into strategic partnerships with the MNO. The MNOs welcome a limited amount of third party innovation, but capabilities such as generating efficient voice streams, or acquiring a geo-location from an indoor location, are often reserved for selected partners that are willing to exchange significant margins (value) in return for access.
• Firewall rules. Many applications that could run on mobile devices will require always-on connectivity, with a static publicly routable IP address that can be provided to other devices and servers. (Think of push-email). This allows the application to be summoned into a session that it did not initiate. Voice telephony is a good example again, where the device carried by the person being called must be reachable via an IP address known to a server outside of the MNO’s network. Unfortunately, the MNO’s configure their firewalls to expire TCP ports such that even though the port and address for the device remains valid the firewall will not allow it to be contacted. While session border controllers (SBCs) exist to solve this problem on the wired Internet that approach is not as effective in the wireless domain since the frequent client-initiated contact to the SBC to keep firewall ports open results in excessive radio activation and therefore poor battery life.
• Others: I’ll list a few more here but this posting is already too long so I won’t elaborate on them
  • Application Certification.
  • Processor cycles available to Applications
  • API performance
  • API functionality (Vocoder access, QOS access)
  • A-GPS functionality access

 

Note that I did not mention device subsidies, since I don’t view them as components of the Walled Garden.

 

So, is the Wireless Walled Garden going to fall down? I think so, but vestiges of it will remain for some time. For example, in all of the hoopla surrounding the “Open Access” debates, the 700 MHz C Block and Google’s pressure on the industry, we heard the large MNOs express willingness to begin loosening their controls and allow “any device” to operate on their networks. But, did they offer to grant VoIP grade QoS to those devices? The answer is no; that topic didn’t even come up and the trade press is not yet knowledgeable enough to ask the question. Open Access in that form is a step forward, but a smaller step than you may have thought.