Becoming Web Integrated – More Than Just a Website

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)

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.

Many of you are also aware of the rapid advancement of Machine to Machine (M2M) communications and the dozens of MVNOs sprouting up to focus on it. That industry uses 2G and 3G devices integrated into equipment, such as oil rigs, billboards and windmill generators, to transfer the real-time and non-real-time data needed to keep performance high and costs low.

But 3G is showing up in lots more places too, and we’ve only just begun to see the spread. I am personally intrigued with the use of 3G to enhance devices that we were familiar with in a prior version . In a previous posting I mentioned the recent arrival of 3G integrated into Personal Navigation Devices. I want to take that theme farther now and start building a list of newly connected devices.

So, here’s the criteria for making it on to the list:

-          Human oriented devices, that is, unlike M2M the purpose of the device must primarily serve for human consumption.

-          One of the following:

o       Devices whose traditional functionality is enhanced via the 3G connectivity.

o       Devices whose traditional functionality is maintained while valuable new functionality is added via the 3G connectivity

o       New devices that weren’t feasible or practical without the 3G connection

-          And of course, primarily voice oriented devices, such as cellphones, are excluded

Here’s the complete list of 3G enabled devices that I am currently aware of:

-          Amazon Kindle – A true breakthrough product, and even more so because I believe that most Kindle owners are only vaguely aware that the thing is using Sprint’s cellular network.

-          Personal Navigation Devices – notable for the absence of Garmin or Motorola

o       Dash

o       Insignia

-          IVC Remote Video Security Camera – Good news for construction companies. Now you can remotely monitor any location that falls within 3G coverage.

-          Vizit Photo Frame (coming Spring ’09) – In my view a digital photo frame is most useful in the role of bring Grandma into the loop of digital photography, allowing her to appreciate the photos you take of the kids without requiring all the messing around with printing. But, how many Grandmas have WiFi????  Send her this frame in the mail. She plugs it in. You send photos to it. Simple.

-          uConnect Web service (aka Autonet Mobile) available in 2009 Chrysler vehicles. Now the kids can keep their iPod Touches operating while the family is on the road.

-          LG Watch (LG-GD910) – OK, just what you always needed, video streaming to your wrist!

Please point me to other products that you know about and I’ll add them and give you credit.

Here’s some devices I’d like enhanced with 3G:

-          YouTube Video Cameras

-          Digital Cameras

-          Handheld Multi-Function Displays for aviation, e.g. the Garmin 696, using Aircell’s EVDO network

I’d love to hear your ideas for products that are ready to be connected.

Motorola has world-class capabilities in 3G devices, and yet their PNDs are not connected. That’s very hard for me to rationalize, and I would love to hear ideas for what their strategy is here.

Hopefully, their next PND will include 3G connectivity and some cool new features that uniquely leverage it.

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:

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.

“I send my daughter off to college with better communications gear than my Katrina response teams were equipped with.” Rear Adm. Robert F. Duncan, Former Commander of the Eighth Coast Guard District.

Most police and fire personnel in the U.S. use analog radio technology (Pause and contemplate that for a moment). In a world where 3G has penetrated almost every corner, there remains a large market of users that have little or no access to the benefits of broadband wireless that we’re becoming accustomed to. My next posting will dig into that deeper.

I continue to admire Garmin for their ongoing innovation, and their PNDs are rapidly increasing functionality as general purpose GPS devices. However my own PND is an off-brand, purchased for $100 at Fry’s, and it even includes spoken street names (aka TTS) and NAVTEQ maps! It’s a Windows CE based platform so I give up some UI elegance in exchange for the low price.

So, one key threat they face is the entry of low cost competitors, some of whom have adopted an open-platform strategy. Garmin remains completely proprietary.

VZ-Navigator has also shifted the landscape, followed by similar offerings from Sprint, Alltel and others. Credible industry analysts are predicting that personal navigation and associated capabilities will be largely taken over by cellphone applications in the near future.

So, a second key threat they face is the emergence of cellphone applications as substitute solutions that can be easily adopted by consumers since they don’t require the user to possess yet another piece of equipment; everybody already has a cellphone.

Virtually all of these GPS solutions depend on geo-content obtained from either NAVTEQ, the leader, or TeleAtlas. I think Tom Tom, the leading competitor to Garmin in the PND space, foresaw how things would unfold and acquired TeleAtlas for the purpose of reverse integrating into what might actually become their primary business. There are many more barriers to entry in the geo-content space than there are for PNDs.

And, of course, NAVTEQ has been acquired by Nokia, the world’s leading cellphone vendor. (The strategy behind that move might be the topic for a future posting.)

So, the third key threat Garmin faces is that the primary suppliers of the geo-content upon which they depend are now allied with competitors.

What should they do? They are at a fork in the road, consisting of the following choices.

First, they could deemphasize their own proprietary approach and embrace third party platforms, then seek to become the navigation software application of choice for every cellphone, Windows device, and OEM’d nav solution in vehicles. Admittedly, they are late to the game here, and their differentiation is not obvious, but it may be their best shot at maintaining a significant share of the navigation market.

Alternatively, they can continue to embrace their proprietary platform approach and crank up the innovation engine to leave competitors in the dust. They may be late to the game here too. For example, they should have been first to market with a “Connected GPS” device, but Dash Navigation and Insignia beat them to the punch. Garmin wasted precious R&D time on the misguided Nuviphone when they should have pursued cellular connectivity as an integrated function, a-la the Amazon Kindle, rather than the primary role of the device.  A Connected GPS device is a GPS device first, and the cellular data connection is complementary.

(By the way, what could Garmin have been thinking when they invested in developing a cellphone? Yes, Apple succeeded, but that’s a very tough business to make a go of.)

With cellular data a GPS device can add many new capabilities, including the obvious ones of real-time traffic reports and traffic based routing/rerouting. As another example, I would love to have a PND that overlaid graphical weather right onto the map display. Think of a mash-up between a Garmin PND and the MyWeather application available as a download on iPhones.

So, I recommend the second approach: A high margin business based on innovation on proprietary platforms, but without the dominant market share they’ve been used to.

From where I sit as an outsider I can’t say they are assured of success long term in the consumer space. They remain well-positioned in aviation and marine products, but the entry barriers are much higher in those markets, especially in aviation where FAA certification is an expensive process.

I’m watching to see what their next move will be. I’m betting that Garmin’s Connected GPS device will soon arrive, but it has already missed the Christmas season. They have some catching up to do!

Most readers here will know about the history of Nextel and the Motorola iDEN technology which served as the cornerstone for their brilliantly executed strategy to win and keep high margin enterprise customers. The Push-to-Talk feature was a key differentiator that other Mobile Network Operators could not match.

And, Nextel had a firm grip on this differentiator, with exclusivity arrangements that ensured they would not find themselves competing with other MNOs who were using the iDEN technology. Nextel even had the “foresight” to lock up the promising CDMA version of PTT, Qualcomm’s QChat technology, with a high dollar exclusivity arrangement, several years before it would come to market. (QChat is a set of proprietary extensions on top of EVDO Rev A.)

Then what happened next? I have a very simple view of the answer. Simply put, Nextel found that the sword of iDEN exclusivity was two-edged: the PTT advantage was nice but the cost of being virtually the only MNO supporting Motorola’s R&D on iDEN was more than they could bear. Nextel could not afford the cost of advancing iDEN to support 3G packet data capabilities nor could they afford the cost of a nation-wide buildout. (There were insufficient iDEN roaming partners to achieve good coverage.) Nextel had found themselves alone on an island, and in the world of wireless that spells certain doom.

And we all know what happened next. Sprint came to the rescue, acquiring Nextel, its iDEN network and those lucrative enterprise customers. And, a “brilliant” strategy fell into place: Since Sprint operated a CDMA based network with a clear path to 3G, the QChat technology that Nextel had previously licensed could be leveraged to provide a path off of the iDEN network onto CDMA. That PTT advantage on iDEN would carryover nicely to CDMA!

Now, this should be feeling like a Déjà vu moment. Here’s why.  The cost of bringing QChat to market was enormous. And who paid for it? Did Alcatel-Lucent, Nortel, Motorola, Huawei or ZTE speculate on that R&D expense, knowing that Sprint would be the only customer? I guarantee you that didn’t happen. Sprint covered the bulk of that R&D risk, subsidizing selected equipment vendors with NRE payments. So, QChat has become yet another two-edged sword. Can Sprint afford to be the only MNO that bears the cost of R&D on QChat? Oh, and by the way, QChat doesn’t roam.

To further disadvantage Sprint, Verizon Wireless used primarily non-proprietary capabilities in EVDO Rev A to deploy a much less expensive version of PTT, leveraging the earlier version of Motorola PTT that had flopped in Verizon. And the result is a darn good PTT solution. Not quite as good as iDEN, and probably not quite as good as QChat, but good enough (think Innovator’s Dilemma…).

So, Sprint is riding a proprietary horse, again, and Verizon Wireless chose a standards-based horse.

Verizon’s check-mate move came when they announced that they would adopt a non-CDMA path for their 4G technology. That leaves Sprint holding the R&D bag not just for QChat, but for all of CDMA if QChat is to survive.

 It is assured that there will be no version of QChat in the 4G networks. Will Sprint keep CDMA + QChat alive forever, or will they fall back to a standards-based PTT that operates on Wimax?

Sprint blew a lot of money on QChat, and ended up right back where they were with iDEN – all by themselves as the sole investor in a technology. Except this time the differential advantage of PTT has almost disappeared, leaving them with a one-edged sword – and it’s the wrong edge!

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.

Increased bits per hertz and bandwidth are very important, but there are other leaps forward in 3G that are equally important. Here’s a few of my favorites:

• Faster channel activation. Packet data connections are supposed to be “always on.” But in wireless it is not practical to keep the physical layer always on, due to the high cost of spectrum and also due to the battery consumption associated with operating the radio. In 3G Wireless a packet session can be activated from dormant to active in far less time than in 2G. The result is a much closer approximation to an “always on” user experience.
• Quality of Service. In the wired Internet the common use of over-provisioning has resulted in almost no need for Quality of Service (QOS) capabilities in the infrastructure. Characteristics such as latency, jitter and packet loss can all be addressed more cost effectively in the wired world by over-provisioning bandwidth (the wonders of fiber optics!). QOS controls become important when over-provisioning is not cost effective, and by using them applications that need good QOS, such VoIP, can efficiently coexist with applications that do not, such as email. 3G networks, beginning with EVDO Rev A and HSPA, have built-in the capability to allow applications to request and receive reservations for the QOS levels that are necessary for a good user experience.
• Broadcast/Multicast. Nextel’s iDEN network includes the ingenious capability to broadcast a single stream to multiple devices in the same cell-sector using a single channel. In their case it is used for the Push-to-Talk service, whereby many users can be participating in the same walkie-talkie conversation, and if several of those users are in the same sector they can all listen using the same channel. 3G Wireless will have a similar capability, although I am not aware that it has been deployed anywhere in the US as of this date. Some of my associates claim that it has been implemented in KDDI’s network in Japan, but I have not heard what types of applications have used it, if any. I’d love to hear any info others have to offer on this.

In my view 3G Wireless represents the transition between pure circuit voice and pure packet voice. 2G Wireless only supported full-duplex voice via circuit telephony. 4G Wireless will only support full-duplex voice via packet data (VoIP). 3G Wireless can support full-duplex voice using either circuits or packet data.

 So in the near term the bandwidth advantages of 3G certainly justify its adoption. But the other advancements that arrive with 3G will enable even greater progress in wireless applications. It’s not just about the speed!