I believe it’s very likely that Google will launch its own smartphone, probably in the first half of 2012, that’s not based on Android.

What is this smartphone NOT? You’re all aware of the Nexus One (launched in January 2010) and the Nexus S (launched in December 2010) smartphones, running the then-latest versions of the Android operating systems. These were developer-centric phones, simply lacking the heavy software customizations to Android usually practiced by HTC, Samsung, Motorola and others.

However, from a hardware perspective, both Nexus devices were designed largely by their makers — HTC and Samsung, respectively. The 2012 Google smartphone I am talking about is very different in at least two aspects:

1. Most importantly, this Google smartphone will not be based on Android. Rather, it would be based on a smartphone-optimized version of Chrome OS. There are several reasons for this.

A. Security — Many current operating systems, including Android and Windows 7, just for starters, have severe security issues. The main reason for this is a combination of two things: First, Android is a very open ecosystem in which applications can enter the device from a variety of angles, including those not vetted sufficiently or at all by a company such as Google itself. Second, the very nature of an operating system which allows applications to be installed is a security risk in and of itself, particularly if the device is not actively managed by a heavy-handed central police such as the famous BlackBerry Enterprise Server from Research In Motion.

A cloud-centric OS such as Chrome breaks this paradigm by not allowing locally installed apps in the traditional sense. This device would only have two major software parts — the OS and the only allowed browser. However, the OS treats the browser as a de-facto hostile application, not allowing it to modify the OS including locally install any applications.

This type of security paradigm would make RIM’s BES mostly obsolete. There would be no traditional need to monitor and restrict the end user device using such comprehensive and active tools. The BES could be replaced by a much simpler management console which would focus more on device access, activity in the browser, and overall account device management only. Google could easily design such a product, causing a lot of headache for RIM.

B. Cost — This type of a cloud-centric smartphone could be produced using less memory, less local storage and a less powerful CPU. It would not negate the need for constant improvement in the areas of display resolution, GPU horsehower and improved (faster) connectivity and input methods such as cameras.

C. Competition — Earlier this year, rumors published by several of the major tech bloggers suggested that Motorola Mobility is working on an unspecified cloud OS for its future smartphone and tablet needs. More recently, Samsung announced that it will introduce its “Cloud Phone” in 2012. It therefore seems clear that at least two of the leading Android licensees are seeking alternatives from the ever-faster spinning Android competitive treadmill.

We are seeing in the discovery documents covered in the Skyhook-Google lawsuit an increasing tension between Google and its licensees, specifically Motorola and Samsung. It looks like all parties have at least some dissatisfaction with what they perceive to be downsides of the very open Android model, which is still largely controlled by Google. Licensees feel it’s hard to differentiate, whereas Google doesn’t want them to differentiate the software, only the hardware.

So where is all of this going? We are increasingly moving into a set of vertical ecosystem silos. RIM, Apple and Hewlett-Packard  are already complete vertical silos in terms of their smartphone/tablet products. Microsoft has a very tightly controlled process with no ability for software differentiation, but is now in bed with what appears to be a highly favored (“more equal than others”) partner in Nokia, to the detriment of Samsung, LG, HTC and Dell. Many signs point to a wholesale merger between Microsoft and Nokia, which may cause the other four aforementioned licensees to drop Microsoft 7 Mobile like a bad habit.

That leaves us with Google. If not only RIM, Apple and HP — but also Microsoft — goes completely vertical with dictatorial end-to-end control, Google would be the odd man out with a chaotic ecosystem lacking in consistency and therefore user experience.

So what would Google do? I believe it is looking to a two-pronged mobile strategy.

1. Continue to Android largely as is but with whatever modifications it can get away with, in order to reduce (but not eliminate) fragmentation. Tools to do this include giving preferential treatment for new OS versions, which we saw most recently with 3.0 Honeycomb. In this case, you may have noted how all 3.0 Honeycomb products announced to date are stock Android, not the skins usually provided by HTC, Samsung and others. This will change soon, at least on the software side, but Google is at least attempting to steer the market for a few months in the direction of its “plain vanilla” offering — which is frankly what most consumers want anyway, at least so far.

2. Offer an all-new OS where it adopts — at a minimum — the Microsoft “total control but with some hardware makers” approach. Google would offer a very tightly controlled smartphone Chrome OS with hardware chassis specs to the usual suspects Samsung, HTC, LG, Motorola, Sony Ericsson, Dell and many others. The emerging problem here would be that at least a couple of these guys — Samsung and Motorola — are or appear to be on their way to their own vertical integration with their own cloud OS ecosystems.

What kind of approach would this be for Google in terms of dealing with the wireless carriers? I think this new Chrome OS smartphone would be treated as one of the current cellular (including LTE soon) tablets such as Apple’s iPad or the Motorola Xoom, just to mention two of the most prominent examples. The Google Smartphone would be sold largely unsubsidized with no contract necessary. One reason this may be feasible is that the cost would be lower as I discussed below. I don’t know if this means a $399 price point or whatever, but I think $299 could be within reach for at least a lower-end version.

It’s pretty clear that an initial version of the Google Smartphone would have to rely on the existing cellular data technologies employed around the world: HSPA, LTE and WiMax. Over time, Google would probably also pioneer the white spaces technology, but it’s unclear if that would make it to market by 2013 or rather by 2014 or 2015. My guess is later rather than sooner for the white spaces networks. For the U.S. market in the near term (2012-2013) it seems clear that LTE would make for a best-selling option given the strong network performance at 700 MHz on both the Verizon and AT&T operators.

What about voice? In this data-only Chrome OS-based Google Smartphone, voice would be pure VoIP using the kind of desktop-lookalike client of Google Voice in combination with Google Talk. This kind of offering from Google is way overdue anyway, it already being over 18 months since it acquired Gizmo5, so one would surely expect it in the Google phone. Carriers would just have to swallow the fact that they are being “reduced” to being a pipe provider — but that’s where they already are on the iPad2 and the Motorola Xoom, for example. As I pointed out in essays published already in 2003, however, there is nothing wrong or demeaning in being a pipe provider. Data is very important and the fundamental way to differentiate a carrier.

What is the remaining dilemma for Google? I think Google has the capacity to entirely spec its own hardware, and therefore simply branding it something like “Google Smartphone” or equivalent (surely they can come up with something better, but you get the point). This maximizes the probability of keeping the project secret until the very end, and could perhaps enhance Google’s ability to maximize the integrated user experience. That said, if Google feels that it could control the user experience built on a tightly controlled hardware chassis spec just like Microsoft Windows 7 Phone, it could at least try to see if a company such as HTC or Dell would be willing to bend to its total wishes and help it along with a branded hardware design.

Several powerful and many recent trends appear to point to the idea that Google will create its own alternative to Android, and that this alternative will be very tightly controlled — at least as controlled as Windows 7 Phone, possibly as tightly controlled as iOS, BlackBerry or HP’s WebOS. It would be able to make more secure devices, built on a pure IP data platform with only VoIP, and counter new competitive moves from Samsung and possibly Motorola Mobility, if they go their own independent cloud OS smartphone ways.

I call this strategy for Google “two torpedoes in the water” — one being Android and the other being Chrome OS.

Timing? This project likely already started several months ago, and given the simplified lack of carrier certification (data-only and over-the-top VoIP tablet mode instead of circuit-switched phone mode), a 2012 launch looks very much realistic.