With more than 15 million Linux handsets having already shipped into the market, most in the last year, the promise of Linux for the mobile phone industry is becoming a reality. Linux handsets have primarily been marketed into Asia so far, with notably strong volumes in Japan. NTT DoCoMo, recognized as an innovator and mindshare leader in the mobile telecommunications space, expects to ship more than 10 million Linux handsets by the end of March 2006 into Japan alone. To put it in perspective, that number represents one Linux handset for every 13 people in Japan. But the growing success of Linux is not a region-specific phenomenon

Linux Handsets - Asia Only?
Asia - Japan in particular - has historically been an early adopter of the latest technologies, and this trend has carried over into the mobile phone market. It's only natural for new mobile trends and innovative technologies to show up in Asia before reaching other parts of the global handset market.

Case in point. Despite an initial focus on the Asian market, Motorola has already begun marketing its Linux handsets in Europe and recently announced that its ROKR E2 handset, unveiled at the Consumer Electronics Show in January, will use its new Linux platform. Although distribution hasn't been announced yet, the ROKR product line has historically been marketed in Europe and the United States. Given Motorola's recent lifestyle-focused ad campaigns and level of promotion around the line, expect to see the ROKR E2 hitting shelves soon - and outside of Asia.

So why is Linux becoming so popular in the mobile space? It's not because consumers are asking for it, that much is certain. Unlike technologies such as Bluetooth or Wi-Fi, you wouldn't expect to find a Linux checkbox on the product packaging. A device operating system isn't there to satisfy consumer preferences; it's an enabler for handset makers to improve functionality, time to market, and success.

The Growth and Transition of an Industry
To understand better why handset manufacturers are trending towards adding Linux handsets to their portfolios, it's important to understand the industry's origins. As with many industries, early players in the mobile market were highly vertically integrated, and early handset makers often had to build many if not all of the components that were used in their devices themselves because of complexity and lack of third-party suppliers. And many handset makers would also build the supporting infrastructure products needed for the carrier networks to deploy mobile services, in essence acting as end-to-end suppliers.

As handsets became more mainstream and the market flooded with competition, producers began to compete with each other on a feature basis and devices became increasingly complicated to build. Having focused primarily on voice for so many years, the software content exploded as device makers began adding graphics, ring tones, text messaging, browsing, cameras, Bluetooth, and most recently, multimedia capabilities (see Figure 1).

This explosion of technology was engineered piecemeal on top of basic software platforms. Each handset was fundamentally different, and handset manufacturers found themselves maintaining dozens of complex and tenuously constructed platforms. Spinning out new models with the latest technology was difficult, but the alternative was to get to market late with yesterday's technologies. The market rewarded companies that hit their schedules and penalized ones that were late. Handset manufacturers would often experience significant shifts in market position due in no small part to their ability to hit schedules.

Differences in Approach
A few top handset manufacturers develop most of the mobile devices sent to market. In fact, the top three handset makers - Nokia, Motorola, Samsung, respectively - account for nearly two-thirds of all handset shipments according to IDC. While there are additional top-tier handset makers with significant volumes, there are also dozens of smaller ones that have recently emerged; many of these falling into the category of the so-called "original design manufacturer" or ODM. Many of these mid-tier and smaller manufacturers lack the vertical integration of the top-tier manufacturers and have to look outside for many of the components necessary to build and integrate a mobile handset design. Initially, third-party offerings focused on more complicated value-added technologies; companies such as OpenWave and Teleca built entire businesses supplying handset makers with the application and middleware components they needed to compete in the mobile market.

New Models for Building Mobile Devices
As with many technology industries, the rate of innovation was rapid in the mobile space and the smaller players quickly caught up with the larger handset manufacturers in terms of pure functionality. Conflicting and varying requirements from competing mobile carriers in diverse geographies also drove increased fragmentation, leading to prohibitive levels of complexity. Building a single device or single platform to satisfy global requirements was a difficult, if not impossible, proposition.

To stay abreast of the new technologies hitting the market and deal with a fragmented customer base, top handset makers increasingly turned to outsourcing non-differentiating software components. The objective was to rely on healthy supplier ecosystems and increased bandwidth to differentiate themselves from their competition. Companies such as Motorola and Siemens even spun out semiconductor divisions (Freescale and Infineon, respectively), and many mobile device makers are currently moving from internal homegrown operating systems to external platforms. Despite this transition to external sourcing, mobile handset makers felt they needed to retain control over the content of their devices to ensure that they stayed competitive in terms of functionality and timing.

As handset makers ventured out to find off-the-shelf components, they discovered a wealth of options available - both from a platform and component perspective. Companies such as Symbian and Microsoft marketed solutions that included fairly complete software stacks (though these solutions still required significant effort to integrate, validate, and deploy).

Although the idea of a complete solution may appeal to some, ODMs in particular, most top-tier handset manufacturers were concerned that relying on these platforms would lead to commoditization. Those familiar with the personal and enterprise computing markets will remember the emergence of a powerful hardware/software duopoly in the late '80s and early '90 that drove the market towards commoditization. Other than a small amount of software and slight variations in the plastic that could surround the final product, very little differentiation was available to the once-dominant technology-driven computing suppliers. Arguably, this led vendors as large as IBM to exit the space. Mobile handsets remain highly complex devices, and consequently highly differentiated. Relying on proprietary commercial operating environments limits flexibility and control, and makes it difficult, if not impossible, for handset makers to address highly fragmented market requirements successfully while still delivering differentiated competitive devices to the market.

Independence, Differentiation, and Efficiency
Preferring to remain unfettered by restrictive proprietary software vendors, many larger handset makers have turned to Linux, though not for the reasons one might think.

Though many assume that the adoption of Linux was driven by a desire to reduce the bill-of-materials cost, as it turns out, most legacy operating systems used in the mobile market were already royalty free - so cost cutting wasn't the prime concern. Because Linux can't be controlled by any one vendor and because developers have access to all the source code, device makers retain full control over their platform architecture and development schedules. This approach also provides the manufacturers with greater control over the phone's presentation and functionality enabling the creation of truly differentiated products.

Unlike competing mobile operating systems, handset makers are leveraging Linux as a common platform across a broad percentage of their portfolio. Because it's easy to program and customize the Linux platform, handset manufacturers can quickly spin out new models with the latest technology and address the latest market trends. For example, Motorola has reportedly stated on multiple occasions that they expect Linux to be used across most of their portfolio, with other operating systems being used primarily in point-devices. NTT DoCoMo, along with suppliers NEC and Panasonic, have announced similar positions. Most recently, Panasonic said it intended to discontinue developing second-generation GSM (Global System for Mobile Phones) devices and focus exclusively on 3G (third-generation) Linux-based handsets.

Given highly fragmented market requirements, the ability to address a variety of customers by making small modifications or extensions to a common platform can reduce many of the challenges facing handset makers today.

Industry Support for Linux
While cost hasn't been a primary factor driving early adopters, various companies are working to cut costs and make it easier to deploy platforms with Linux. Over the past 12-18 months, various independent organizations such as the Open Source Development Labs (OSDL), the Consumer Electronics Linux Forum (CELF), and the Linux Phone Standards Forum (LiPS) have launched initiatives seeking to increase the adoption of mobile Linux globally. Individual companies have also launched initiatives to improve mobile Linux adoption. Through its Mobilinux Open Framework Program, MontaVista is working with leading third parties to port their offerings and develop pre-integrated mobile phone solutions (see Figure 2).

Some of these companies include Aplix, Bluefin Mobile, Certicom, Digital Airways, Esmertec, Jataayu, Red Bend, Sky Mobile Media, Teleca, Texas Instruments, and Trolltech.

By pre-integrating these modular technologies on top of Mobilinux, MontaVista and its partners are helping to address many of the challenges associated with the integration and interaction of voice and data software components, and making it easier for handset developers to design Linux-based mobile phones.

Fast Forward
Clearly a lot of expectations are riding on the success of the Open Source platform in the mobile space, and a large number of companies are working together to further its progress. With vendors continuing to improve the performance and capability of the operating system, and continuing to work with the rapidly growing mobile Linux ecosystem, building Linux phones is becoming easier all the time. In the next year, expect to see Linux-based handsets hitting shelves in North America, and establishing a more significant presence in Europe. But don't expect to see Linux on the product packaging; that's not the point.