If 2.5 and 3G data services take off there will be literally billions of pieces of data flying around the airwaves, but where will they be stored? Catherine Haslam explains the options.
According to the ARC group, in excess of 20% of mobile data users will be using MMS by 2007. And after 2007, ARC’s research suggests MMS usage will grow exponentially with 50% of all usage being content-to-person, although early growth will be driven by person-to-person messaging. This equates to just over 2 billion messages a month and 25 billion a year, a considerable percentage of which will need to be stored somewhere. Even if the take-up is slower — 200 million messages per month in 2004 — as suggested by World Wireless Forum, the need for users to have a mechanism to save and store messages is clear. Add to this demand the requirement to store other information such as games and MP3, not to mention email, and the need to sort out storage for wireless generated and terminated content becomes pressing.
There are three main options to storing content in the wireless world — embedded storage in the handset; external memory card and servers in the network — and a case can be made for all.
Embedded storage
Mobile handsets have traditionally been stand-alone units and as such there is a user and operator expectation that the handset will be able to support all the functions of the terminal. Users may be willing to accept that text messages are transitory, but the personal and evocative nature of picture messages will inevitably lead to a demand to be able to store them somewhere.
As has happened in the PC market, memory capabilities for mobiles are improving. However, the memory capacity of any terminal is the result of a delicate balancing act based on such things as capacity, power consumption, size and cost. Up until recently, the principle requisite was for the phone to operate effectively and provide comparatively little in the way of data storage.
According to Mary Curtis, 2G/2.5G international product marketing manager baseband at Philips semiconductors, 4MB of S-RAM combined with up to 16MB of Nor Flash used primarily to run communications aspects, are supported in the volume phones of today. In contrast, newer developments, SDRAM which offers a much higher RAM capability in excess of 256MB, and Nand Flash which was developed specifically for data storage and is much more effective for larger sizes of Flash memory, have yet to be seen being contemplated for anything but the high-end phones. Indeed, although high-end products including these are expected by the end of the year according to Curtis, there are few signs of these capabilities being included in mass market products for some time.
Balancing act
The reason that such functionality remains the domain of the high-end feature phones is basically down to price pressure. However, behind this is a complicated balancing act which means that if other savings can be made which reduce the demands of other essential elements of a chipset, more can be invested in memory; more in terms of power, space and money. As Curtis explains, “Everything comes down to system design.” (The trends of the chipset market will be examined in detail in the May edition of Mobile Europe).
In addition, there is the question of what happens to the data if something happens to the phone — it breaks, is lost or stolen? The personal and therefore valuable nature of the data stored on the device means its loss will create animosity that will directed at the handset manufacturer, the operator, or both.
Another option is therefore to store data on an external resource as is used for other electronic goods such as digital cameras. Indeed, a number of existing and soon-to-be-released mobile terminals already have external memory slots. This option overcomes several problems of the embedded memory. Firstly, the resources required to support the slot are fixed, while the memory capability is limited only by the user. This leads us to the second benefit (for the handset manufacturer and operator at least), the cost falls on the user. The handset manufacturer and/or operator is then also provided with the flexibility to offer additional functionality which will only work when additional memory is added through the expansion slot.
Obviously, such an approach has appeal but it does have a down side. Firstly, it puts a cost barrier in the way of certain applications and/or storage and secondly, it could tie users to specific handset models as the memory card market is proprietary. Currently, there are three main players — Multimedia Cards (MMCs); SD Cards and Sony Memory Sticks.
MMCs can be used in: Nokia’s 9110, 9210, 3300 and 3560; Siemens’ SL442, SL45 and CL35; Palm’s M130 and Handspring’s Treo 90. SD Cards are usable in the Nokia 9210i, Orange Microsoft SPV, Palm M130, Treo 90 and Siemens s55, while, not surprisingly, the Sony Memory Stick Duo is the sole preserve of Sony Ericsson’s P600 and P800 terminals, although it is transferable to any Sony product. Although not equivalent to the cost of upgrading a handset, the cost is not insignificant if you believe that the target is teenagers. The MMC and SD cards weigh in at around UKP25 (â‚-36.50) for 64MB, while the Sony Ericsson Memory Stick is virtually double the price for the same storage.
Even if the price and multiple standards are not enough to stifle usage, the question remains as to whether even these external memory extensions provide enough storage. Obviously, data can be transferred to a PC (or Mac) and that will be useful as a back-up but if, as is expected, much of the data is to be used with the mobile device, it will need to be kept in a way hat it can be used when mobile. If you consider that a 64KB card stores around one hour of MP3, then it will get pretty expensive to store music downloads in this way, although it may be more suitable for games.
The final option is to store data on the network and provide users with access to it, preferably via fixed and mobile devices. Vodafone is set to provide its Live! subscribers with space free in the form of a message library and others will follow this model. It has the advantage for the subscriber of minimising the hardware costs, while for the operator, it is hoped it will drive traffic as data is accessed and transferred. The downside is that it requires an investment from the already over-stretched resources of mobile operators. However, Mirapoint for one, is offering storage as part of its overall MMS solution.
The Mirapoint storage platform is MMSC agnostic and comes complete with an application which triggers an SMS that contains a URL to identify where the recipient can locate the MMS that was intended for them, providing some form of legacy support. It is also possible that such functionality will encourage the purchase of MMS-capable handsets as those without such handsets will get a taste of the application.
According to Mirapoint’s EMEA sales director, Steve Ashmore, “MMS storage is just the beginning as operators begin to offer storage services for all forms of messaging — SMS, EMS, MMS, voicemail and e-mail.”
Free from ties
Mirapoint is not the only provider of storage systems, but, according to Ashmore it has two USPs which set it apart from the competition. Firstly, its storage system is not tied to a particular MMSC. He explains, “Openwave ties MMSC with storage and while this is open to some extent, they interface together better than with third party options.” Secondly, Mirapoint has adopted what Ashmore describes as the “Cisco IP routing model.” This means that those elements of the system which are mature have been built in hardware rather than software. This both increases processing capabilities and also massively reduces the levels of integration required. “Anything that is part of the hardware is obviously pre-integrated and therefore we offer an out-of-the-box solution.”
Intuwave has also included a data storage offering as part of a larger middleware platform and its co-founder, Jeremy Burton believes that the company’s msafetynet service, which represents over-the-air back-up of data provides an “attractive value proposition for the user and for the operator which can use it as a mechanism to reduce churn.”
Ultimately, there is no simple or single answer to the storage question. The need will inevitably be met by a combination of the available memory resources. Burton believes that the applications will dictate the storage mechanism used. He says, “For extremely private data that the user wants to keep close to them, the terminal is the obvious choice but for volatile information that is constantly changing such as stock prices, it doesn’t make sense to cache or store that locally.”
For Curtis this then correlates to the well-defined segments of the handset market, “The amount of data storage is based on the target segment of the handset. At the high end, performance is everything and therefore storage capabilities will be included sooner rather than later, at the low end cost is key,” she says. For this lower cost but mass market, at least, it would seem that operators face little option but to offer storage facilities as a service. The only question is whether or not to charge for it as an entity in itself?