The Symbian OS (779886), страница 78
Текст из файла (страница 78)
The radio hardwareis likely to vary between phone makers, from custom silicon to standardbought-in parts, and the data channel might be old-fashioned serial or,more recently, USB.Non-voice data on 2G and 2.5G and all data on 3G is packetized(into TCP/IP packets). The DSP/DAC steps are omitted but otherwise theTHE PHONE OPERATING SYSTEM371packets follow the same path through the signaling stack and are tunneledas GSM/2.5G/3G frames.The basic hardware design choices for a Symbian phone revolvearound whether to use one processor or two and, if two, how to connectthem.
The two-CPU option treats the baseband, or phone side, as a completely separate hardware subsystem and interfaces it to an applicationsubsystem with its own application processor (an ARM or StrongARMCPU, for example), running Symbian OS and applications. The singleCPU option creates a single hardware system, and shares it betweenan RTOS and Symbian OS or (enabled by the Symbian real-time EKA2kernel) uses Symbian OS exclusively to host both the baseband and theapplication stack on a single CPU.While to some extent how licensees architect their phones aroundSymbian OS and the design choices they make are opaque (and oftenjealously guarded), the consequences clearly impact the operating systemdesign and the assumptions it makes about the environment in which itruns.The hardware design options are as shown in Figure 15.1.A: Two processors connected by a fast serial bus: CPU A is the basebandprocessor and hosts an RTOS, which in turn hosts the baseband stack.CPU B is the application processor and hosts Symbian OS, on topof which is layered a bespoke or licensed user interface that hostsapplications.B: A custom package with two CPUs and shared memory at the registerlevel.
CPU A is the baseband processor and hosts an RTOS, whichin turn hosts the baseband stack. CPU B is the application processorand hosts Symbian OS, on top of which is layered a bespoke orlicensed user interface that hosts applications.C: A single processor hosts both the RTOS and Symbian OS. RTOS runsthe baseband stack and Symbian OS runs the user-side processes;the two operating systems have a mutual agreement to share theCPU, RAM, device drivers, and other system resources.D: A single processor hosts Symbian OS with real-time kernel EKA2.Symbian OS, abstracted by a custom ‘personality layer’, runs thebaseband stack, device drivers and user-side application processes.But telephony is not just a matter of getting raw data to the baseband.The baseband needs to be under application control, which means thatthere must be application interfaces to the phone side from the applicationside. On a typical Symbian phone, the phone application is simple andcan therefore be relatively hard-wired to the phone side, but phone372JUST ADD PHONEBasebandStackAppsAppsGUIGUISymbianOSRTOSBasebandStackSymbianOSRTOSCPUCPUCPUCPUHardware PlatformHardware PlatformA: EKA1-based system – 2 CPUs,bus connectedB: EKA1-based system – 2 CPUs,shared memoryAppsAppsGUIGUIBasebandStackBasebandStackSymbianOSSymbianOSRTOSCPUCPUHardware PlatformHardware PlatformC: EKA1-based system – 1 CPU, sharedby Symbian OS and Partner OSD: EKA2-based system – 1 CPU,Baseband Stack directly hosted bySymbian OSFigure 15.1 Hardware design optionsbooks, call logs, messaging (SMS, MMS, email, and so on) require accessto phone protocols and data services (such as networking and Webbrowsing) need to control the phone side in a modem-like fashion.While the telephony application that the user sees is relatively simple,the underlying engine which sits beneath it is quite complex.
It needs tohandle a number of cases such as ensuring that emergency calls are alwayspossible, even in low-memory conditions, and it must interoperate withhardware accessories such as headsets, as well as specific call-handlingand over-the-air (OTA) settings protocols (e.g., call-handling and SIMtoolkit functions).THE PHONE OPERATING SYSTEM373Supporting New HardwareThe real-time kernel (EKA2) is still valuable even in designs that retain aseparate, dedicated RTOS or partner operating system. Whether or not ithosts the baseband, EKA2 has advantages on the application side too.Ian Hutton:Hosting the telephony stack directly on Symbian OS requires real-time capability – and this is an issue. But the other argument for EKA2 is that it willallow mobile-phone manufacturers to integrate more multimedia hardware.The increased complexity of the hardware puts demands on the operatingsystem that are increasingly hard to sustain, just with the level of interrupts andso on, and that’s where EKA2 makes the difference.
So enabling the integrationof more and more hardware without compromising performance, which iswhat we are witnessing with phones, is the real bonus.The other aspect of new hardware – cameras, audio-codecs, highresolution displays and multiple displays, multiple radio interfaces andnew memory formats such as NAND flash, to give the most obviousexamples – is that it needs drivers and, increasingly (and especially formore exotic hardware), the drivers are likely to be proprietary ratherthan supplied by the operating system.
Easy integration of third-party andpartner drivers becomes a significant matter. EKA2 has been designedwith these needs in mind.Supporting ServicesNetworks are driven by services and services are supported by phones.As voice services become increasingly commoditized, existing nonvoice services such as messaging, browsing and new services (alternativenetwork access through Wi-Fi, broadcast TV, presence and navigation)become more important.Supporting FeaturesThere is no good definition of what features make a smartphone into asmartphone, a mid-range phone into a mid-range phone or a low-endphone into a low-end phone.4 Typically the measure is Bill of Materials(BOM) cost; as manufacturing techniques improve and Moore’s law54In [Lindholm et al.
2003, p. 172], ‘smartphone’ is still an emerging category definedby its balance of phone, personal productivity, imaging and gaming features.5Moore’s Law, derived from an article by Gordon Moore of Intel, originally publishedin Electronics Magazine in 1965, is popularly formulated as predicting a doubling ofcomputing power every 18 months.374JUST ADD PHONEHighFeaturesBill ofMaterialsIn 2007, best guesses are thatentry-level phones will target BoMcosts of less than $20, while highend phones based on Symbian OSand competitors (Linux, Microsoft)will target costs of less than $100.Feature phones, the so-called ‘midtier’, are likely to be aiming at lessthan $50.LowTimeFigure 15.2 BOM costs fall but feature pressure risescontinues to hold, BOM costs fall (See Figure 15.2).
As volumes continueto push inexorably upwards, driving marginal costs down, so the cost ofa given feature set drops inexorably.Meanwhile feature pressure (the demand to pack more and morefunctionality into phones) exerts a degree of counter pressure, drivingROM/RAM peripheral hardware requirements up. (For example, thecamera-phone has evolved into the multi-camera-phone; still pictureshave become video sequences; the ringtone phone has become a musicplaying phone and has subsequently evolved into a direct competitor toMP3 players.) It’s not clear whether that means that the BOM costs arenot falling as fast as they might and therefore the line between high-endand mid-range is not falling as quickly as it might or whether it just meansthat we are all migrating to the high end.Supporting the User ModelPhones are branded goods, fashion items, consumer appliances (see[Lindholm et al.
2003, Introduction]), and a host of other unlikely thingsthat drive user expectations for how they behave, how easy they are touse and what they do. Phones have hard keys and soft keys; some havekeyboards, some have pens, some accept voice commands; all may raiseissues about handedness and screen orientation.
Users expect ‘natural’interaction models, associate interaction styles with brands, and fuel bothperformance pressure and feature pressure (where the performance andfeatures of mid-range phones are increased to match high-end phones).Supporting the MarketPhone manufacturers typically want rapid product cycles as part of theirdrive towards increased volume of sales. Feature pressure (again with ahelping hand from Moore’s law and the economics of volume) drivesTHE PHONE OPERATING SYSTEM375a rapid technology-proliferation cycle. The pace of the drive towardsincreased volume of sales continually quickens and the cycle times fortechnology moving from research labs to products speed up.The result is a demand for continuously greater agility from suppliers,including the operating system supplier, as well as continuously greaterpredictability, fuelled by the relatively long product lead time coupledwith a short product cycle and lifetimes.There is additional pressure on a system when it becomes a platform(being a supplier is easier: there is only one customer to please).1997: The State of the ArtIt is worth recalling what a mobile phone was in 1997, the year the PsionSeries 5 was launched.Mobile phones had been mass-market products since perhaps 1994–5,depending on which geographical area you look at, with the UK’spenetration of a little over 20% about average for Europe, excludingScandinavia (which was more than twice that).
Penetration in the USAwas a little ahead of the UK, Canada a little behind, and in the rest ofthe Americas almost non-existent, as it was in China [Haikio 2002, pp.157–9].Worldwide, mobile phone sales were just under 108 million units (in2005, just under 800 million units were sold).6 Motorola was dominantwith 23% market share, Nokia was a little behind with 19%, and Ericssonwas a little more behind with just under 15%.
Nokia, indeed, seemed tohave stumbled, issuing profit warnings in both 1995 and 1996. Otherwise,Vodafone in the UK had a little over two million customers (15.5 millionin 2005) and had just introduced a pay-as-you-go service. The WAPForum had just been created, with the first WAP phones two years or soaway.Motorola’s phone of the year was the SlimLite (see Figure 15.3a),with a 4×16 character monochrome display and a 100-entry phonebookmemory. Among Nokia’s hot phones was the 3110 (see Figure 15.3b),in which a new, easy-to-use ‘one key’ (the Navi-key) user interfacedebuted.
The design context was still dominated by users’ propensity touse anything that looked like a dedicated ‘Call’ key to try to get a dialingtone before keying in a number. One of Navi-key’s goals was to help6The statistics and product specifications in this and the following paragraphs comefrom public sources. Some useful URLs include- www.gartner.com/press releases/asset 132473 11.html- www.gartner.com/5 about/press room/pr19990208a.html- http://en.wikipedia.org/wiki/List of mobile network operators- www.gsmarena.com/motorola slimlite-78.php- http://en.wikipedia.org/Nokia 7610- www.paconsulting.com/news/by pa/1997/by pa 19970115.htm376JUST ADD PHONE(a)(b)Figure 15.3 a.
![](https://s.studizba.com/z.php?f=/templates/si/i/noavatar.png&w=100&h=100&t=1"){kind=avatar}
