偷得浮生半桶水(半日闲), 好记性不如抄下来(烂笔头). 信息爆炸的时代, 学习是一项持续的工作.
全部博文(1750)
分类: LINUX
2009-12-02 15:17:47
大多数的智能机(和许多不那么智能的手机机)一般含有两个处理器。操作系统、用户界面和应用程序都在Application Processor(AP)上执行,比如ARM芯片的CPU。而手机射频通讯和控制软件,则运行在另一个独立的CPU上,这个CPU称为Baseband Processor(BP)。AP与BP通过定义好的控制链路通信,比如串行控制线,或者AP发出的GPIO。
把射频功能放在BP上执行的主要原因是:射频控制函数(信号调制、编码、射频位移等)都是高度时间相关的。在主CPU运行这些函数的唯一可行方法是只有运行实时操作系。
另外一个使用BP的好处是一旦它被设计和认证为好了的,不管你采用的操作系统和应用软件怎么变化,它都可以正确的执行功能(它的通讯功能)。另外,操作系统和驱动的bug也不会导致设备发送灾难性的数据到移动网络中。同时,FCC要求在设备上跑的,与蜂窝网通讯的整个软件栈需要认证。由于AP和BP是分开的设备,手机设计者可以更加自由的设计用户界面和应用软件。
Most smart phones (and many not-so-smart phones) will contain two processors. The OS, user interface, and applications will run on the main applications processor (AP), such as an ARM-based CPU. The actual phone radio communications and control software, however, are seperated out and run on a seperate CPU, called the Baseband Processor (BP). The AP communicates to the BP over a defined control link, such as a serial connection or via GPIO lines coming from the AP.
The main reason for seperating out the radio functionality into the BP is because radio control functions (signal modulation, encoding, radio frequency shifting, etc.) are highly timing dependant. The only practical way to run these functions on the main CPU is if you are running a realtime OS.
The other benefit of utilizing a BP is that once it is designed and certified, you know that it is going to function properly regardless of application and OS changes. Also, application and driver bugs won't be able to cause malfuncitons in the phone radio, which would otherwise have disasterous consequences to the mobile phone network. Also, the FCC requires that the entire software stack running on a device which communicates with the cellular network be certified.
Since the AP is considered a seperate device from the BP (which creates the radio signals), phone designers are more free to do as they please with the user interface and applications.