Energy-Efficient Hand-Held Multimedia Systems - Designing the Swiss Army Knife of Computing -

Abstract

The trend in wireless terminals has been to shrink a general-purpose desktop PC into a package that can be conveniently carried. Even PDAs have not ventured far from the general-purpose model, neither architectural nor in terms of usage model. Both the notebook and the personal computer generally use the same standard PC operating system such as Windows (CE) or Unix, same applications, use the same communication protocols and use the same hardware architecture. The only difference is that portable computers are smaller, have a battery, a wireless interface, and often use low power components [2].

Even though battery technology is improving continuously and processors and displays are rapidly improving in terms of power consumption; battery life and battery weight are issues that will have a marked influence on how hand-held computers can be used. Energy consumption is becoming the limiting factor in the amount of functionality that can be placed in these devices. More extensive and continuous use of network services will only aggravate this problem since communication consumes relatively much energy.

Another key challenge of mobile computing is that many attributes of the environment vary dynamically. Mobile devices face many different types of variability in their environment [3]. Therefore, they need to be able to operate in environments that can change drastically in short term as well as long term in available resources and available services. Merely algorithmic adaptations are not sufficient, but rather an entirely new set of protocols and/or algorithms may be required. For example, mobile users may encounter a complete different wireless communication infrastructure when walking from their office to the street [4]. A possible solution is to have a mobile device with a reconfigurable architecture so that it can adapt its operation to the current environment and operating condition. Adaptability and programmability should be major requirements in the design of the architecture of a mobile computer.

We are entering an era in which each microchip will have billions of transistors. One way to use this opportunity would be to continue advancing our chip architectures and technologies as just more of the same: building microprocessors that are simply complicated versions of the kind built today. However, simply shrinking the data processing terminal and radio modem, attaching them via a bus, and packaging them together does not alleviate the architectural bottlenecks. The real design challenge is to engineer an integrated mobile system where data processing and communication share equal importance and are designed with each other in mind. Just integrating current PC or PDA architecture with a communication subsystem, is not the solution. One of the main drawbacks of merely packaging the two is that the energy-inefficient general-purpose CPU, with its heavyweight operating system and shared bus, becomes not only the center of control, but also the center of data flow in the system and a main cause of energy consumption [1].

Clearly, there is a need to revise the system architecture of a portable computer if we want to have a machine that can be used conveniently in a wireless environment. A system level integration of the mobile’s architecture, operating system, and applications is required. The system should provide a solution with a proper energy-efficient balance between flexibility and efficiency through the use of a hybrid mix of general-purpose and the application-specific approaches. The key to energy efficiency in future mobile systems will be designing higher layers of the mobile system, their system architecture, their functionality, their operating system, and indeed the entire network, with energy efficiency in mind. Furthermore, because the applications have direct knowledge of how the user is using the system, this knowledge must be penetrated into the power management of the system.