IMEC developed a design approach for run-time cross-layer control algorithms that enables energy-efficient operation of reconfigurable wireless communication systems. The controller approach has been validated on IMEC’s software-defined radio platform. A reduced energy consumption of up to a factor 3 is achieved by exploiting the available scalability in the baseband and the front-end. The approach is compliant with current and future wireless communication standards.
Energy consumption is one of the major challenges in designing reconfigurable radio platforms. To bridge the energy gap of software defined radios compared to hardwired radios while still providing the required quality of service, IMEC’s cross-layer controller exploits the scalability inherently available in the reconfigurable radio terminal.
The cross-layer performance/power optimization consists of design-time definition and exploration and run-time controlling. The controller follows at run-time the dynamics in the application requirements and the propagation conditions to achieve low power operation. Based on the run-time dynamics, the terminal is reconfigured at all levels including front-end and analog-to-digital converters, baseband and MAC processing (algorithms and platform), and video codec (algorithms and platform). In this way energy can be saved with respect to conventional controllers. The amount of scalability determines the achievable energy gain. This allows also legacy systems with more limited scalability to gain from this method.
Application in real-life cases has been used to validate the approach. For data transmission over an 802.11a wireless LAN link using IMEC’s proprietary software-defined radio digital baseband and front-end, the cross-layer optimization resulted in a reduced energy of up to a factor of three.
The results were achieved by exploiting the output power versus linearity versus DC power consumption trade-off of the power amplifier and traditional baseband parameters such as constellation and code rate. Overhead on the terminal is limited to a 2 Kbytes look-up table for storing design-time computed parameter settings and to a load of below 1% on the 200MHz ARM926 processor for executing the run-time optimization algorithm. Also design time effort for developing the controller is limited: the controller is automatically generated from behavioral models of the radio components.
The approach was also successfully applied to a WiMAX (802.16e) scenario showing similar gains. The design approach for the power/performance optimization controller is now ready for transfer to the industry for use in upcoming mobile communication systems.
More information on the cross-layer approach
The cross-layer performance/power optimization approach consists of design-time definition and exploration and run-time controlling. In a first step, the run-time controllable parameters are defined that significantly impact the performance and energy consumption at system level. Next, the relevant external variables representing the system dynamics are defined that can be tracked at run-time. The system energy and performance behavior is then modeled in function of the available parameters and external environment variables.
The complete system optimization is partitioned in local optimization of the different sub-blocks. Based on this design-time exploration and pruning, a look-up table is created that connects the energy quality behavior of the different system blocks with their configuration. Based on this system characterization, the policy is derived that will be used at run-time to determine the optimal configuration for the whole system.
Finally, the run-time cross-layer controller tracks the external variables, selects the current scenario and runs the optimization policy to select the optimal configuration points of each layer. In this way, energy consumption is minimized while the required performance is provided.
IMEC is a world-leading independent research center in nanoelectronics and nanotechnology. Its research focuses on the next generations of chips and systems, and on the enabling technologies for ambient intelligence. IMEC’s research bridges the gap between fundamental research at universities and technology development in industry. Its unique balance of processing and system know-how, intellectual property portfolio, state-of-the-art infrastructure and its strong network of companies, universities and research institutes worldwide position IMEC as a key partner for shaping technologies for future systems. IMEC is headquartered in Leuven, Belgium, has a sister company in the Netherlands, IMEC Nederland, concentrating on wireless autonomous transducer solutions, and has representatives in the US, China and Japan. Its staff of more than 1500 people includes more than 500 industrial residents and guest researchers. In 2006, its estimated revenue (P&L) was EUR 227 million.