Battery Life Estimation of Mobile Embedded Systems full report
project report tiger|
Active In SP
Joined: Feb 2010
01-03-2010, 10:52 PM
Since battery life directly impacts the extent and duration of mobility, one of the key considerations in the design of a mobile embedded system should be to maximize the energy delivered by the battery, and hence the battery lifetime. To facilitate exploration of alternative implementations for a mobile embedded system, in this paper we address the issue of developing a fast and accurate battery model, and providing a framework for battery life estimation of Hardware/Software (HW/SW) embedded systems. We introduce a stochastic model of a battery, which can simultaneously model two key phenomena affecting the battery life and the amount of energy that can be delivered by the battery: the Rate Capacity effect and the Recovery effect. We model the battery behavior mathematically in terms of parameters that can be related to physical characteristics of the electro-chemical cell. We show how this model can be used for battery life estimation of a HW/SW embedded system, by calculating battery discharge demand waveforms using a power co-estimation technique. Based on the discharge demand, the battery model estimates the battery lifetime as well as the delivered energy. Application of the battery life estimation methodology to three system implementations of an example TCP/IP network interface subsystem demonstrate that different system architectures can have significantly different delivered energy and battery lifetimes.
Debashis Panigrahi y, Carla Chiasserini z, Sujit Dey y, Ramesh Rao y,
Anand Raghunathan x and Kanishka Lahiri y
As the need for mobile computation and communication increases, there is a strong demand for design of Hardware/ Software (HW/SW) Embedded Systems for mobile applications. Maximizing the amount of energy that can be delivered by the battery, and hence the battery life, is one of the most important design considerations for a mobile embedded system, since it directly impacts the extent and duration of the systemâ„¢s mobility. To enable exploration of alternative implementations for a mobile system, it is critical to develop fast and accurate battery life estimation techniques for embedded systems. In this paper, we focus on developing such a battery model, and provide a framework for battery-life estimation of HW/SW embedded systems. Previous research on low power design techniques [1, 2], tries to minimize average power consumption either by reducing the average current drawn by a circuit keeping the supply voltage fixed or by scaling the supply voltage statically or dynamically. However, as shown in this paper, designing to minimize average power consumption does not necessarily lead to optimum battery lifetime. Additionally, the above techniques assume that the battery subsystem is an ideal source of energy which stores or delivers a fixed amount of energy at a constant output voltage. In reality, it may not be possible to extract the energy stored in the battery to the full extent as the energy delivered by a battery greatly depends on the current discharge profile. Hence, accurate battery models are needed to specifically target the battery life and the amount of energy that can be delivered by a battery in the design of a mobile system. The lifetime of a battery, and the energy delivered by a battery, for a given embedded system strongly depend on the current discharge profile. If a current of magnitude greater than the rated current of the battery is discharged, then the efficiency of the battery (ratio of the delivered energy and the energy stored in the battery) decreases, in other words, the battery lifetime decreases [3, 10]. This effect is termed as the Rate Capacity Effect. Additionally, if a battery is discharged for short time intervals followed by idle periods, significant improvements in the delivered energy seem possible [11, 13]. During the idle periods, also called Relaxation Times, the battery can partially recover the capacity lost in previous discharges. We call this effect as the Recovery Effect. An accurate battery model, representing fine-grained electro-chemical phenomenon of cell discharge using Partial Differential Equations (PDE), was presented in . However, it takes prohibitively long (days) to estimate the battery lifetime for a given discharge demand of a system. Hence, the PDE models cannot be used for design space exploration. Some SPICE level models of battery have been developed [6, 7], which are faster than the PDE model. However, the SPICE models can take into account the effect of Rate Capacity only. Based on the Rate Capacity effect, a systemlevel battery estimation methodology was proposed in [4, 5]. Recently, a Discrete-Time battery model was proposed for high-level power estimation . Though it is faster than the previous models, it does not consider the Recovery effect. In this paper, we describe a stochastic battery model, taking into account both the Recovery effect and the Rate Capacity effect. The proposed model is fast as it is based on stochastic simulation. Also, by incorporating both Recovery and Rate Capacity effects, it represents physical battery phenomena more accurately than the previous fast models. We also show how this model can be used for estimating the battery lifetime and the energy delivered by the battery for a HW/SW system, by calculating battery discharge demand waveforms using a power co-estimation technique . Based on the discharge demand, the battery model estimates the battery lifetime as well as the delivered energy. Finally, we demonstrate how this framework can be used for system level exploration using a TCP/IP network interface subsystem. The results indicate that the energy delivered by the battery and the lifetime of the battery can be significantly in- 1 creased through architectural explorations. The rest of the paper is organized as follows. Section 2 motivates the need for an accurate battery life estimation methodology by illustrating that the battery life and the energy delivered by the battery can be affected significantly by tradeoffs at the system level. Section 3 provides background on the physical phenomena inside a battery, which affect the the battery lifetime as well as the delivered energy. The proposed battery model is described in section 4. The methodology used to calculate current waveforms is described in section 5. In section 6, we demonstrate how the battery life estimation methodology can be used to evaluate alternate implementations in the design of Battery Efficient Systems. Section 7 concludes the paper and explores future research.