Computing Battery Lifetime Distributions
DSN '07 Proceedings of the 37th Annual IEEE/IFIP International Conference on Dependable Systems and Networks
xTune: A formal methodology for cross-layer tuning of mobile embedded systems
ACM Transactions on Embedded Computing Systems (TECS)
| Hi-index | 0.00 |
In the design process of an (embedded) computer system there are several important attributes the developer has to take care of: first of all, the final product should do the right thing, we then speak of functional correctness. Second, the performance should be adequate, expressed in measures such as throughput, delay or loss probability. Third, when relying on a battery as power source, it becomes increasingly important that the system behaves in an energy-aware manner. We could assess any of the three attributes in isolation, using completely different sets of models and tools. However, since the alteration of one of the attributes most surely also affects the other two, an integrated framework where all aspects can be evaluated and balanced is definitely desirable. We present such an integrated approach, but focus on the evaluation of battery lifetime. The system under consideration is represented by a stochastic workload model which then is combined with a battery model. In doing so, several design alternatives in the behaviour of the system can be compared early in the design process and the optimum with respect to functionality, performance and energy-consumption can be chosen.