Transition Activity Estimation for Digital Signal Processing Systems
Authors: Calomarde, Antonio; Rubio, Antonio; Saludes, Jordi
Source: Journal of Low Power Electronics, Volume 1, Number 3, December 2005 , pp. 217-225(9)
Publisher: American Scientific Publishers
Key:
- Free Content
- New Content
- Subscribed Content
- Free Trial Content
- Free Content
- New Content
- Subscribed Content
- Free Trial Content
Abstract:
Accurate high-level power estimation is required for complex systems design in a low time-to-market scenario. In this paper, a new transition activity technology-independent efficient technique to evaluate power consumption of DSP systems is presented. Temporal correlation and transition activity of signals for such systems are derived. Based on the spectral distribution of signals our proposal has the advantage of using data directly manipulated by the designer, simple and easy to apply even for complex circuits. With only the knowledge of the input signal spectrum and the transfer function from the input to the circuit internal nodes an efficient estimation of the temporal correlation, covariance, and variance are given. A model for computing transition activity in terms of the previous variables is developed, allowing getting the transition activities from the signal spectrum. The spectral analysis gives new criteria to obtain power optimization. We apply the estimation methodology to two typical examples of Digital Signal Processing Systems, IIR filters, and Sigma Delta Modulators, validating the method at RTL level of abstraction.Keywords: LOW POWER DESIGN; POWER ESTIMATION; SIGNAL CORRELATION; TRANSITION ACTIVITY
Document Type: Research article
DOI: 10.1166/jolpe.2005.043
Key:
- Free Content
- New Content
- Subscribed Content
- Free Trial Content
- Free Content
- New Content
- Subscribed Content
- Free Trial Content
Click here for Page Help