基于л型RLC模型估计电源树同步切换噪声

Switching noise estimation on P/G networks based on di stributed RLC model

深亚微米工艺下,高密度器件的高频同步切换噪声可严重影响超大规模集成电路(verylargescaleintegratedcircuit,VLSI)的可靠性和信号完整性.基于一类VLSI电源树(powersupplytree,PST)的分段π型RLC模型,利用切换事件驱动机制、节点和参数重组建立切换电流规整传播路径,并以二极点模型近似切换电流传递过程,实现电源树同步切换噪声(simultaneousswitchingnoise,SSN)估计.仿真表明,该方法可加速同步切换噪声模拟过程,提高效率,并保持较高模拟精度.

Under very deep sub micron (VDSM) technology, highe r device densities and faster switching speed of VLSIs will cause large switching current that degrade the performance, reliability and signal integrity. Based on л-typed RLC models of power supply lines, a simultaneous switching noise (SSN) estimation approach was presented. To simplify,the event-driven mechanism, nod e and parameter regrouping was used to set up the regular path that transmitted switching currents to the destination. Simulation results showed that our method could accelerate the SSN estimation process, improve the efficiency obviously, and have good accuracy.