FFT processing through faulty memories in OFDM based systems

In the widely used OFDM (Orthogonal Frequency Division Multiplexing) systems, the FFT and IFFT pair are integral parts used to modulate and demodulate the data constellation on the sub-carriers. Within such systems, embedded buffering memories occupy a large portion of the area and hence directly control the overall metrics of the system including power consumption and cost. This paper presents a unified statistical model that accurately reflects the impact of random embedded memory failures due to power management policies on the overall performance of an OFDM-based communication system. The proposed model expands the design space, by allowing the designer to replace the faulty hardware with perfect hardware while propagating the resulting distribution (due to voltage overscaling) through the system. The proposed framework enables system designers to efficiently and accurately determine the effectiveness of novel power management techniques and algorithms that are designed to manage both hardware failures and communication channel noise, without the added cost of lengthy system simulations that are inherently limited and suffer from lack of scalability.