Random access detection fft/ifft processing method and device

Abstract

The present invention relates to a random access detection FFT / IFFT processing method and device, the method comprising: receiving an input data signal; switching between a burst mode and a pipeline mode according to the type of the input data signal, and completing the FFT of the input data signal processing or IFFT processing. The present invention aims at the characteristics of small FFT operation amount and large IFFT operation amount during random access processing in the LTE system, and adopts a set of devices to realize the parallel processing of two sequences by means of burst processing during FFT operation, Perform block floating-point operations on the sequence; when processing a large number of IFFT sequences later, use the high-throughput fixed-shift pipeline IFFT processing method to serially process one line of data, using a set of circuit resources to the greatest extent to improve the processing accuracy and balance the FFT The asymmetry of processing traffic between the IFFT and IFFT operations saves resource overhead during random access detection.

Description

technical field [0001] The present invention relates to the communication field, in particular to a random access detection FFT / IFFT processing method and device. Background technique [0002] In the LTE (Long Term Evolution, long-term evolution) system, the random access function needs to complete the process of performing circular convolution correlation between the random access sequence of the receiving antenna and the local root sequence. Due to the complexity of time-domain circular convolution, according to its nature, the convolution sequence can be transformed into a frequency-domain sequence by FFT (FastFourierTransform, Fast Fourier Transform), and the FFT transformation value of the received random access sequence can be multiplied by the local root sequence operation, and then perform IFFT (InverseFastFourierTransform, Inverse Fast Fourier Transform) on the multiplied sequence to obtain a time domain convolution correlation sequence. [0003] For FFT, it is don...

AI Technical Summary

Problems solved by technology

However, due to the characteristics of the implementation of the FFT device, the shift detection of the block floating point cannot be performed during the processing.

Therefore, only a fixed shift can be used for each level, and the accuracy is poor

When performing FFT transformation, when completing a random access detection, the number of FFT transformations that each antenna needs to complete is at most 2 times (corresponding to random access formats 2 and 3), but because the sequence of transformation is antenna The receiving sequence often has random data, so the precision requirements for data processing are high

[0012] Therefore, when the existing FFT device performs FFT and IFFT processing, in order to meet the two kinds of transformation processing requirements, it cannot process FFT and IFFT operations in parallel, but can only use a separate device to perform FFT or IFFT operations, which has a large resource overhead

Images