Cyclic Autocorrelation based Blind OFDM Detection and ...

Cyclic Autocorrelation based Blind OFDM Detection
and Identification for Cognitive Radio
Ning Han, Guanbo Zheng, Sung Hwan Sohn, Jae Moung Kim
INHA-WiTLAB, INHA University
253 Younghyun-dong, Nam-gu, 402-751
Incheon, Korea
neil_han@ieee.org, gbzheng@gmail.com, sunnyshon@gmail.com, jaekim@inha.ac.kr
Abstract—Cognitive radio is considered as a promising technique
to increase the utilization of limited spectral resource. The key
issue in cognitive radio is to design a reliable spectrum sensing
method that is able to detect the signal in the target channel as
well as to recognize different signals. In this paper, focusing on
classifying different OFDM signals, we propose a two-step
detection and identification approach. The key parameters to
separate different OFDM signals are the subcarrier spacing and
guard interval. A simple but reliable peak detection method is
adopted in the first step, while a peak searching method is used
to determine the length of guard interval. Simulations are
carried out in AWGN to verify the validation of the proposed
method. It is shown that our method can satisfy the detection
and identification requirement with a low false alarm
probability.
Keywords-cognitive radio; spectrum sensing; OFDM;
cyclostationary; signal detection and identification
I. INTRODUCTION (HEADING 1)
It is commonly believed that there is a scarcity of spectrum
availability at frequencies that can be economically used for
wireless communications, especially in the bands below 3 GHz
[1]. However, according to the measurements taken in [2], the
shortage is mainly caused by the spectrum policy under which
the spectrum is licensed to a limited number of
implementations. Cognitive radio provides the opportunity to
utilize the vacant spectrum resources while helping to prevent
interference to primary users that own the frequency bands. It
is defined as an intelligent wireless communication system that
is aware of its surrounding environment, and learns from the
environment and adapts its internal states to statistical
variations in the incoming RF stimuli by making
corresponding changes in certain operating parameters in real
time [3]. The new functionality (spectrum sensing) requires
that the radio is able to separate the vacant frequency bands
from those filled with primary user signals accurately.
It is well known that by splitting a single high-rate data
This work was supported by the Korea Science and Engineering
Foundation (KOSEF) through the National Research Lab. Program funded by
the Ministry of Education, Science and Technology (No. M10600000194-
06J0000-19410). This work was supported by the Korea Science and
Engineering Foundation (KOSEF) grant funded by the Ministry of Education,
Science and Technology (MEST) (No. R01-2006-000-10266-0(2008)).
978-1-4244-2108-4/08/$25.00 © 2008 IEEE
stream into a number of lower rate subcarriers, OFDM holds
several advantages including robustness against frequency
selective fading or narrowband interference [4]. Recently, due
to the booming of WiFi and WiMAX, OFDM based systems
are becoming the trend for next generation wireless
communications. Therefore, the detection method separating
OFDM signal from other single carrier signal or random noise
comes to the frontier. [5] proposed to exploit the embedded
periodicity among the subcarriers in OFDM signal. [6]
developed several criteria based on the time domain
periodicity introduced by the cyclic prefix in DVB-T OFDM
symbol. These approaches require either the number of
subcarriers or the spacing between consecutive subcarriers in a
priori. However, different OFDM systems usually own their
unique parameters due to various applications. Even in a single
OFDM system, there are several operation modes with
different parameters to achieve various transmission data rates.
These make the detector almost impossible to know the
information in advance. Thus, the existing methods are
impractical to detect OFDM signal blindly.
In this paper, by considering the uncertainties in the
detector, we proposed a time domain cyclostationarity based
approach to detect and identify OFDM signal from random
noise. The key parameters to discriminate different OFDM
signals are the subcarrier spacing and duration of guard
interval. The proposed approach consists of two steps. The first
step is to detect the OFDM signal from random noise simply
by recognizing the symmetric peaks in the autocorrelations,
which is a special case of the cyclic autocorrelation function.
The subcarrier spacing is calculated as long as the OFDM
signal is detected. In the second step, by searching the cycle
frequencies, the length of guard interval is calculated to
recognize different OFDM signals. The main advantage is that
the proposed method does not require the FFT process. Since
the number of subcarrier is unknown, the mismatch of FFT
parameters will reduce the performance of any FFT based
detection methods.
The rest of our paper is organized as follows; Section II
describes the time domain cyclostationarity of OFDM signal in
terms of cyclic correlation function (CAF). Section III presents
the proposed detection procedure and criteria based on the
blind assumption. Simulation results are discussed in Section