Algorithms for recognition of OFDM based systems

1. Recognition of OFDM based systems and opportunist spectrum access

An opportunist base station is a base station able to sense its spectral surrounding to determine which bands are used and which bands are free. If an opportunist base station exists, the opportunist terminal has to be able to find it, and to recognize an opportunist base station from a primary user (the user whose bands may be used in an opportunist manner).

We focus on this issue in this demonstration. As many standards are now based on an OFDM modulation, we assume that both primary user and secondary user (opportunist base station) used OFDM based modulation but with different parameters.

If the opportunist terminal receives an OFDM signals, it has to know which kind of base station has transmitted this signal. It can therefore estimate the OFDM parameters of this signal. Of course, this kind of algorithms also apply to military applications.

For more information about the algorithms used in this tutorial, you can refer to [1] [2] [3] [4].

2. Description of this demonstration

This demonstration has been done with 2 PC running matlab and windows, each one connected to an USRP with a daughterboard RFX2400. On the first PC, 4 OFDM signals have been generated with different parameters, and transmitted to the second computer thanks to the USRP.

The objective on the second PC is to estimate the OFDM parameters of the transmitted signals.

The received signal can be downloaded from this page: [acqui1.dat], [acqui2.dat], [acqui3.dat] and [acqui4.dat]. A matlab script to open the file is also available for downloading : [ReadAcquiSignal.m]. For example: Note that the signal described in these files are sequences of bursts, each burst being a replica of the signal of interest. The first treatment consists hence in selecting one of these burst.

3. Performance of the tested algorithms

3.1. Parameters used to generate the OFDM signals

This is the list of parameters used to generate the signals. These signals have been sampled at $T_e=1 \mus$

• Acqui #1: $NT_c=320 \mus$, $D/N=1/4$
• Acqui #2: $NT_c=640 \mus$, $D/N=1/32$
• Acqui #3: $NT_c=512 \mus$, $D/N=1/8$
• Acqui #4: $NT_c=3072 \mus$, $D/N=1/16$

3.2. Cyclic-correlation based method [1]

The following parameters have been estimated:

• Acqui #1: $NT_c=320 \mus$, $D/N=1/4$
• Acqui #2: $NT_c=640 \mus$, $D/N=1/32$
• Acqui #3: $NT_c=512 \mus$, $D/N=1/8$
• Acqui #4: $NT_c=3072 \mus$, $D/N=1/16$

The cost function for the system #2 is the following one:

3.3. Kurtosis based method [4]

Note that for the following methods, the frequency and the time synchronisation offsets have to be estimated.

• Acqui #1: $NT_c=320 \mus$, $D/N=1/4$
• Other acqui : in progress

The cost function for the system #1 is the following one:

References:

[1] A. Bouzegzi, P. Jallon et P. Ciblat : A second order statistics based algorithm for blind recognition of OFDM based systems, IEEE Global Telecommunications Conference (GLOBECOM), Novembre 2008.
[2] A. Bouzegzi, P. Ciblat et P. Jallon : Matched Filter based algorithm for blind recognition of OFDM systems, IEEE Vehicular Technology Conference (VTC), Calgary (Canada), Septembre 2008.
[3] A. Bouzegzi, P. Ciblat et P. Jallon : Maximum Likelihood based methods for OFDM intercarrier spacing characterization, IEEE International Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC), Septembre 2008.
[4] A. Bouzegzi, P. Jallon et P. Ciblat : A fourth-order based algorithm for characterization of OFDM signals, IEEE Workshop on Signal Processing Advances in Wireless Communications (SPAWC), Juillet 2008.

Attached files:

[ AF1 ] acqui1.dat
[ AF2 ] acqui2.dat
[ AF3 ] acqui3.dat
[ AF4 ] acqui4.dat
[ AF5 ] ReadAcquiSignal.m