Ambiguity Function Analysis For OFDM Radar

Signals
Jun Wang, Bocheng Zhang, Peng Lei
School of Electronic and Information Engineering, Beihang University, Beijing, China
Email: wangj203@buaa.edu.cn, bocheng@buaa.edu.cn, peng.lei@buaa.edu.cn

Abstract: Orthogonal Frequency Division Multiplexing signal with the matched filter firstly [4]. In the process of
(OFDM) is an efficient modulation method which is widely signal transmission, delay and Doppler frequency shift will be
used in many mobile communication systems. Applying introduced which will produce the signal distortion [5]. In
OFDM signals to radar systems will make the combination of order to analyze the mismatch between the received signal and
radar and communication systems in one single system matched filter in the condition of different delay and Doppler
platform possible. In this paper, the auto-correlation and cross- shift, the ambiguity function is defined [4][5].
correlation components of the ambiguity function are
researched for the analysis of the performance of the OFDM s t represents the transmission signal and the ambiguity
signal resolution. The impact of sparse subcarriers on the function could be represented as [6],
resolution of the delay and the Doppler is also discussed.
f
Finally, the conclusions are verified. F W , fd ³ f
s t s t  W e j 2S fd t dt (1)
Keywords: OFDM Radar; Ambiguity Function; Sparse with W denoting the delay, f d being the Doppler frequency,
Subcarriers; Ambiguity; Resolution
s t being the complex conjugate of s t .

I. INTRODUCTION And Eq. (1) can be rewritten using Parseval’s theorem as [4]
Orthogonal Frequency Division Multiplexing (OFDM), a f
multicarrier modulation method encoding digital data on F W , fd ³f
S f S f  f d e  j 2S f W df (2)
multiple orthogonal subcarriers, can achieve high-speed data
transmission through parallel low-speed subcarriers. It has where S f is the Fourier transforms of s t .
many advantages, such as high spectral efficiency, robustness
against Inter Carrier Interference (ICI) and Inter Symbol In the radar system, the matched filter can be used to obtain the
Interference (ISI), smaller error rate, implementation information of the target’s distance. Therefore, the mismatch
modulation / demodulation by IFFT / FFT, etc [1]. OFDM is between targets with different state of motion can be obtained
widely used in Wireless Local Area Network (WLAN), the by analyzing the ambiguity function of the radar signal, that is,
3rd/4th Generation mobile communication and Digital Audio the radar resolution of distance and velocity. In summary, the
Broadcasting (DAB) systems [2]. Since the superior ambiguity function is an effective tool for analyzing the range /
performance of high rate communication under the condition velocity resolution of radar waveform.
of high-speed mobile, OFDM is also used in waveform design
of intelligent transportation systems [3]. III. THE STRUCTURE OF OFDM SIGNAL
Ambiguity function is an important tool to analyze and design OFDM is a multi-carrier modulation. It converts high-speed
radar signal. By investigating the ambiguity function of the serial communication data into N parallel data streams, then
radar waveform, we can obtain the resolution of range and modulates the parallel communication data on N orthogonal
Doppler, and analyze the impact of different waveform subcarriers, and finally add those N modulated signals to get
parameters on the radar resolution. one OFDM symbol. The duration of a single OFDM symbol is
longer, but it is possible to achieve a higher communication
The paper is organized as follows. First, a brief introduction of rate by carrying N modulation data in one OFDM symbol, that
the ambiguity function will be given. Then, the structure and is, OFDM implement high-speed serial data communication by
the parameters of OFDM signal are introduced. In the multiple lower rate subcarriers which are parallel and
following section, the ambiguity function of the OFDM signal orthogonal. OFDM signals can be expressed as [2]
will be derived and the impact of sparse subcarriers on the
M 1 N 1
ambiguity function will be presented and discussed. Moreover,
¦ ¦e
j 2S f 0  k f t  lTssym
sl t p t  lTsym (3)
the results of the analysis will be verified with MATLAB l 0 k 0
simulation. Finally, a conclusion will be given. where M denotes the number of OFDM symbols, N denotes
the number of OFDM subcarriers, f 0 is the carrier frequency,
II. AMBIGUITY FUNCTION
f denotes the frequency difference between two adjacent
In order to make the output of the receiver possess the
maximum SNR, the receiver usually processes the received
subcarriers, Tsym denotes the elementary OFDM symbol Let W 0 , we can get the zero time cut along the Doppler
frequency axis, which is also called zero-delay cut,
duration, p t denotes the function of rectangular window.
j 2S fd Tsym
e 1 N N 1 N 1
1
All subcarriers overlapped on the spectrum are orthogonal as F 0, f d ˜ ¦ ¦ (8)
the case of f 1 Tssym is fulfilled. j 2S f d k1 0 k2 0, k2 z k1 k1  k2 f  fd
Let f d 0 and 0 d W  Tsym , we can get the zero Doppler cut
1 Tsym
³
j 2S f 0  k f t  j 2S f 0  l f t
e e dt along the time-delay axis, which is also called zero-Doppler
Tsym 0
cut,
1 Tsym
³ e j 2S k l ft
dt (4) F W,0
Tsym 0

§ W · N 1 j 2S kW N 1
e j 2S k2W f  e j 2S k1 fW (9)
­1, k l Tsym ¨1  ¸¸ ¦ e  ¦
f
® ¨ T j 2S k1  k2
¯0, else © sym ¹k 0 k1 , k2 0, k1 z k2

(a)

Fig. 1 Spectrum of OFDM signal

IV. THE ANALYSIS OF AMBIGUITY FUNCTION

In the case that the signal processing union of the radar system
is single OFDM symbol, the ambiguity function can be
expressed as
f
F W , fd ³ f
s t s t  W e j 2S fd t dt
(5)
F auto W , f d  F cross W , f d
where
(b)
N 1 f
F auto W , f d = ¦ e j 2S kW f
³ p t p t W e j 2S f d t
dt (6) Fig. 2 The ambiguity function of single OFDM symbol with 8
f
k 0 subcarriers and the Doppler range is (a) [-40 40] (b) [-8 8]
F cross W , f d
N 1 N 1
In the case that the signal processing union of the radar system
f (7) is multi OFDM symbols, the ambiguity function can be
=¦ ¦
j 2S » k1  k2
³
j 2S k2W f f  f d º¼ t
e p t p t W e dt
k1 0 k2 0, k2 z k1
f expressed as
F auto W , f d which denotes the ambiguity function when
k1 k2 , is the auto-correlation component of each subcarrier F W , fd
and the energy of the ambiguity function mainly concentrates f M 1 N 1

³ ¦ ¦
j 2S ¬ª k1  k2 f  f d ¼º t
in this part. F cross W , f d which demotes the ambiguity p t  l1Tsym p t  W  l2Tsym ˜ e e j 2S k2W f dt
f
l1 , l2 0 k1 , k2 0
function when k1 z k2 , is the cross-correlation component
(10)
between the subcarriers and shows as the interference between When 0 d W  Tsym , we can get
the adjacent channels.
F W , fd F 0, f d M
N 1 N 1 N N
¦ ¦
MTsym MTssym j 2S ª¬ k1  k2
³W ³W
f  f d º¼ t 1 1
e j 2S k2W f
e j 2S fd t dt  e j 2S k2W f
e dt N e j 2S f d T  1 M M
e
j 2S ¬ª k1  k2 M f  f d º¼ T
 1 (17)
k1 k2 0 k1 , k2 0, k1 z k2 ˜ ¦ ¦
M j 2S f d k1 0 k2 0, k2 z k1 j 2S ª¬ k1  k2 M f  f d º¼
(11)
Let W 0 , we can get the zero-delay cut,
It is easy to get that the maximum value for F 0, f d M
occurs
e
j 2S f d MTsym
1 N 1
e
j 2S f d MTsym
1 at f d 0, r M f , r2M f , r3M f . When the subcarriers
F 0, f d N
j 2S f d
 ¦
k1 , k2 0, k1 z k2 j 2S ª¬ k1  k2 f  f d º¼ become sparser, the amplitude of the side-lobe becomes higher
and the spacing between the main lobe and the grating lobe is
(12) M f showing in Fig. 4.

(a) (b)

Fig. 3 The ambiguity function of multi OFDM symbol with 8

subcarriers, 4 symbols

If K subcarriers are chose from N subcarriers of the OFDM

signal for communication and radar system, the N subcarriers
can be assigned to N K users. Following, the impact of sparse
subcarriers on the ambiguity function is analyzed.
(c) (d)
The OFDM signal with sparse subcarriers can be expressed as
N
Fig. 4 The impact of sparse subcarriers on zero-delay cut with
1
M 64 subcarriers and (a) M 1 , (b) M 2 ,(c) M 4 ,(d)
s t ¦e j 2S kM ft
p t (13) M 8
k 0
Let f d 0 and W t 0 , we can get the zero-Doppler cut with
where M N K is the coefficient of sparse subcarriers, and
sparse subcarriers
the ambiguity function of the OFDM signal with sparse
subcarriers can be expressed as F W ,0 M
F auto W ,0 M
 Fcross W ,0 M
(18)
F W , fd F auto W , f d  Fcross W , f d (14) where
M M M
with N
M
1

N
1
F auto W , 0 M
= ¦ T  W e j 2S M fkW
(19)
M f k 0
F auto W , f d M ¦e
k 0
j 2S kW M f
³ f
p t p t W e j 2S f d t
dt (15) N
M
1
N
M
1

=¦ ¦
T
³W e
j 2S M f k1  k2 t
F cross W , f d F cross W , 0 M
e j 2S M fk2W
dt (20)
M k1 0 k2 0, k2 z k1

(16) As the energy of the ambiguity function is mainly concentrated

N N
1 1
M M f
¦ ¦ e on F auto W , 0 M , the influence of F cross W , 0 M will be ignored,
j 2S ¬ª k1  k2 M f  f d º¼ t
³f p t p t  W e
j 2S k2W M f
dt
k1 0 k2 0, k2 z k1 then the zero-Doppler cut will be expressed as
Let W 0 , we can get the zero-delay cut with sparse
subcarriers F W , 0 M | F auto W , 0 M
sin S N f W (21)
T W ˜
sin S M f W
It is easy to get that the maximum value for F W , 0 M
occurs Symbol Table Column Head Value

nN M Number of OFDM symbols 256

at W Tsym where n is integer. So when the subcarriers
M 'f Subcarrier spacing 50kHz
become sparser, the spacing between the main lobe and the
grating lobe in the zero-Doppler cut will become more narrow Tsym Duration of OFDM symbol 20­s
showing in Fig. 5. B Totla signal bandwidth 3.2MHz

'd Radar range resolution 46.8750m

'v Radar velocity resolution 4.9656m/s

At the receiver, each OFDM symbol is processed by match
filter to obtain the location of the targets, then extract target
Doppler information by doing FFT transforming to the
received multiple symbols of the same distance gate [7].
After matched filtering and pulse accumulation, the Range-
(a) (b) Doppler distribution of two targets can be got, shown as in Fig.
6.

(c) (d)
(a) (b)
Fig. 5 The impact of sparse subcarriers on zero-Doppler cut
with 64 subcarriers and (a) M 1 , (b) M 8 ,(c) M 16 ,(d)
M 32
The spacing between the main lobe and the grating lobe will be
wider in the zero-delay cut, and be more narrow in the zero-
Doppler cut when the subcarriers become sparser. For example,
when M 4 , the spacing between the main lob and the grating
lobe in the zero-delay cut of the sparse OFDM signal will be
quadruple, while that in its zero-Doppler cut becomes a quarter.
(c) (d)
V. SYNTHESIS ALGORITHM OF OFDM SIGNALS
By studying the characteristics of the ambiguity function, we Fig. 6 The simulation figure and the ambiguity function with
can analyze the performance of the radar resolution. Next the the impact of sparse subcarriers and (a) M 1 , (b) M 4 ,(c)
identification with two targets by specific location and specific M 1 ,(d) M 4
speed will be simulated. The processing unit is the OFDM It can be seen from Fig. 6 that the two targets are already
signal with 64 subcarriers and 256 symbols, and subcarrier obvious, the location and speed of the targets can be measured
spacing is 'f 50kHz which leads to the duration of the by the method of a matched filtering and pulse accumulation.
However the maximum unambiguous range of the sparse
OFDM signal is Tsym 20Ps , and the carrier frequency is
OFDN signal with M 4 is 750m, while that in the stand
f c 5.9GHz . It is assumed that the location of the two targets OFDM signal is 3000m.As the subcarriers become sparser, the
is 600m and 650m, the speed of the two targets is 200m/s and range ambiguity will be introduced.
210m/s, and the sampling frequency is 20MHz. An overview
on all system parameters is provided in Table I. VI. CONCLUSIONS
TABLE I. SYSTEM PARAMETERS
The signal models of one symbol OFDM and multi OFDM
symbols are constructed, and the ambiguity function is derived.
Symbol Table Column Head Value The ambiguity functional shapes show the perfect thumbtack
shape substantially, but there is Doppler ambiguity on the
fc Carrier frequency 5.9GHz
Doppler axis. Increasing the OFDM symbols can reduce the
N Number of subcarriers 64 Doppler ambiguity. The Doppler ambiguity can be reduced
with sparse subcarriers, but this will introduce the range
ambiguity. The result of the analysis indicates that the OFDM
signal can be used to obtain high resolution Doppler frequency
and delay, and these conclusions have certain reference
significance to OFDM radar waveform design and signal
processing.

REFERENCES

[1] Yong Soo Cho, Jaekwon Kim, Won Young Yang, Chung G. Kang.
MIMO-OFDM Wireless Communications with MATLAB. Singapore:
John Wiley & Sons; 2010.
[2] Hermann Rohling. OFDM: Concepts for Future Communication
Systems. Springer: Springer Science & Business Media; 2011.
[3] Christian Sturm, Thomas Zwick, Werner Wiesbeck. “An OFDM System
Concept for Joint Radar and Communications Operations.” IEEE 69th
Vehic. Tech. Conf., pp. 1–5, Apr.2009.
[4] Bassem R. Mahafza. Radar Systems Analysis and Design Using
MATLAB Second Edition. Huntsiville: CRC Press, 2005.
[5] Mervin C. Budge, Shawn R. German. Basic Radar Analysis. London:
Artech House, 2015.
[6] Merrill I. Skolnik. Introduction to Radar Systems. Washington: Tata
McGraw Hill, 2001.
[7] Gabriel Lellouch, Amit Mishra, Michael Inggs. “Impact of the Doppler
modulation on the range and Doppler processing in OFDM radar.” 2014
IEEE Radar Conference. Pp. 0803 – 0808, May 2014.