4
th
International Conference on Information, System and Convergence Applications
13 -18 August 2017, Ulaanbaatar, Mongolia
Anti-aliasing filters design for three-band overlaped signals
receiver
1st Hongmei Wang, 2nd Faguang Wang*, 3rd Shiyin Li, 4th Sanghyuk Lee
1st-4th Authors –School of Information and Control Engineering, China University of Mining and Technology, Xuzhou,
China, e-mail address: wfaguang@cumt.edu.cn
5th Author - Electrical and Electronic Engineering, Xi’an Jiaotong-Liverpool University, Suzhou, China
Abstract—Simultaneously receive multi-band RF signals is one of requirement in software defined radio(SDR).
Aliasing between received signals are frequent occurrence and limit the flexibility of SDR receiver. This paper
proposed one anti-aliasing algorithm to separate three overlapped signals. This method can reduce hardware
burden and use software to solve the aliasing problem.
Keywords-Bandpass sampling; multi-band signals; receiver; third order sampling
I.
INTRODUCTION
How to digitize RF signals is the first problem faced
by SDR(software defined radio),i.e., how to sample
interested signal[1]. BPS(bandpass sampling) which can
use the sampling rate much lower than the highest
frequency of the RF signal is widely used in SDR
receiver. With the development of wireless
communication, multi-band receiver takes more
application. But when receive multi-band signals,
aliasing between multi-signals occur frequently. Most
researchers solve this problem to select sampling
frequency to avoid aliasing. Which limit the selection of
sampling frequency[2-4]. Second order BPS is proposed
to reduce the sampling rate and process aliasing. For the
second order BPS system, most of the studies are based
on orthogonal signal processing to anti-aliasing. For
example, [5] using the Hilbert transform after sampling,
while in literature [6], 90 degree hybrid crossover
signals are sampled. In [7], two sampling device
generates a phase difference of 90 degrees. These studies
allow the sampling system use smaller sampling
frequencies, but can only eliminate the aliasing of the
signal’s own image and cannot process aliasing between
the two signals. In the previous work, second order BPS
with adjustable time delay was used to realize the
reception of two overlapped signals[8]. In this paper,
Third order BPS structure is designed to receive and
process three signals.
This paper is organized as follows. In Section II,
Third order BPS structure is given and the spectra of
three sampled channel is analyzed. In Section III, design
method of anti-aliasing filters were given and the effect
of parameters is discussed. Section IV presents the
simulation result, and Section V provides concluding
remarks.
1
Foundation item: This work was supported by the national natural
science foundation of China(61601464) and (51541410).
II. STRUCTURE OF THIRD ORDER BPS
Assume the primary bandpass signal
()Rf
is band
limited to B and is sampled by a sampling rate of
2
s
B
. Any signal in the frequency zone of index n
expressed by
(1/2) (1/2)nffnf
s
(1)
is aliased into the first Nyquist zone
B
, where the
first Nyquist zone is the frequency zone with index zero.
Structure of third-order BPS system is shown in
figure 1.
1
()T
123
SSS
1
S
A
S
S
C
S
2
()T
Figure 1 Structure of third-order BPS system
Assume three RF signals
1
S
,
2
S
and
3
S
with spectra
1
()Rf
,
2
()Rf
and
3
()Rf
, as shown in figure 2, are
sampled by third order BPS using structure of figure 1.
All the three signals have positive and negative spectrum
defined as
()
i
Rf
and
()
i
Rf
(i=1,2,3). The three
sampled channel are denoted as channel A, channel B
and channel C. As shown in figure 3, define spectrum in
channel A as:
1122
() () () () ()
AA A A A
fRfRfRfRf
33
() ()
AA
RfRf
(2)