The
14m
IEEE
2003
International
Symposium
on Persona1,lndoor and
Mobile
Radio
Communication Proceedings
A
Channel-Estimation-Based
Equalization Algorithm
for
Fast Frequency
Hopping
Radio
Jinhuia
Sun,
Jiandong
Li,
Lijun
Jin
State Key Lab. of Integrated Service Networks
Xidian University
jhsun@xidian,edu.cn
Xi’an, Shaanxi.
710071,
China
ADstmct-This paper presents a channel estimation based
equalization algorithm for high speed. fast hopping radio. The
equalizer coefficients are directly calculated from the impulse
response
of
the channel which
is
obtained by rapid channel
estimation in time domain employing
cyclically
extended
binary sequence. Simulation is done under general two-path
channel model and rapidly fading channels. Theoretical analysis
and computer simulation results show that compared with
traditional adaptive coefficients adjusting equalizer or
MLSE
receiver, this algorithm is more robust and has less
computational overhead, which is desirable under limited time
resources.
K~II
ords-fr.equerrcy
hopping;
clmiiiel
estimntion;
channel
equiilizntiori
I.
INTRODUCT~ON
In
frequency hopping system, the frequency hopping rate of
tactical
hopping radio is becoming faster and faster for high
interference resist ability and data transmission rate. High-data-
rate transmission and fast frequency hopping have brought
forth many problems such as: PLL frequency synthesizer
should have fast frequency hopping ability, the receiver of
frequency hopping system should be able to set up rapid
frequency synchronization and tracking. High data
transmission rate has made intersymbol interference
(ISI)
an
iinneglectable problem. This paper focuses on the channel
equalization algorithm
to
reduce
[SI.
In our experimental high
speed
radio,
the frequency hopping rate is
1
000hops/sec,
SO
all
signal processing must be finished within
lins
for one hop data.
Since the processing time is very limited, fast channel
estimation and channel equalization algorithm with low
Computation burden are required to track time-varying channels.
It
has been realized that Decision Feedback Equalizer (DFE)
structure has good performance
to
channels with severe
amplitude distortion due to selective fading, and its
performance is similar to the optimal but much more complex
MLSE
receiver
[I1.
For
a
long time, how to adjust DFE
coefficients has been an important research subject.
As
a
kind
of
adaptive inverse modeling system, the optimum equalizer
solution is influenced by the power spectrum density of noise.
When it tracks rapidly fading channels, adaptive filter
algorithm often goes to diverge, which leads
to
equalizer
fai
I
ure
.
This paper
is
supported by
863
hightech key project
(700
I
A
12303
I),
TRAPOYT, key project
of
MOE.
0-7803-7822-9/03/$17.00
0
2003
IEEE
Xiaojun
Wu
Department of Computer Science and Engineering
Northwestern Polytechnical University
depender@yahoO.com
Xi’an, Shaanxi,
7
10072,
China
Some researchers have been studying the method of
calculating the equalizer’s coefficients of rapidly fading
channels directly from the channel impulse response. It has
been proved to be an efficient method, especially compared
with traditional iterative adjustment methods. In
[
11,
the
equalizer’s coefficients are determined by the autocorrelation
of the impulse response of the channel using Cholseky
fractorization. In
[2],
B.Farhang Boroujeny presented an
iterative algorithm to derive the equalizer’s coefficients
in
order
to
avoid complex computation when the number
of
channel parameters is very large.
Both methods are helpful to channel equalization on burst
mode communication, such as in Meteor Burst Communication
(MBC)
systems and GSM system. In our high speed VHF
hopping radio, voice or data short burst has
a
training sequence
in the beginning. The training sequence is used
to
estimate
the
channel impulse response. The
channel-estimation-based
equalization can be used. We employ fast time domain channel
estimation, because it is less sensitive to channel noise. Based
on the Minimum Mean Square Error (MMSE) rule,
Gauss-
Seidel algorithm (an iterative algorithm to find out the solution
of set of linear equation) is adopted
as
the mapping method
to
derive equalizer’s coefficients from channel parameters
calculated in channel estimation. The output of equalizer
is
detected, then fed back to the feedback filter.
This paper is organized as follows: Section
I1
introduces
our simulation system model; Section
111
gives the principle of
fast channel estimation; in Section
IV,
we will discuss channel
estimation based equalization algorithm and the simulation
results in detail. Finally, we conclude the whole paper
in
Section
V.
11.
SYSTEM MODEL
The basic M-PSK system block diagram is shown
in
Figure
1.
Information bits are changed from serial stream into
parallel stream and then mapped into constellation points. The
mapped data together with the training sequence is quadrature
amplitude modulated and transmitted over the physical
channel. In the receiver the received signal is down converted
into baseband signal and the equalization is done before the
symbol is detected. The whole system before equalization
is
equal to
a
discrete complex baseband channel model, as
735
Authorized licensed use limited to: Communication University of China. Downloaded on April 08,2010 at 11:03:02 UTC from IEEE Xplore. Restrictions apply.
资源评论
liang888888
- 粉丝: 0
- 资源: 2
最新资源
- Opencv+ROS自编相机驱动
- python绘制爱心表白专用
- 基于Jupyter实现的深圳市道路交通事故数据分析+源码（毕业设计&课程设计&项目开发）
- 车辆车牌检测源码和报告，使用python编写，下载即可运行，可做毕业设计
- ptgame-master1.zip
- GSDML-V2.3-wenglor-wenglor ident-20161007-112500.xml
- stm32心率检测keil5工程
- GSDML-V2.2-Murrelektronik-IMPACT67-20120315.xml
- GSDML-V2.31-Murrelektronik-MVK-MPNIO-F-20150903.xml
- 通过C#上位机与库卡（KUKA）机器人进行TCP通讯，实现实时位置返回及运动控制
资源上传下载、课程学习等过程中有任何疑问或建议，欢迎提出宝贵意见哦~我们会及时处理！
点击此处反馈
