Methods and apparatus for narrow band interference detection and suppression in ultra-wideband systems
View Patent ↗An exemplary method is disclosed to accurately estimate the center frequency of a narrow-band interference (NBI). The exemplary method uses multi-stage autocorrelation-function (ACF) to estimate an NBI frequency. The exemplary method allows an accurate estimation of the center frequency of NBI in an Ultra-Wideband system. A narrow band interference (NBI) estimator based on such a method allows a low complexity hardware implementation. The exemplary method estimates the frequency in multiple stages. Each stage performs an ACF operation on the received signals. The first stage gives an initial estimation and the following stages refine the estimation. The results of all stages are combined to produce the final estimation. An apparatus based on such a multi-stage narrow band interference frequency detector is also disclosed to improve the accuracy by combining various filters with the detector.
1. A computational method for narrow-band interference detection based on discrete signal processing in a radio frequency receiver, the method comprising:
initially estimating, via a first autocorrelation function (ACF) unit, a narrow-band interference frequency in a narrow-band interference frequency detector using a first stage of a multi-stage autocorrelation-function applied to an input signal;
estimating, via a second ACF unit, at least one residue frequency based on at least one other autocorrelation stage of the multi-stage autocorrelation-function, the input signal and the initial frequency estimate; and
combining the results of multiple stages of the multi-stage autocorrelation-function for final estimation, wherein
the final estimation produces an estimate of the narrow-band interference frequency based on the combined results of the multi-stage autocorrelation-function,
the initial frequency estimate of the first autocorrelation stage is computationally expressed as
f
^
i
,
1
=
1
2
π
T
arg
{
ACF
(
m
;
M
,
1
)
}
wherein arg{ACF (m; M,1)} is the angular function that returns the angle of a complex number, and wherein ACF(m;M,1) is the autocorrelation function (ACF) defined as
ACF
(
m
;
M
,
1
)
=
∑
l
=
0
M
-
1
R
(
m
+
l
)
R
*
(
m
+
l
+
1
)
where R(m+l) is a discrete received signal R at time instant m+l, m being an index of a first sample in a first segment, and l being an offset index; ( )* denotes a conjugation operator; and M is summation window size of the ACF.
2. The method according to claim 1 , wherein the narrow-band interference frequency being estimated is a center frequency of the narrow-band interference.
3. The method according to claim 1 , wherein each stage of the multi-stage autocorrelation-function is based on processing an autocorrelation-function computing operation on a received signal.
4. A narrow band interference estimator apparatus based on a hardware implementation of a process embodying the method according to claim 1 .
5. The narrow band interference estimator according to claim 4 , further includes an adaptive filter combined with the narrow band interference estimator.
6. A multi-stage narrow band interference frequency detector comprising:
a first autocorrelation stage to produce, via a first autocorrelation function (ACF) unit, an initial frequency estimate of a narrow band interference based on an input signal; and
at least one other autocorrelation stage to estimate, via a second ACF unit, a residue frequency based on the input signal and the initial frequency estimate, wherein an output estimate {circumflex over (f)} i of the multi-stage narrow band interference frequency detector is based on a combination of the initial frequency estimate of a narrow band interference and the at least one estimate of the residue frequency, wherein
the residue frequency estimate of the at least one other autocorrelation stage g, where g>1, is computationally expressed as
f
^
i
,
g
=
1
2
π
K
g
T
arg
{
ACF
(
m
;
M
,
K
g
)
exp
(
-
j
2
π
K
g
T
∑
n
=
1
g
-
1
f
^
i
,
n
)
}
where
arg
{
ACF
(
m
;
M
,
K
g
)
exp
(
-
j2π
K
g
T
∑
n
=
1
g
-
1
f
^
i
,
n
)
}
is the angular function that returns the angle of a complex number; ACF(m;M,K g ) is the autocorrelation function (ACF) defined as
ACF
(
m
;
M
,
K
g
)
=
∑
l
=
0
M
-
1
R
(
m
+
l
)
R
*
(
m
+
l
+
K
g
)
where R(m+l) is a discrete received signal R at time instant m+l, m being an index of a first sample in a first segment, and l being an offset index; ( )* denotes a conjugation operator; K g is the “lag” of stage g; and M is summation window size of the ACF.
7. The multi-stage narrow band interference frequency detector according to claim 6 , wherein the lag K s of the final stage (K g , where g=S) satisfies max(B i )<1/K s T; max(B i ) indicates a maximum bandwidth of the detectable narrow band interference;
and the lag K g of stage g increases with the value of g.
8. A multi-stage narrow band interference frequency detector comprising:
a first autocorrelation stage to produce, via a first autocorrelation function (ACF) unit, an initial frequency estimate of a narrow band interference based on an input signal; and
at least one other autocorrelation stage to estimate, via a second ACF unit, a residue frequency based on the input signal and the initial frequency estimate, wherein an output estimate {circumflex over (f)} i of the multi-stage narrow band interference frequency detector is based on a combination of the initial frequency estimate of a narrow band interference and the at least one estimate of the residue frequency, wherein
a sampled baseband signal is used as the input signal and has a sampling interval of T=I/B u wherein B u is Bandwidth,
the output estimate {circumflex over (f)} i of the multi-stage narrow band interference frequency detector is a weighted combination of the initial frequency estimate of a narrow band interference and the at least one estimate of the residue frequency, and a center frequency narrow band interference can be any value in the range [-B u /2, B u /2].