Adaptive Double-threshold Joint Spectrum ... - Semantic Scholar
JOURNAL OF COMPUTERS, VOL. 8, NO. 10, OCTOBER 2013
2565
Adaptive Double-threshold Joint Spectrum Sensing Based on Energy Detection XiaoLin Jiang School of Electronics and Information Engineering,Harbin Institute of Technology, Harbin,,china School of Electronics and Information Engineering,Heilongjiang University of science and technology, harbin, china Email: [email protected]
Xuemai Gu School of Electronics and Information Engineering,Heilongjiang University of science and technology, harbin,china Email:[email protected]
Yingpei Lin Department of Electronic Engineering, Shanghai Jiao Tong University, Shanghai, China Email:[email protected]
Abstract—Energy sense of adaptive double threshold joint spectrum perception algorithm using multiple cognitive, and after double threshold for comprehensive judgment, and threshold value with the discretion of the signal to noise ratio (SNR) adaptive adjustment, will eventually each SU decision data fusion center for spectrum using the information, make perception performance effectively improve. In this paper, the detection probability of the proposed algorithm, the virtual detection probability and frequency of testing performance expression in this paper, and by using the MATLAB simulation get signal-to-noise ratio and detection probability, the relationship between the frequency of testing. The results show that, with the traditional single threshold energy spectrum method under the perception, compared with less effective section the detection time, improving the detection probability, can significantly improve the cognitive wireless network spectrum perception performance. Index Terms—Double threshold; Average frequency of testing
Energy
detection;
I. INTRODUCTION Along with the growing demand of wireless communication, the demand os wireless spectrum communication system is also a corresponding increasing, leading to increasingly scarce spectrum resources, becoming a new bottleneck restricting the development of wireless communication. However, assigned to the existing spectrum resources in many wireless systems in time and space, there are different degrees of idle, serious lack of use[1].So people begin to seek permission without spectrum licensing users of in all legal users do Project supported by the National Nature Science Foundation of China (NSFC) under Grant No. 60832009,Heilongjiang University of Science and Technology talented persons training project. Project supported by HarBin Science and Technology Bureau under Grant No. 2013RFQXJ107. Manuscript received March 13, 2013; revised April 14, 2013; accepted April 16,2013;Corresponding author is Xumai Gu.
© 2013 ACADEMY PUBLISHER doi:10.4304/jcp.8.10.2565-2569
not have any impact on the case, and find the chance to access the authorized band internal and communication, in order to effectively improve the utilization of the limited spectrum, this idea is Cognitive Radio(CR). Spectrum. Sensing refers to the specific time and space which was detected on the utilization of spectrum resources, continuously, effectively detect whether the use of the band is authorized users, so the unused spectrum was identified. Spectrum sensing is one of the key technologies of cognitive radio system[2],the performance perceived will have a direct relationship with the cognitive wireless network. Spectrum sensing energy detection method does not need to know the main user prior knowledge, it is simple and easy to implement, but because of the influence of vulnerable to channel multipath fading, shadowing effect interference, leading to correct detection probability versus SNR decreased and decreased rapidly, in low signal to noise ratio of spectrum sensing performance is poor. Based on energy detection adaptive dual energy detection limit of cooperative spectrum sensing algorithm using multiple cognitive after double threshold decision, and the threshold value adjustment with high and low signal to noise ratio adaptive, eventually every SU decision data center is obtained after spectral information fusion, completely the sentence whether it is the primary user signal, make improve perceived performance effectively.
II.ENERGY DETECTION METHOD The detection process energy detection method can be integrated into a binary hypothesis, Block diagram is shown in Figure 1.
2566
JOURNAL OF COMPUTERS, VOL. 8, NO. 10, OCTOBER 2013
In the formula, Γ（.,.) says the incomplete gamma function, define it as ∞
Γ（a,b)= ∫ t a −1 exp(−t )dt b
Figure 1
Similarly, we can get the correct detection
Energy detection principle
probability. In the H1, the correct detection probability is Directly filtering of signal sample values, modulo,
Γ（
then square can be obtained, or do N point FFT of the
Pd =
received signal, converted to the frequency domain, then the frequency domain signal modulus square also can be. Energy detection principle is the energy accumulation in a certain frequency range, if the accumulated energy
(4)
If you use the right tail probability distributions to represent it,and
values above a certain threshold, that signals exist, If less
∞
Qx 2 = ∫ p (t )dt , t ≥ 0
than a certain threshold value, the only noise is present. It
N
t
is the starting point of signal plus noise energy larger than the energy of noise[3].
N λ ， 2 ） 2 2(σ w + σ s2 ) N Γ（ ) 2
(5)
and Qx（ x）= 2
Supposing:
N
⎧ H 0 , x ( n ) = w( n ) , n = 1, 2,... N ⎨ ⎩ H 1 , x ( n ) = s ( n ) + w( n ) (1) In the formula, s ( n ) is the main input of user information, is assumed to be zero mean, the variance of normal distribution of
σ s2 , w ( n ) is noise information, if
⎧ ⎪2Q （ x） N = 1 ⎪ 1 1 ⎪ k− exp(− x) ( N −1）/2 (k − 1)!(2 x) 2 ⎪ 2 , N is odd （ x) + ⎨2Q ∑ (2k − 1)! π k =1 ⎪ 1 ⎪ k− ⎪ 1 ( N −1）/2 ( x / 2) 2 , N is even （ ⎪ − 2 x） ∑ k! k =1 ⎩
the white Gauss noise, it is assumed to be zero mean, the variance of normal distribution of
σ s2 ,H0 said
(6) test end So, the probability of false alarm for
only noise,H1 said the test end user information and noise, assumption on the noise sample values are independent
Pf = Qx（ 2 N
and identical distribution, and signal sampling values are independent of each other[4],in the H0 in the H1
x ~ N (0, σ w2 I )
2 s
summation, output statistics of Y can be expressed as
Y = ∑ x(n)2 N
If the given thresholds for
(2)
λ 2(σ + σ s2 ) 2 w
） (8)
Energy detection method for general use single threshold detection, when the decision statistic is greater than threshold decision signal exists, otherwise the decision signal does not exist. If the use of double
than λ0 , then the decision signal does no exist; between the decision statistics between
Γ（
© 2013 ACADEMY PUBLISHER
N
the decision signal exists; If the decision statistic is less
alarm probability can be expressed as
Pf =
Pd = Qx（ 2
threshold, if the decision statistic is greater than λ1 , then
λ , and the false N λ ， ） 2 2σ w2 N Γ（ ) 2
(7)
The probability of correct detection for ,
x ~ N [0, (σ + σ ) I ] also in N sampling , 2 w
λ ） 2σ w2
λ0
and
λ1 ( λ0 < λ1 )do
not make judgments and once again until the test decision success or reach the maximum detection times[5]. (3)
JOURNAL OF COMPUTERS, VOL. 8, NO. 10, OCTOBER 2013
Let each detection time is N, the maximum frequency
2567
of testing for Nmax ,energy detection method for detecting
non-coverage of the two regions, R0 and R1 ,they called
time adaptive specific steps are as follows: a) The detection sample square time N get sum of
the decision regions. When the observation r falls within
decision statistics T, set the threshold λ0 and λ1 .
the region R0 is determined that
H0
is correct(become
b) Comparison decision statistic and the two threshold size, if T
> λ1 that signals exist, other T < λ0 ,that signals
does not exist, and the detection process is finished, otherwise the step c. c) If the detection times to
Nmax
a select region
H 0 );other
the observation r falls within the
R1 is determined that H1 is correct(become a
without judgment of
success, also believe that there is no signal[6].Detection times not to A to step a. In this detection method, if the SNR is very high, then the detection of few times, even once testing is complete, but low SNR, to reduce the probability of success, to improve the detection performance by increasing the frequency of testing [7].
select H1 ).
III. PERFORMANCE ANALYSIS OF DOUBLE-THRESHOLD
under observation for the corresponding dual threshold
SPECTRUM SENSING
Order to facilitate the analysis of PH 01 , PH 02 , PH 03 in
H0
and
PH11 , PH12 , PH13 of H1 denoted segmentation
value falls more than λ1 ,
λ0 − λ1 ,or
less than λ0 ,
Analysis model of double threshold hypothesis testing for radio signal detecting system, this process is
each segment probability, 0 and 1 corresponds to two
represented by the diagram of a model. First, we
assumptions
H0
and
H1 . H0 under
the assumption
hypothesized that PU users have the signal generation, the input signal y(t),in the simple case, there are only two kinds of assumptions
H 0 or H
1
, decision output is
only one of the two assumptions.
Figure 2
Dual-threshold hypothesis testing model
is: The Observation value falls more than λ1 is the probability that is: PH = Pr{T > λ1 | H0 } = Q 2 ( χ 01
λ1 ) 2σ w2
The probability of observed values fall between
λ0
N
and
λ1 is : PH = Pr{T < λ0 | H 0 } = 1 − Qχ ( λ0 2 ) 2 N
02
2σ w
With respect to each hypothesis, we can get a view of measurement, it is a random variable generated according to certain rules of transmission. In general, with the probability of transmission machine to represent
The Observation value falls less than λ0 is the probability that is:
the process. Probability of transmission mechanism
PH03 =1− PH01 − PH02 = Qχ2 (
according to the known conditional probability density of
(
)
(
N
)
p r | H 0 and p r | H1 measurements r , r is a Under point observation space R.
λ0 λ ) −Qχ ( 12 ) 2 2σw 2σw 2 N
H0 false alarm probability can be expressed is:
Hypothesis testing wich the final problem to solve is :Judgment on the basis of the observables, determining which A and B two kinds of hypothesis is correct. This is
Nmax
Pf = PH01 (1 + ∑ PH03 n ) n =1
equivalent to the observation space R is divided into Average detection time of expression:
© 2013 ACADEMY PUBLISHER
(9)
2568
JOURNAL OF COMPUTERS, VOL. 8, NO. 10, OCTOBER 2013
Nf =
1 − PH 03 N max 1 − PH 03
(10)
It can be found that there was a direct relationship between the number of false alarm probability and the two threshold and detection, When we set these parameters, false alarm probability is constant, and by changing these parameters to achieve the desired
H1
probability of false alarm.
under the assumption is:
The Observation value falls more than λ1 is the probability
that
is: PH = Pr{T > λ1 | H 0 } = Q 2 ( χ 11 N
λ1
2(σ + σ s2 ) 2 w
)
detection probability with the curve of the signal-to-noise ratio under the same probability of false alarm and detection time conditions; figure 4 is a curve of the relationship of the number of signal-to-noise ratio of the adaptive energy detector for detecting. Little difference in SNR-5dB following sections two methods can be seen in FIG, the partially adaptive energy detection method will be better than the above-5dB energy detection method, this Figure 4 detects the frequency curve consistent,the -5dB section decreased very significantly increase the number of detections as the signal-to-noise ratio, Saving the detection time, this is a fixed detection time compared to energy detection method to improve the detection probability. Figure 4, with the signal-to-noise ratio to reduce the number of detection in low signal-to-noise ratio segment declined slightly, it makes detection and rapid judgment 0 aborted due to increase in the probability of undetected under the low signal-to-noise ratio.
The probability of observed values fall between λ0 and
λ1 is : PH = Pr{T < λ0 | H0} = 1− Qχ ( 2 N
12
λ0
2(σ + σ s2 ) 2 w
The Observation value falls less than
λ0
)
is the
probability that is:
PH13 = 1 − PH11 − PH12 = Qχ 2 ( N
λ0
2(σ + σ ) 2 w
2 s
) − Qχ 2 ( N
λ1
2(σ + σ s2 ) 2 w
)
Figure 3 Double threshold energy and single-threshold energy detection performance comparison
The probability of detection and the average number of times of detection can be expressed
1 − P13Nmax Pd = P11 1 − P13
Nd =
(11)
1 − PH13 Nmax 1 − PH13
(12)
IV. SIMULATION AND VERIFICATION In the process of conducting simulation, to be used to compare the detection method with the general energy, assuming that these two methods have the same probability of false alarm and detection time, next on the Spectrum sensing mechanism analysis. Figure 3 is a performance comparison of the detection method of adaptive energy detection method and the general energy,
© 2013 ACADEMY PUBLISHER
Figure 4 Average number of detections and signal-to-noise ratio of the simulation curve
JOURNAL OF COMPUTERS, VOL. 8, NO. 10, OCTOBER 2013
V. CONCLUSION This paper proposes a cooperative spectrum sensing algorithm based on energy-aware adaptive two-door limited, employing a plurality of cognitive, and comprehensive judgment after double threshold, the ultimate fusion of the of each SU judgment data center spectrum usage information, to improve perceived performance. Relative to the general energy-aware spectrum detection algorithm, this proposed algorithm fully consider the degree of reliability of the different perception of the user, if the received signal energy is between the two threshold values, then Adaptive threshold using multiple judgments strategy, generated verdict judgment sent to the center for data fusion. Finally, through MATLAB simulation signal-to-noise ratio and the probability of detection, the relationship between the number of detections, simulation results show that the algorithm compared with the general energy spectrum detection algorithm, the valid section detection time, improve the detection probability, significantly improve spectrum perceived performance.
2569
Distributed Opportunistic Spectrum Networks[J],Journal of Communications, 2009,Vol 4(10), pp.728-740 [7] Xu Jianxia LiuHuiHeng, LiuKeZhong. Cognitive radio algorithm is a kind of double threshold [J]. Journal of wuhan university of technology and management engineering (information). 2011,Vol.7(04),pp.134-138 [8] Ding Hanqing Yang Guwei, Zhao Zhiyuan. Based on adaptive double threshold based on cooperative spectrum sensing[J]. Journal of jilin university (engineering science).2011(02),pp.77-81
Xiaolin Jiang received the B.E. degree in Heilongjiang University of science and technology ,Harbin, China, in 2002 and the M.E. degree in communication and information system from Harbin Institute of Technology, China, in 2008. He is currently working toward the Ph.D. degree with the Department of Electronic Engineering, Harbin Institute of Technology,, Harbin, China. His research interests include the areas of wireless communication and signal processing, including cognitive radios and statistical signal processing.
REFERENCE [1] Zhu Qing Song Chunlin. Based on the fusion rules of double threshold cooperative spectrum sensing algorithm[J]. Journal of computer applications,2011,Vol.31(8)pp.2040-2043 [2] Owayed Abdullah Alghamdi,,Optimal and Suboptimal Multi Antenna Spectrum Sensing Techniques with Master Node Cooperation for Cognitive Radio Systems[J],Journal of Communications, 2011,Vol.7(6),pp.512-523 [3] ZhangLei Huang Guangming. Based on the credibility of cognitive radio collaborative spectrum detection [J]. Journal of computer applications, 2010, Vol.30( 9) ,pp.2519-2521． [4] Zhu Jiang Huang Benxiong, furong wang, et al. A new type of cooperative spectrum sensing in cognitive radio network method [J]. Small microcomputer system,2010,Vol.31( 2) ,pp.193-197． [5] Robert C. Qiu, Changchun Zhang, Towards A Large-Scale Cognitive Radio Network Testbed: Spectrum Sensing, System Architecture, and Distributed Sensing[J],Journal of Communications, 2012,Vol 7(7) , pp.552-566 [6] Ammar Alshamrani, Xuemin Shen. A Cooperative MAC with Efficient Spectrum Sensing Algorithm for
© 2013 ACADEMY PUBLISHER
Xuemai Gu received the B.E., M.E. and Ph.D. degrees in Harbin Institute of Technology, Harbin, China, in 1982 , 1985 and 1987 respectively, and then he has been with the Department of Electronic information and communication Engineering, Harbin Institute of Technology, Harbin, China, where he is currently a Professor and the dean of the school of electronics and information engineering, Harbin Institute of Technology, Harbin, China,. His current research interests include Data communication systems, multiple access communication network technology, satellite communication system simulation design, personal communication system and satellite tt&c and communication system, etc. Yingpei Lin received the B.E. degree in electronic information science and technology from Yantai University, Yantai, China, in 2003 and the M.E. degree in communication and information system from Nanjing Institute of Electronic Technology, Nanjing, China, in 2007. And the Ph.D. degree in the Department of Electronic Engineering, Shanghai Jiao Tong University, Shanghai, China,in 2012. His research interests include the areas of wireless communication and signal processing, including cognitive radios and statistical signal processing.
2565
Adaptive Double-threshold Joint Spectrum Sensing Based on Energy Detection XiaoLin Jiang School of Electronics and Information Engineering,Harbin Institute of Technology, Harbin,,china School of Electronics and Information Engineering,Heilongjiang University of science and technology, harbin, china Email: [email protected]
Xuemai Gu School of Electronics and Information Engineering,Heilongjiang University of science and technology, harbin,china Email:[email protected]
Yingpei Lin Department of Electronic Engineering, Shanghai Jiao Tong University, Shanghai, China Email:[email protected]
Abstract—Energy sense of adaptive double threshold joint spectrum perception algorithm using multiple cognitive, and after double threshold for comprehensive judgment, and threshold value with the discretion of the signal to noise ratio (SNR) adaptive adjustment, will eventually each SU decision data fusion center for spectrum using the information, make perception performance effectively improve. In this paper, the detection probability of the proposed algorithm, the virtual detection probability and frequency of testing performance expression in this paper, and by using the MATLAB simulation get signal-to-noise ratio and detection probability, the relationship between the frequency of testing. The results show that, with the traditional single threshold energy spectrum method under the perception, compared with less effective section the detection time, improving the detection probability, can significantly improve the cognitive wireless network spectrum perception performance. Index Terms—Double threshold; Average frequency of testing
Energy
detection;
I. INTRODUCTION Along with the growing demand of wireless communication, the demand os wireless spectrum communication system is also a corresponding increasing, leading to increasingly scarce spectrum resources, becoming a new bottleneck restricting the development of wireless communication. However, assigned to the existing spectrum resources in many wireless systems in time and space, there are different degrees of idle, serious lack of use[1].So people begin to seek permission without spectrum licensing users of in all legal users do Project supported by the National Nature Science Foundation of China (NSFC) under Grant No. 60832009,Heilongjiang University of Science and Technology talented persons training project. Project supported by HarBin Science and Technology Bureau under Grant No. 2013RFQXJ107. Manuscript received March 13, 2013; revised April 14, 2013; accepted April 16,2013;Corresponding author is Xumai Gu.
© 2013 ACADEMY PUBLISHER doi:10.4304/jcp.8.10.2565-2569
not have any impact on the case, and find the chance to access the authorized band internal and communication, in order to effectively improve the utilization of the limited spectrum, this idea is Cognitive Radio(CR). Spectrum. Sensing refers to the specific time and space which was detected on the utilization of spectrum resources, continuously, effectively detect whether the use of the band is authorized users, so the unused spectrum was identified. Spectrum sensing is one of the key technologies of cognitive radio system[2],the performance perceived will have a direct relationship with the cognitive wireless network. Spectrum sensing energy detection method does not need to know the main user prior knowledge, it is simple and easy to implement, but because of the influence of vulnerable to channel multipath fading, shadowing effect interference, leading to correct detection probability versus SNR decreased and decreased rapidly, in low signal to noise ratio of spectrum sensing performance is poor. Based on energy detection adaptive dual energy detection limit of cooperative spectrum sensing algorithm using multiple cognitive after double threshold decision, and the threshold value adjustment with high and low signal to noise ratio adaptive, eventually every SU decision data center is obtained after spectral information fusion, completely the sentence whether it is the primary user signal, make improve perceived performance effectively.
II.ENERGY DETECTION METHOD The detection process energy detection method can be integrated into a binary hypothesis, Block diagram is shown in Figure 1.
2566
JOURNAL OF COMPUTERS, VOL. 8, NO. 10, OCTOBER 2013
In the formula, Γ（.,.) says the incomplete gamma function, define it as ∞
Γ（a,b)= ∫ t a −1 exp(−t )dt b
Figure 1
Similarly, we can get the correct detection
Energy detection principle
probability. In the H1, the correct detection probability is Directly filtering of signal sample values, modulo,
Γ（
then square can be obtained, or do N point FFT of the
Pd =
received signal, converted to the frequency domain, then the frequency domain signal modulus square also can be. Energy detection principle is the energy accumulation in a certain frequency range, if the accumulated energy
(4)
If you use the right tail probability distributions to represent it,and
values above a certain threshold, that signals exist, If less
∞
Qx 2 = ∫ p (t )dt , t ≥ 0
than a certain threshold value, the only noise is present. It
N
t
is the starting point of signal plus noise energy larger than the energy of noise[3].
N λ ， 2 ） 2 2(σ w + σ s2 ) N Γ（ ) 2
(5)
and Qx（ x）= 2
Supposing:
N
⎧ H 0 , x ( n ) = w( n ) , n = 1, 2,... N ⎨ ⎩ H 1 , x ( n ) = s ( n ) + w( n ) (1) In the formula, s ( n ) is the main input of user information, is assumed to be zero mean, the variance of normal distribution of
σ s2 , w ( n ) is noise information, if
⎧ ⎪2Q （ x） N = 1 ⎪ 1 1 ⎪ k− exp(− x) ( N −1）/2 (k − 1)!(2 x) 2 ⎪ 2 , N is odd （ x) + ⎨2Q ∑ (2k − 1)! π k =1 ⎪ 1 ⎪ k− ⎪ 1 ( N −1）/2 ( x / 2) 2 , N is even （ ⎪ − 2 x） ∑ k! k =1 ⎩
the white Gauss noise, it is assumed to be zero mean, the variance of normal distribution of
σ s2 ,H0 said
(6) test end So, the probability of false alarm for
only noise,H1 said the test end user information and noise, assumption on the noise sample values are independent
Pf = Qx（ 2 N
and identical distribution, and signal sampling values are independent of each other[4],in the H0 in the H1
x ~ N (0, σ w2 I )
2 s
summation, output statistics of Y can be expressed as
Y = ∑ x(n)2 N
If the given thresholds for
(2)
λ 2(σ + σ s2 ) 2 w
） (8)
Energy detection method for general use single threshold detection, when the decision statistic is greater than threshold decision signal exists, otherwise the decision signal does not exist. If the use of double
than λ0 , then the decision signal does no exist; between the decision statistics between
Γ（
© 2013 ACADEMY PUBLISHER
N
the decision signal exists; If the decision statistic is less
alarm probability can be expressed as
Pf =
Pd = Qx（ 2
threshold, if the decision statistic is greater than λ1 , then
λ , and the false N λ ， ） 2 2σ w2 N Γ（ ) 2
(7)
The probability of correct detection for ,
x ~ N [0, (σ + σ ) I ] also in N sampling , 2 w
λ ） 2σ w2
λ0
and
λ1 ( λ0 < λ1 )do
not make judgments and once again until the test decision success or reach the maximum detection times[5]. (3)
JOURNAL OF COMPUTERS, VOL. 8, NO. 10, OCTOBER 2013
Let each detection time is N, the maximum frequency
2567
of testing for Nmax ,energy detection method for detecting
non-coverage of the two regions, R0 and R1 ,they called
time adaptive specific steps are as follows: a) The detection sample square time N get sum of
the decision regions. When the observation r falls within
decision statistics T, set the threshold λ0 and λ1 .
the region R0 is determined that
H0
is correct(become
b) Comparison decision statistic and the two threshold size, if T
> λ1 that signals exist, other T < λ0 ,that signals
does not exist, and the detection process is finished, otherwise the step c. c) If the detection times to
Nmax
a select region
H 0 );other
the observation r falls within the
R1 is determined that H1 is correct(become a
without judgment of
success, also believe that there is no signal[6].Detection times not to A to step a. In this detection method, if the SNR is very high, then the detection of few times, even once testing is complete, but low SNR, to reduce the probability of success, to improve the detection performance by increasing the frequency of testing [7].
select H1 ).
III. PERFORMANCE ANALYSIS OF DOUBLE-THRESHOLD
under observation for the corresponding dual threshold
SPECTRUM SENSING
Order to facilitate the analysis of PH 01 , PH 02 , PH 03 in
H0
and
PH11 , PH12 , PH13 of H1 denoted segmentation
value falls more than λ1 ,
λ0 − λ1 ,or
less than λ0 ,
Analysis model of double threshold hypothesis testing for radio signal detecting system, this process is
each segment probability, 0 and 1 corresponds to two
represented by the diagram of a model. First, we
assumptions
H0
and
H1 . H0 under
the assumption
hypothesized that PU users have the signal generation, the input signal y(t),in the simple case, there are only two kinds of assumptions
H 0 or H
1
, decision output is
only one of the two assumptions.
Figure 2
Dual-threshold hypothesis testing model
is: The Observation value falls more than λ1 is the probability that is: PH = Pr{T > λ1 | H0 } = Q 2 ( χ 01
λ1 ) 2σ w2
The probability of observed values fall between
λ0
N
and
λ1 is : PH = Pr{T < λ0 | H 0 } = 1 − Qχ ( λ0 2 ) 2 N
02
2σ w
With respect to each hypothesis, we can get a view of measurement, it is a random variable generated according to certain rules of transmission. In general, with the probability of transmission machine to represent
The Observation value falls less than λ0 is the probability that is:
the process. Probability of transmission mechanism
PH03 =1− PH01 − PH02 = Qχ2 (
according to the known conditional probability density of
(
)
(
N
)
p r | H 0 and p r | H1 measurements r , r is a Under point observation space R.
λ0 λ ) −Qχ ( 12 ) 2 2σw 2σw 2 N
H0 false alarm probability can be expressed is:
Hypothesis testing wich the final problem to solve is :Judgment on the basis of the observables, determining which A and B two kinds of hypothesis is correct. This is
Nmax
Pf = PH01 (1 + ∑ PH03 n ) n =1
equivalent to the observation space R is divided into Average detection time of expression:
© 2013 ACADEMY PUBLISHER
(9)
2568
JOURNAL OF COMPUTERS, VOL. 8, NO. 10, OCTOBER 2013
Nf =
1 − PH 03 N max 1 − PH 03
(10)
It can be found that there was a direct relationship between the number of false alarm probability and the two threshold and detection, When we set these parameters, false alarm probability is constant, and by changing these parameters to achieve the desired
H1
probability of false alarm.
under the assumption is:
The Observation value falls more than λ1 is the probability
that
is: PH = Pr{T > λ1 | H 0 } = Q 2 ( χ 11 N
λ1
2(σ + σ s2 ) 2 w
)
detection probability with the curve of the signal-to-noise ratio under the same probability of false alarm and detection time conditions; figure 4 is a curve of the relationship of the number of signal-to-noise ratio of the adaptive energy detector for detecting. Little difference in SNR-5dB following sections two methods can be seen in FIG, the partially adaptive energy detection method will be better than the above-5dB energy detection method, this Figure 4 detects the frequency curve consistent,the -5dB section decreased very significantly increase the number of detections as the signal-to-noise ratio, Saving the detection time, this is a fixed detection time compared to energy detection method to improve the detection probability. Figure 4, with the signal-to-noise ratio to reduce the number of detection in low signal-to-noise ratio segment declined slightly, it makes detection and rapid judgment 0 aborted due to increase in the probability of undetected under the low signal-to-noise ratio.
The probability of observed values fall between λ0 and
λ1 is : PH = Pr{T < λ0 | H0} = 1− Qχ ( 2 N
12
λ0
2(σ + σ s2 ) 2 w
The Observation value falls less than
λ0
)
is the
probability that is:
PH13 = 1 − PH11 − PH12 = Qχ 2 ( N
λ0
2(σ + σ ) 2 w
2 s
) − Qχ 2 ( N
λ1
2(σ + σ s2 ) 2 w
)
Figure 3 Double threshold energy and single-threshold energy detection performance comparison
The probability of detection and the average number of times of detection can be expressed
1 − P13Nmax Pd = P11 1 − P13
Nd =
(11)
1 − PH13 Nmax 1 − PH13
(12)
IV. SIMULATION AND VERIFICATION In the process of conducting simulation, to be used to compare the detection method with the general energy, assuming that these two methods have the same probability of false alarm and detection time, next on the Spectrum sensing mechanism analysis. Figure 3 is a performance comparison of the detection method of adaptive energy detection method and the general energy,
© 2013 ACADEMY PUBLISHER
Figure 4 Average number of detections and signal-to-noise ratio of the simulation curve
JOURNAL OF COMPUTERS, VOL. 8, NO. 10, OCTOBER 2013
V. CONCLUSION This paper proposes a cooperative spectrum sensing algorithm based on energy-aware adaptive two-door limited, employing a plurality of cognitive, and comprehensive judgment after double threshold, the ultimate fusion of the of each SU judgment data center spectrum usage information, to improve perceived performance. Relative to the general energy-aware spectrum detection algorithm, this proposed algorithm fully consider the degree of reliability of the different perception of the user, if the received signal energy is between the two threshold values, then Adaptive threshold using multiple judgments strategy, generated verdict judgment sent to the center for data fusion. Finally, through MATLAB simulation signal-to-noise ratio and the probability of detection, the relationship between the number of detections, simulation results show that the algorithm compared with the general energy spectrum detection algorithm, the valid section detection time, improve the detection probability, significantly improve spectrum perceived performance.
2569
Distributed Opportunistic Spectrum Networks[J],Journal of Communications, 2009,Vol 4(10), pp.728-740 [7] Xu Jianxia LiuHuiHeng, LiuKeZhong. Cognitive radio algorithm is a kind of double threshold [J]. Journal of wuhan university of technology and management engineering (information). 2011,Vol.7(04),pp.134-138 [8] Ding Hanqing Yang Guwei, Zhao Zhiyuan. Based on adaptive double threshold based on cooperative spectrum sensing[J]. Journal of jilin university (engineering science).2011(02),pp.77-81
Xiaolin Jiang received the B.E. degree in Heilongjiang University of science and technology ,Harbin, China, in 2002 and the M.E. degree in communication and information system from Harbin Institute of Technology, China, in 2008. He is currently working toward the Ph.D. degree with the Department of Electronic Engineering, Harbin Institute of Technology,, Harbin, China. His research interests include the areas of wireless communication and signal processing, including cognitive radios and statistical signal processing.
REFERENCE [1] Zhu Qing Song Chunlin. Based on the fusion rules of double threshold cooperative spectrum sensing algorithm[J]. Journal of computer applications,2011,Vol.31(8)pp.2040-2043 [2] Owayed Abdullah Alghamdi,,Optimal and Suboptimal Multi Antenna Spectrum Sensing Techniques with Master Node Cooperation for Cognitive Radio Systems[J],Journal of Communications, 2011,Vol.7(6),pp.512-523 [3] ZhangLei Huang Guangming. Based on the credibility of cognitive radio collaborative spectrum detection [J]. Journal of computer applications, 2010, Vol.30( 9) ,pp.2519-2521． [4] Zhu Jiang Huang Benxiong, furong wang, et al. A new type of cooperative spectrum sensing in cognitive radio network method [J]. Small microcomputer system,2010,Vol.31( 2) ,pp.193-197． [5] Robert C. Qiu, Changchun Zhang, Towards A Large-Scale Cognitive Radio Network Testbed: Spectrum Sensing, System Architecture, and Distributed Sensing[J],Journal of Communications, 2012,Vol 7(7) , pp.552-566 [6] Ammar Alshamrani, Xuemin Shen. A Cooperative MAC with Efficient Spectrum Sensing Algorithm for
© 2013 ACADEMY PUBLISHER
Xuemai Gu received the B.E., M.E. and Ph.D. degrees in Harbin Institute of Technology, Harbin, China, in 1982 , 1985 and 1987 respectively, and then he has been with the Department of Electronic information and communication Engineering, Harbin Institute of Technology, Harbin, China, where he is currently a Professor and the dean of the school of electronics and information engineering, Harbin Institute of Technology, Harbin, China,. His current research interests include Data communication systems, multiple access communication network technology, satellite communication system simulation design, personal communication system and satellite tt&c and communication system, etc. Yingpei Lin received the B.E. degree in electronic information science and technology from Yantai University, Yantai, China, in 2003 and the M.E. degree in communication and information system from Nanjing Institute of Electronic Technology, Nanjing, China, in 2007. And the Ph.D. degree in the Department of Electronic Engineering, Shanghai Jiao Tong University, Shanghai, China,in 2012. His research interests include the areas of wireless communication and signal processing, including cognitive radios and statistical signal processing.
