CN103117817B - A kind of frequency spectrum detecting method under time-varying fading channels - Google Patents
A kind of frequency spectrum detecting method under time-varying fading channels Download PDFInfo
- Publication number
- CN103117817B CN103117817B CN201310007794.XA CN201310007794A CN103117817B CN 103117817 B CN103117817 B CN 103117817B CN 201310007794 A CN201310007794 A CN 201310007794A CN 103117817 B CN103117817 B CN 103117817B
- Authority
- CN
- China
- Prior art keywords
- channel
- time
- state
- varying
- current time
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Mobile Radio Communication Systems (AREA)
- Monitoring And Testing Of Transmission In General (AREA)
Abstract
The present invention proposes a kind of lower frequency spectrum perception algorithm of time-varying fading channels, a kind of dynamical state space model is designed to describe the time-varying characteristics of authorized user's state and time varying channel gain, regard authorized user's working condition and fading channel gain as two hiding system modes, and introduce single order FSMC models and portray time-varying slow fading channel gain time-varying migration characteristic, using reception signal energy accumulation and as systematic perspective measured value；On this basis, a kind of brand-new frequency spectrum sensing method is proposed for time-varying fading channels, fully excavate authorized user's state prior probability and fading channel conditions transfer characteristic, Combined estimator is implemented to authorized user's state and time varying channel gain, greatly improve frequency spectrum perception performance under fading channel, in the case where that need not implement complicated multi-node collaboration and perceive, good detection performance (see photo) can be also obtained.For the program using accumulated energy as observed quantity, new method also remains the advantage short without authorization user signal feature and detection time simultaneously.
Description
Technical field
The present invention is directed to frequency spectrum detection problem under time-varying fading channels, and design proposes a kind of dynamical state space system model
(Dynamic State-space Model, DSM), it is based respectively on two state Markov state probability metastasis models and single order
Finite state Markov channel (Finite-state Markov Channel, FSMC) model, by authorized user (also referred to as
Primary user) state and time varying channel states are as hidden state (Hidden States)；On this basis, a kind of joint is proposed
Estimate channel gain and the new frequency spectrum sensing method of authorized user's state, and utilize particle filter (Particle
Filtering, PF) posterior probability involved in technology estimation non-stationary system, it is achieved thereby that for time-varying fading channels
Gain and the Combined estimator of primary user's state.Belong to the communications field.
Background technology
With the fast development of Modern wireless communication technology and being continuously increased for user's request, Radio Transmission Technology is
Constantly strided forward towards the high code check direction in big broadband.In today that frequency spectrum resource situation in short supply increasingly highlights, wireless system designer
Take up the availability of frequency spectrum that lifting has authorized frequency range, it is intended to realize a kind of more flexible, efficient radio spectrum resources pipe
Operational version is managed, so as to provide prerequisite condition for the development of high-speed radiocommunication business of future generation.In the promotion of this background
Under, cognitive radio technology (Cognitive Radio, CR) arises at the historic moment, turn into be expected to alleviate at present frequency spectrum resource it is exhausted with
One of new technology of most application prospect of this relatively low principal contradiction of its utilization rate.CR detects wireless frequency spectrum by real-time perception
Behaviour in service (i.e. whether primary user uses frequency range), when authorizing frequency range dereliction user job, unauthorized user (also referred to as time
Want user) by opportunistic dynamic access and use the idle frequency spectrum.Therefore, frequency spectrum perception (Spectrum Sensing) will
It is the key component in cognitive radio technology.
Frequency spectrum perception is substantially distinguished from signal detection and receive process in general communication control processor, and it is without accurate recovery
Received signal, and only need to detect some special frequency channel on specific geographical area and period, if it there are primary user's letter
Number.At present, existing spectrum sensing scheme can be divided into three kinds of detection modes, i.e., energy detector (Energy Detection,
ED), matched filter (Matched Filter Detection, MFD) and cyclostationary characteristic detection
(Cyclostationary Detection).Compared to other two methods, case of energy detection schemes because its computation complexity is low,
Without authorization signal prior information, and many advantages such as detection time are short, turn into the spectrum sensing scheme that is widely used at present it
One.Unfortunately, energy measuring is easily influenceed by channel status, specifically, decision threshold and reception in ED
The energy of signal is closely related, thus in time-varying fading channels, the characteristic that channel gain changes over time undoubtedly will greatly
Increase detection difficulty, and significantly reduce the frequency spectrum detection performance in practical application.Although existing research has contemplated that Time-varying fading is believed
The statistical probability distribution density feature in road, but the statistical probability distribution is only capable of the instantaneous accidental characteristic of reflection fading channel, not
Its property with time migration and variation can be described, thus is also difficult to time varying channel is described and followed the trail of, causes ED detections property
Can be not good enough.
The present invention proposes the dynamical state space model under a kind of time-varying fading channels, to deep description primary user's state
With the time-varying migration characteristic of time varying channel gain the two hidden states；On this basis, design proposes to believe for Time-varying fading
A kind of brand-new frequency spectrum sensing method in road, the observation signal using accumulated signal energy as DSM, and fully excavate authorized user's work
Make the prior probability of state and the state transfer characteristic of fading channel gain, primary user's state and time varying channel gain are implemented
Combined estimator (Joint Estimation), greatly improves the frequency spectrum perception performance in actual fading channel, without implementing
In the case that complicated multi-node collaboration perceives, good detection performance can be also obtained.Due to the program using accumulated energy be used as see
Measurement, new method also remain the advantage short without primary user's signal characteristic and detection time while improving performance.
The content of the invention
The present invention proposes a kind of new frequency spectrum detection model for time-varying fading channels, by authorized user's working condition and declines
Fall channel gain and regard two hiding system modes as, and introduce single order FSMC models and increase effectively to portray time-varying slow fading channel
Benefit with time-shift characteristic, while using energy accumulation of the authorization user signal received in special time window and as being
The observation of system；Based on newly-established DSM, further design proposes a kind of brand-new frequency spectrum sensing method, fully discovers and uses and award
The time-varying characteristics of user working status and the state transition information of fading channel gain are weighed, based on maximum posteriori criterion
(Maximum A Posterior Probability, MAP) and stochastical sampling theory (Random Sampling) particle filter
Technology (Particle Filtering, PF), while estimate the posterior probability of time varying channel gain and authorized user's working condition,
And then realize Combined estimator and the real-time tracking to primary user's state and channel gain.New departure is ensureing that perception algorithm is low multiple
On the premise of miscellaneous degree and requirement of real-time, the frequency spectrum perception performance in time-varying fading channels is greatly improved, so as to be distribution
The design of cognition wireless network provides a kind of scheme of great application potential with realizing.
The present invention uses following technical scheme.
First, this method establishes equivalent information sequence (the Equivalent state of primary user's working condition
Sequence), characterized using 0,1 sequence of tool single order Markov property, namely " 0 " represents that authorized user is in silent shape
State, i.e., the frequency range is idle this moment；" 1 " represents that authorized user is in transmitting working condition, i.e., the frequency range is occupied this moment.Together
When, this method characterizes time-varying slow fading Rayleigh channel using single order FSMC models, and channel gain is divided into some discrete shapes
State, between state transfer determined by Probability State transfer matrix (Probability Transition Matrix, PTM).Specifically
Ground, the state transition probability are determined and provided by single order Markov Chain characteristic.
Secondly, this method establishes a dynamical state space system model, for perceiving user, by channel status and
Authorized user's state will sample the signal energy of reception and as observation as hidden state in certain time window；
Finally, our subtraction unit realizes the connection for authorized user's working condition and time varying channel states based on observation
Estimation is closed, using the Markov state transfer characteristic of channel gain, channel gain amplitude is estimated in real time；It is basic herein
On, sequential renewal and estimation are carried out to the posterior probability of authorized user's working condition using particle filter technology, so as to final real
Existing channel gain and the Combined estimator of authorized user's state.
It is an advantage of the invention that：
1) technical solution of the present invention is applied to the frequency spectrum perception under complicated time-varying wireless transmission environments, is cognitive radio skill
Art frequency spectrum perception provides a kind of brand-new theory, and establishes solid foundation for its practical application；
2) present invention proposes a kind of dynamical state space model of frequency spectrum perception system for time-varying fading channels, can be more
Effectively to reflect frequency spectrum perception inherent mechanism；
3) it is different from existing frequency spectrum sensing method for the static probability distribution modeling method of fading channel, new departure energy
Channel time-varying characteristics are fully demonstrated, thus more meet cognitive radio technology actual working environment, designed method also can be more
To be effectively applied to reality；
4) existing scheme is different from just for primary user's state to be estimated and ignore the limitation of time varying channel gain,
New departure carries out combining real-time estimation for channel gain and authorized user's state first, thus greatly improves Time-varying fading
Frequency spectrum perception performance under channel；Meanwhile based on case of energy detection schemes, new frequency spectrum aware scheme possesses low implementation complexity
The short advantage with detection time；
5) invention makes full use of the prior information of time varying channel and authorized user's working condition, and uses particle filter skill
Art effectively overcomes the non-stationary non-Gaussian feature that observation signal (cumlative energy) shows, and greatly improves time-varying fading channels
Under frequency spectrum perception performance.
Brief description of the drawings
Fig. 1 is the fading channel conditions block diagram based on FSMC models.
Fig. 2 carries out frequency spectrum perception flow chart to perceive user.
Fig. 3 is that new method frequency spectrum perception detects accuracy and traditional ED performance simulations comparison diagram.
Embodiment
The present invention sets up the frequency spectrum perception dynamical state space model under time-varying fading channels, while uses maximum a posteriori
Canon of probability carries out Combined estimator to time varying channel gain and primary user's state, and realizes that non-stationary is non-using particle filter technology
Parameter Estimation in the case of Gauss.Dynamic system model and frequency spectrum perception process are illustrated respectively below.
1. under the time-varying fading channels that the present invention establishes shown in frequency spectrum perception dynamical state space model such as formula (1)：
In formula, xnPrimary user's state at n-th of moment is represented, is changed according to certain state transition function f ()；
anThe channel fading amplitude gain at n-th of moment is represented, is also updated according to particular state transfer function h (), this method
By authorized user's state x in designnWith channel gain state anAs two hidden states (Hidden states), for cognition
For user, its observation signal ynSampled signal energy in as specific observation time window and, as shown in formula (2)
Wherein, M=Ts×fspRepresent to perceive cycle TsInterior sampling number, fspFor sample frequency.H0And H1Assume inspection for two
Test, correspond respectively to primary user's signal be present and in the absence of primary user's signal.The main target of frequency spectrum perception is seen by receiving
Survey signal energy value to judge hypothesis testing, and finally determine to whether there is primary user's signal in the current detection cycle.
1) for observation signal yn, when in the absence of authorization user signal, i.e. xnFor " 0 " when, reception signal ynObey the free degree
For M center chi square distribution, yn~χM 2；When authorization user signal, i.e. x being presentnFor " 1 " when, reception signal takes ynIt is from the free degree
M non-central chi square distribution yn~χM 2(κ), wherein non-centrality parameter κ=M (anxn)2。
2) for the sake of for convenience of analyzing, authorized user's status switch is further abstracted as the binary system that amplitude is " 0 " or " 1 "
Sequence of symhols.In practice, authorized user's status switch carries out state transfer according to single order Markov Chain.Correspondingly, Ma Er
Section husband is represented by：
p(xn=j | xn-1=i)=ΠijI, j=1,2 (3)
Wherein, transfering state probability matrix can be further represented as：
In formula, parameter μ and λ are determined by specific applied business.
3) in analyzing, it is assumed that time-varying fading channels obey Rayleigh slow fading characteristic, the i.e. probability distribution of accidental channel gain
For rayleigh distributed.Channel gain is divided into K discrete state, K >=3.It is further assumed that in the case of slow fading, channel magnitude
Durations TaFrequency spectrum detection period T will be much larger thans, while TaAnd the integral multiple in primary user's status switch cycle；Further profit
The single order Markov switching feature showed with time-varying channel gain, namely channel status can only be to itself or adjacent shapes
State carries out saltus step, then has：
p(an=j | an-1=i)=pijI, j=0 ..., K-1 (4)
In practice, transition probability pijBy the number, the time varying channel rapidity of fading and channel gain of discrete equivalent channel state
Probability distribution combines decision.
2. being based on above-mentioned dynamical state space model, the present invention is further to two implicit system state (i.e. authorized users
State and channel gain) carry out Combined estimator.Estimation detection algorithm depends on maximum a posteriori probability and sequential detection thought,
Mainly include following three steps：(a) pre- judgement, (b) channel gain estimation and renewal and (c) based on maximum a posteriori probability
It is as shown in Figure 2 based on primary user's state estimation of sequential detection particle filter, corresponding scheme implementation process.Below for above-mentioned three
Individual part describes in detail：
1) pre- judgement
Adjudicate in advance mainly roughly to estimate whether include primary user's signal, so as to for different situations design subsequent
Algorithm for estimating, realize the estimation to channel gain and authorized user's state.It should be noted that it is different from traditional frequency spectrum perception
Pre- judgement, it is mainly that subsequent estimation provides a condition judgment mechanism that this programme, which uses pre- judgement, although pre- court verdict
Error probability is higher, but subsequent algorithm can further be corrected to this result, so as to finally lift the entirety of frequency spectrum perception
Performance.
Specifically, adjudicate in advance by contrasting the size between predetermined threshold value γ and reception signal energy, it is final to determine
Pre- judgement output result on primary user's state.If yn<γ, then pre- court verdict xn' it is " 0 ", namely find to be somebody's turn to do through pre- judgement
Moment frequency range is idle；If yn>=γ, then pre- court verdict xn' it is " 1 ", expression obtains the moment frequency range through pre- judgement and is authorized to use
Family takes.In practice, the predetermined threshold value is related to sampling number and minimum fading channel amplitude.
2) estimation of channel gain
By means of dynamical state space model and pre- court verdict xn', this programme is by real-time update channel gain.Examine simultaneously
Considering time varying channel has slow fading characteristic, i.e. its coherence time TaMuch larger than frequency spectrum detection period Ts, for the sake of easy analysis,
Further simplification is assumed to be Ta=LTs, wherein L is the integer more than 1
(a) as pre- court verdict xn' when being " 1 ", can be further by all identical anticipations within a channel magnitude cycle
Certainly result (xn'=1) LnIndividual detection cycle observation (and free degree) is accumulated (Ln≤ L), cumulative observations amount is respectivelyAnd Mn=LnM is represented.So, with accumulation degree LnIt is continuously increased, algorithm for estimating can be in same channel
Constantly channel magnitude estimate is modified in the amplitude cycle, so as to improve accuracy of estimation.It should be noted that accumulation degree
Ln0 is set to again at the time of channel gain changes；Afterwards, increase will be believed with detection cycle and will be constantly updated, that is, will be had：
On this basis, the fading channel amplitude at current time can be estimated based on Maximize criterion, such as
Shown in formula (6)：
Wherein, xn'=1 represents that pre- court verdict is 1, p (an|an-1) represent that the transfer of the n-th moment fading channel amplitude is general
Rate, its value are determined by FSMC channel properties and state-transition matrix.p(Yn|xn'=1, an) it is then likelihood function, its value depends on
Observational equation and observation noise.Under conditions of pre- court verdict is " 1 ", cumulative observations value YnNon-central chi square distribution is obeyed,
Have：
Wherein, σ2For Gaussian noise variance, λ is referred to as noncentral distribution parameter,mi=an。IMn’/2-1(x) it is
First kind Mn/ 2-1 rank modified Bessel functions.
(b) when pre- court verdict is " 0 ", it is necessary to further be divided to two class situations to discuss.If current time is channel
The amplitude jump moment, then utilize channel gain prior probability p (an|an-1) the channel gain estimate of renewal this moment, that is, haveIf current time is the channel magnitude non-toggle moment, the channel magnitude of previous moment is directly utilized
Estimate then has as channel magnitude this moment
3) primary user's state sequence is estimated
In view of conversion of the channel gain state transfer not with primary user's working condition in dynamical state space model is protected
Synchronization is held, thus is inherently seen, dynamical state space model shows non-stationary property at cognitive user end；At the same time,
Operated in observation for nonlinear transformation；This non-stationary, non-Gaussian feature are undoubtedly by for the sequential detection band of main User Status
Carry out serious challenge.Particle filter technology based on sequential importance sampling can then tackle such practical challenges, then to primary user
Working condition carries out Sequential Estimation.Particle filter mainly carries weight w using one group(i)Discrete particle x(i)To approach complexity
Posterior distrbutionp p (x), that is, there is p (x)=∑iw(i)δ(x-x(i)).Wherein, discrete particle x(i)And its weight w(i)Will be with new observation
Value carries out sequential renewal.In the specific implementation, particle filter mainly includes following four step：(a) particle, (b) foundation are generated
Likelihood function more new particle weights of importance, (c) resampling and (d) are according to particle and respective weights estimation primary user's state.
A) it is to be directed to a specific distribution sampling process to generate on particle essential, namelyThe distribution is different from Posterior distrbutionp probability, also known as importance function.We
Optimal importance function is used in case, as shown in formula (8)
In formula,For the vector formed from initial 1st moment to (n-1) moment particle.According to primary user's state transition
During memoryless characteristic, further have
B) according to the likelihood function newly obtained, the weight of new particle caused by renewal, such as formula (9)
When particle is 0, the likelihood function meets center chi square distribution, its probability distribution such as formula (10)：
When particle is 1, the likelihood function meets non-central chi square distribution, shown in its probability distribution such as formula (11)：
Wherein, M represents unit element number of samples,For noncentral distribution parameter.
Although c) weight more new algorithm provides a kind of recursive algorithm of great application potential for the estimation of unknown state,
There is also sample degeneracy problem during being somebody's turn to do.That is, after recursion iteration several times, most of particle weights of importance will
Tend to 0, so as to cause to estimate the decline of performance.An effective scheme for overcoming sample degeneracy is resampling.
Resampling main thought is, once occur obvious degradation phenomena (such asLess than certain
Individual threshold value), then the low particle of weights is further eliminated on the basis of importance sampling, and replicate and retain the high particle of weights, with
This reaches the phenomenon that particle suppresses to degenerate.By resampling, a new particle set will be finally produced, the new grain in the set
Son comes from an independent identically distributed sample setAnd each particle weights are 1/P, wherein P is total number of particles.
D) by particle and its weight can approximate complicated Posterior probability distribution, as shown in formula (12)
Similarly, have
The approximate posterior probability obtained by above formula, the estimate of primary user's state can be obtained.I.e. approximate posterior probability is larger
State be primary user estimated state.
The above-mentioned frequency spectrum perception algorithm based on channel gain and primary user's Combined estimator is emulated, in Rayleigh slow fading
Frequency spectrum perception performance curve under channel is as shown in Figure 3.In the figure, black (- * -) red (---) blueness (---) is bent
Line represents Doppler frequency shift f respectivelydCorrect verification and measurement ratio is perceived for 0.1,0.05,0.02 time-frequency spectrum, and respective channel coherence time Ta
With detection cycle TsBetween relation by Ta=Ts/fdDetermined.Dotted line represents the frequency spectrum obtained using conventional energy detection method
Perceptual performance, solid line represent the frequency spectrum perception performance of the combined estimation method proposed using this programme.Can be with from simulation result
Find out, the new method designed by the present invention obtains significant performance boost compared with conventional method.Work as fdWhen=0.02, believe in height
Make an uproar than in the case of, new method improves about more than 10dB compared with conventional energy detection method.It is also noted that energy measuring declines in time-varying
It is difficult to obtain preferable detection performance to fall under channel, i.e. correct detection probability is difficult to reach 1, and by contrast, new departure is in SNR
100% correct detection probability can be obtained during=8dB, frequency spectrum perception performance when can be obviously improved under changing environment, thus
In actual applications also by more advantage.
Claims (6)
1. a kind of frequency spectrum perception implementation method, the high-performance frequency spectrum perception under time-varying fading channels can be realized；It is characterized in that：
The time-varying characteristics of channel gain are taken into full account, using the dynamical state space model newly proposed, realizes and is worked for primary user
State and the Combined estimator of time-varying slow fading channel gain, the detection algorithm of the Combined estimator depend on maximum a posteriori probability
With sequential detection；
Combined estimator is carried out to time varying channel gain and primary user's state using observation signal：Primary user's state is carried out first thick
Slightly pre-estimation, on this basis using the channel gain under maximum a-posteriori estimation time-varying slow fading, is finally further utilized
Particle filter technology reevaluates to authorized user's state；
Wherein, it is described to the rough pre-estimation of primary user's state progress, including：
Compare the size of reception signal energy and predetermined threshold value, obtain judged result, wherein, if the reception signal energy
Less than the predetermined threshold value, then the judged result is that current time frequency range is idle, if the reception signal energy is more than
Equal to the predetermined threshold value, then the judged result is current time frequency range authorized user occupancy；
Channel gain under the time-varying slow fading using maximum a-posteriori estimation, including：
When the judged result is the frequency range free time at current time, judge whether the current time is the channel cycle head moment；
If the current time is the channel cycle head moment, turned according to last moment channel magnitude estimate and channel magnitude priori
Probability is moved, estimates the channel status at current time；
If the current time is the non-channel cycle head moment, directly by the use of last moment channel magnitude estimate as it is current when
The channel status at quarter；
When the judged result is that current time frequency range authorized user takes, judge whether the current time is channel week
Moment phase head；
If the current time is the channel cycle head moment, likelihood function value is calculated using the observation at current time, utilizes institute
State the channel status at likelihood function value and channel magnitude prior probability Combined estimator current time；
If the current time is the non-channel cycle head moment, estimation likelihood is calculated using observation is accumulated in channel symbol periods
Functional value, utilize the estimation likelihood function value and the channel shape at channel magnitude prior probability Combined estimator current time calculated
State；
It is described that authorized user's state is reevaluated using particle filter technology, including：
Particle is generated according to sequential importance function；
The weights of importance of the particle is calculated according to likelihood function；
The weights of importance is normalized, and carries out resurveying particle；
According to the normalization result of the weights of importance and the particle resurveyed, judge that the time-frequency band at current time uses
State.
2. according to the method for claim 1, it is characterised in that：It is as two using time varying channel states and primary user's state
System hidden state, and described by means respectively of a single order Markov Chain and FSMC models, and by special time window
The energy and conduct systematic perspective measured value of sampled signal.
3. according to the method for claim 1, it is characterised in that：This method direct basis sampling number, minimum fading channel
Gain and noise variance select a threshold value；Using relative size relation between the threshold value and observation signal, according to a preliminary estimate
Go out the working condition of primary user, so as to provide basis for subsequent estimation.
4. according to the method for claim 1, it is characterised in that：Channel gain is updated according to pre- court verdict, no
Same pre- court verdict determines different estimation strategies, specifically when it is " 0 " to estimate primary user's state, directly uses priori shape
State transition probability updates channel gain；And when it is " 1 " to estimate primary user's state, then when being estimated using maximum posteriori criterion
Become channel gain.
5. according to the method for claim 1, it is characterised in that：It is long-range using time-varying slow fading channel lower channel coherence time
In the characteristic of detection cycle, cumulative observations value in the case of identical pre- judgement in some detection cycles and tired is further made full use of
The product free degree, so as to constantly correct the estimate of channel magnitude, effectively increase the estimated accuracy of channel magnitude.
6. according to the method for claim 1, it is characterised in that：On the basis of acquired channel gain estimate, adopt
With particle filter technology, posterior probability is approached using the particle and its weights of importance of one group of sequential renewal, and dependent on most
Big posterior probability criterion obtains the real-time Sequential Estimation of primary user's state, improves frequency spectrum detection accuracy.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310007794.XA CN103117817B (en) | 2013-01-09 | 2013-01-09 | A kind of frequency spectrum detecting method under time-varying fading channels |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310007794.XA CN103117817B (en) | 2013-01-09 | 2013-01-09 | A kind of frequency spectrum detecting method under time-varying fading channels |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103117817A CN103117817A (en) | 2013-05-22 |
| CN103117817B true CN103117817B (en) | 2018-02-02 |
Family
ID=48416100
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310007794.XA Active CN103117817B (en) | 2013-01-09 | 2013-01-09 | A kind of frequency spectrum detecting method under time-varying fading channels |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103117817B (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104954088A (en) * | 2014-03-28 | 2015-09-30 | 中国科学院声学研究所 | Frequency spectrum detection method based on partially observable Markov decision process model |
| CN103916969B (en) * | 2014-04-17 | 2018-02-09 | 北京邮电大学 | Combined authorization user perceives and Link State method of estimation and device |
| CN104333424B (en) * | 2014-10-16 | 2018-01-05 | 北京邮电大学 | A kind of frequency spectrum detection and unknown noise variance tracking method of estimation and device |
| CN104852874B (en) * | 2015-01-07 | 2019-04-19 | 北京邮电大学 | Adaptive Modulation recognition methods and device under a kind of time-varying fading channels |
| CN104994046B (en) * | 2015-07-14 | 2018-01-05 | 宁波大学 | A kind of interframe frequency spectrum sensing method in cognitive radio system |
| CN106685549B (en) * | 2016-11-17 | 2020-06-12 | 北京邮电大学 | Primary user spectrum sensing method and device |
| CN106788816B (en) * | 2016-11-30 | 2020-11-13 | 全球能源互联网研究院有限公司 | Channel state detection method and device |
| CN109150339B (en) * | 2017-06-28 | 2021-07-16 | 北京石油化工学院 | Frequency spectrum sensing method and system based on Rayleigh fading channel signal envelope |
| CN111817813B (en) * | 2020-06-18 | 2021-01-19 | 南京审计大学 | Signal time delay-frequency shift parameter estimation and target signal recovery method based on signal sampling multivariate hypothesis test |
| CN116866925A (en) * | 2023-06-30 | 2023-10-10 | 中山大学 | Credibility-based unmanned aerial vehicle frequency spectrum sensing method and system |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101242333A (en) * | 2008-02-29 | 2008-08-13 | 浙江大学 | A multi-address access method based on channel sensing and prediction |
| CN102256261A (en) * | 2011-07-14 | 2011-11-23 | 东北大学 | Dynamic spectrum access method with network cognition ability |
| CN102595570A (en) * | 2012-01-11 | 2012-07-18 | 北京邮电大学 | Hidden Markov model based spectrum accessing method for cognitive radio system |
| CN102625319A (en) * | 2012-04-06 | 2012-08-01 | 电信科学技术研究院 | Method and device for realizing wireless cognitive sensor network |
| CN102711115A (en) * | 2012-05-24 | 2012-10-03 | 上海交通大学 | Multiuser distributed access method of opportunistic spectrum resources in cognitive radio network |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| AU2011207344B2 (en) * | 2010-01-21 | 2015-05-21 | Asthma Signals, Inc. | Early warning method and system for chronic disease management |
-
2013
- 2013-01-09 CN CN201310007794.XA patent/CN103117817B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101242333A (en) * | 2008-02-29 | 2008-08-13 | 浙江大学 | A multi-address access method based on channel sensing and prediction |
| CN102256261A (en) * | 2011-07-14 | 2011-11-23 | 东北大学 | Dynamic spectrum access method with network cognition ability |
| CN102595570A (en) * | 2012-01-11 | 2012-07-18 | 北京邮电大学 | Hidden Markov model based spectrum accessing method for cognitive radio system |
| CN102625319A (en) * | 2012-04-06 | 2012-08-01 | 电信科学技术研究院 | Method and device for realizing wireless cognitive sensor network |
| CN102711115A (en) * | 2012-05-24 | 2012-10-03 | 上海交通大学 | Multiuser distributed access method of opportunistic spectrum resources in cognitive radio network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103117817A (en) | 2013-05-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103117817B (en) | A kind of frequency spectrum detecting method under time-varying fading channels | |
| US9681270B2 (en) | Device localization based on a learning model | |
| Huang et al. | Machine-learning-based data processing techniques for vehicle-to-vehicle channel modeling | |
| CN110346654B (en) | Electromagnetic spectrum map construction method based on common kriging interpolation | |
| CN104168075B (en) | Frequency spectrum detecting method and device in the case of a kind of without knowledge of noise covariance | |
| CN104333424B (en) | A kind of frequency spectrum detection and unknown noise variance tracking method of estimation and device | |
| CN107886160B (en) | BP neural network interval water demand prediction method | |
| CN104657418A (en) | Method for discovering complex network fuzzy association based on membership transmission | |
| CN105517019A (en) | Method for detecting LTE (Long Term Evolution) network performance by using integrated regression system | |
| CN113259325A (en) | Network security situation prediction method for optimizing Bi-LSTM based on sparrow search algorithm | |
| CN103916969A (en) | Combined authorized user perception and link state estimation method and device | |
| CN103220054B (en) | A kind of cognitive radio frequency spectrum sensing method based on Gabor algorithm and system | |
| CN103209005B (en) | The pre-examining system of frequency hop sequences of a kind of graphic based model | |
| CN106972899A (en) | A kind of cooperative frequency spectrum sensing method excavated based on multi-user's history perception data | |
| CN109219055B (en) | Main user duty ratio estimation method | |
| CN106296727A (en) | A kind of resampling particle filter algorithm based on Gauss disturbance | |
| CN105634634A (en) | Asynchronous channel perception method with unknown timing | |
| CN108400826A (en) | A kind of frequency spectrum sensing method based on circulant matrix eigenvalue | |
| CN105429913A (en) | Multi-level detection and identification method based on characteristic value | |
| Hassan et al. | Measurement‐based determination of parameters for non‐stationary TDL models with reduced number of taps | |
| CN105119668A (en) | Iterative spectrum sensing method based on double judgment | |
| Zhao et al. | Blind spectrum sensing for cognitive radio over time-variant multipath flat-fading channels | |
| Sun et al. | Joint detection scheme for spectrum sensing over time‐variant flat fading channels | |
| Liu et al. | Prediction of wireless network connectivity using a Taylor Kriging approach | |
| Mann et al. | State estimation and fault detection and identification for constrained stochastic linear hybrid systems |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |