CN110351884B - Spectrum opportunity access method based on double-layer multi-arm tiger machine statistical model - Google Patents

Spectrum opportunity access method based on double-layer multi-arm tiger machine statistical model Download PDF

Info

Publication number
CN110351884B
CN110351884B CN201910581217.9A CN201910581217A CN110351884B CN 110351884 B CN110351884 B CN 110351884B CN 201910581217 A CN201910581217 A CN 201910581217A CN 110351884 B CN110351884 B CN 110351884B
Authority
CN
China
Prior art keywords
channel
channels
access
sensing
secondary user
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
Application number
CN201910581217.9A
Other languages
Chinese (zh)
Other versions
CN110351884A (en
Inventor
张周
闫野
蒋品
邓宝松
赵维维
付军峰
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Tianjin (binhai) Intelligence Military-Civil Integration Innovation Center
National Defense Technology Innovation Institute PLA Academy of Military Science
Original Assignee
Tianjin (binhai) Intelligence Military-Civil Integration Innovation Center
National Defense Technology Innovation Institute PLA Academy of Military Science
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Tianjin (binhai) Intelligence Military-Civil Integration Innovation Center, National Defense Technology Innovation Institute PLA Academy of Military Science filed Critical Tianjin (binhai) Intelligence Military-Civil Integration Innovation Center
Priority to CN201910581217.9A priority Critical patent/CN110351884B/en
Publication of CN110351884A publication Critical patent/CN110351884A/en
Application granted granted Critical
Publication of CN110351884B publication Critical patent/CN110351884B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B17/00—Monitoring; Testing
    • H04B17/30—Monitoring; Testing of propagation channels
    • H04B17/309—Measuring or estimating channel quality parameters
    • H04B17/318—Received signal strength
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B17/00—Monitoring; Testing
    • H04B17/30—Monitoring; Testing of propagation channels
    • H04B17/309—Measuring or estimating channel quality parameters
    • H04B17/336—Signal-to-interference ratio [SIR] or carrier-to-interference ratio [CIR]
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B17/00—Monitoring; Testing
    • H04B17/30—Monitoring; Testing of propagation channels
    • H04B17/391—Modelling the propagation channel
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W74/00—Wireless channel access
    • H04W74/08—Non-scheduled access, e.g. ALOHA

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Physics & Mathematics (AREA)
  • Electromagnetism (AREA)
  • Quality & Reliability (AREA)
  • Mobile Radio Communication Systems (AREA)

Abstract

The invention discloses a spectrum opportunity access method based on a double-layer multi-arm slot machine statistical model. The method comprises the following steps: firstly, modeling a channel sensing and accessing process by using a double-layer multi-arm tiger machine statistical model; then, analyzing the operation scene of the secondary user, and if the secondary user belongs to a cognitive network under a single master user station, performing spectrum opportunity access by adopting an isomorphic channel perception access method; and if the cognitive network belongs to the cognitive network under the multi-master user station, performing spectrum opportunity access by adopting a heterogeneous channel perception access method. The invention ensures the profit loss performance of O (lnt) in limited time, the profit loss of the O (lnt) expression algorithm is linearly changed along with the lnt curve, the statistical asymptotic validity is realized under the condition of enough time, and the high-efficiency data transmission can be completed on the premise of not causing harmful interference to authorized users.

Description

Spectrum opportunity access method based on double-layer multi-arm tiger machine statistical model
Technical Field
The invention relates to the technical field of wireless networks, in particular to a spectrum opportunity access method based on a double-layer multi-arm slot machine statistical model.
Background
With the increasing scarcity of wireless spectrum resources, the problem of low utilization rate of the existing spectrum is obvious. As an effective means for alleviating the spectrum shortage problem, spectrum opportunity access becomes a research hotspot. The core idea of the opportunity access method is that the secondary user is allowed to transmit data by using the dynamically identified spectrum opportunity without influencing the normal work of the authorized user in the cognitive network. In order to accurately identify the opportunities, the secondary user has channel sensing capability, and can access the authorized channel opportunistically according to the sensing result, so that the spectrum utilization rate is improved. Most of the existing technical schemes are provided for the spectrum access problem under the condition that the spectrum statistical information is known, and the existing technical schemes are difficult to operate or poor in performance under the condition that the spectrum information is unknown.
Compared with the spectrum access research under the condition of known statistical information, the channel sensing access research under the condition of unknown statistical information is still in an initial stage, and the main challenge is that the sensing capability of a secondary user is limited, an optimal channel needs to be found for sensing under the condition of unknown statistical information, and the channel is accurately estimated on the basis of sensing observation data; after the channel is sensed in real time, the optimal frequency spectrum is dynamically selected for opportunistic access and data transmission based on the current observation data and historical data statistical information. In the conventional method, the idle probability information of the authorized spectrum is assumed to be known by the secondary user, and in each time slot, the secondary user always selects the optimal channel set for sensing and accesses all sensed idle channels. However, in an actual operation situation, especially in a mobile network or networking in an unknown regional environment, the statistical information of the licensed spectrum generally needs to be obtained through long-time measurement and data analysis, accurate information needs to pass through a channel sensing process for a sufficiently long time, and the statistical information is more difficult to obtain in an emergency condition or an enemy wireless environment. In order to fully detect the optimal channel set, the secondary user must sense the non-optimal channel for a sufficient number of times, accurately estimate channel statistical information by using sensing data, and distinguish the optimal authorized channel from the suboptimal authorized channel. Since the accurate acquisition of the channel information needs to go through a learning process with a certain duration, in the process, the secondary user erroneously perceives a non-optimal channel in each time slot or accesses the non-optimal channel, which causes data collision between the secondary user and the authorized user and causes loss of channel access revenue. Therefore, the efficient secondary user channel perception access method is designed by taking the optimal channel perception access income with known statistical information as a reference so as to realize the lowest income loss, is one of the unsolved problems, and has important significance for improving the network spectrum utilization rate.
The current main channel sensing access method has the following defects:
1) the actual environment adaptability is poor. Most of the existing methods are perception access methods under the condition that statistical information is known, and the existing methods cannot operate under the condition that the statistical information is unknown or the statistical information is difficult to obtain in time. The existing method considers that the number of sensing and access channels is the same, but in the practical situation, the receiving energy and the transmitting energy of a secondary user are greatly different, a plurality of channels need to be sensed under the condition that the power energy is limited, and partial high-quality channels are preferably accessed, and the traditional method does not consider the situation.
2) The method has large loss of income. The conventional channel sensing access method under the condition of unknown statistical information is mostly proposed based on the design of a classical multi-arm slot machine mathematical model, and aiming at the cognitive access under the condition of limited physical network or physical factors, a secondary user faces the limitation of power, energy or hardware, so that the sensing idle channel cannot be completely accessed.
Disclosure of Invention
The invention aims to provide a spectrum opportunity access method based on a double-layer multi-arm slot machine statistical model, which is wide in application range and low in yield loss.
The technical solution for realizing the purpose of the invention is as follows: a spectrum opportunity access method based on a double-layer multi-arm slot machine statistical model comprises the following steps:
step 1, modeling a channel sensing and accessing process by using a double-layer multi-arm slot machine statistical model;
step 2, analyzing the operation scene of the secondary user, and entering step 3 if the secondary user belongs to a cognitive network under a single main user station; if the cognitive network belongs to the cognitive network under the multi-master user station, entering the step 4;
step 3, adopting a isomorphic channel perception access method to carry out spectrum opportunity access;
and 4, performing spectrum opportunity access by adopting a heterogeneous channel perception access method.
Further, the channel sensing and accessing process is modeled by using the double-layer multi-arm slot machine statistical model in the step 1, and the channel sensing and accessing process specifically comprises the following steps:
the secondary user can sense a plurality of channels in each time slot and select part of the channels to perform opportunistic access; modeling a channel perception and access process through a double-layer multi-arm slot machine statistical model, wherein at least access yield loss in an O (lnt) relation with time t exists in any physically feasible method, and the O (lnt) represents that the yield loss of an algorithm linearly changes along with an lnt curve;
setting a cognitive network to have N authorized channels, wherein a master user accesses the channels according to time slots, and the idle probability of the master user channel meets theta1＞θ2＞…＞θN；
In a single time slot, the secondary user can simultaneously sense M channels, M is less than N, and at most, K sensing idle channels are accessed, K is less than or equal to M, and the method is used
Figure BDA0002113233180000021
And
Figure BDA0002113233180000022
respectively, the detection probability and the false alarm probability of channel i.
Further, the analysis of the secondary user operation scenario in step 2 is specifically as follows:
based on a statistical model of a double-layer multi-arm slot machine, the method aims at cognition under a single-master user station and a plurality of master user stationsTwo typical scenes of the network adopt two different methods of an isomorphic channel sensing method and a heterogeneous channel sensing method to sense channels respectively, under the condition of a single main user station, the energy of signals received on each main channel is similar and spectrum sensing parameters are the same within a certain distance range between a secondary user and a single main user, and the requirements are met
Figure BDA0002113233180000031
Under the condition of multiple main user stations, the distances between a secondary user and different main user stations are different, and the energy of received signals on each main channel is different, so that the situation belongs to the sensing situation of heterogeneous channels; for channel i, there are different perceptual parameters.
Further, the spectrum opportunity access is performed by using the homogeneous channel sensing access method in step 3, which specifically includes:
under the sensing scene of homogeneous channels, the sensing performance of the secondary user to all the primary user channels is the same, and the detection characteristic meets the detection probability
Figure BDA0002113233180000032
And false alarm probability
Figure BDA0002113233180000033
At time slot T, the secondary user records T (T) ═ T (T)1(t),T2(t),...,TN(t)) and Y (t) ═ Y1(t),Y2(t),...,YN(T)), wherein Ti(t) represents the number of time slots when channel i has been sensed before time slot t, Yi(t) represents the number of time slots before time slot t where the perceived channel i is idle;
the isomorphic channel perception access method comprises the following specific steps:
step 3.1: initialization: by using
Figure BDA0002113233180000037
Sensing all N main channels in each time slot, randomly selecting a sensing idle channel to access in each time slot, and updating Ti(t) and Yi(T) wherein Ti(t) and Yi(t) has the same meaning as above; wherein
Figure BDA0002113233180000038
Represents rounding up;
step 3.2: the process is cycled along with time:
step 3.2.1: at time slot t, for all grant channels i 1,2iIs estimated value of
Figure BDA0002113233180000034
And determining a set of perceptual channels M (t) comprising the largest M coefficients
Figure BDA0002113233180000035
A corresponding channel;
step 3.2.2: the channels of the secondary user perception set M (T) use the channel set I (T) representing perception idle, and the recorded data T is updated in sequencei(t) and Yi(t)；
Step 3.2.3: if set I (t) is not empty, access channel set
Figure BDA0002113233180000036
Returning to the step 3.2.1; otherwise, no channel is accessed.
Further, the spectrum opportunity access is performed by using the heterogeneous channel sensing access method in step 4, which specifically includes:
under the sensing scene of the homogeneous channel, the detection probability and the false alarm probability of the secondary user to different channels are different, and the secondary user can sense M channels in a single time slot, namely sensing
Figure BDA0002113233180000041
Any one of a set of potential channels, the set of perceived channels being denoted as
Figure BDA0002113233180000042
Wherein
Figure BDA0002113233180000043
To take a combination of M different channels from N different channels at a time,m is more than or equal to 0 and less than or equal to N, and the calculation formula is
Figure BDA0002113233180000044
Collection
Figure BDA0002113233180000045
In (1), let mij J 1,2, M denotes a set MiThe jth channel in the set if the secondary user perceives the set M at time slot tiUse of
Figure BDA0002113233180000046
Indicating a set of channels that are perceptually idle, using T before time slot Ti(t) denotes perception MiNumber of time slots of, Yi(T) denotes a set of channels that are perceived as idle, Ti,j(t), j ═ 1, 2.. M denotes perception MiAggregation and access channel mijNumber of time slots of, Yij(t) denotes perception MiAggregation and access channel mijIn time slots of the channel mij(ii) the accumulated revenue obtained; secondary user maintenance record Ti(t)、Yi(t)、Ti,j(t) and Yi,j(t) a statistic;
the method for sensing and accessing the heterogeneous channel comprises the following specific steps:
step 4.1: initialization: for the
Figure BDA0002113233180000047
Performing sensing M continuously in successive time slotsiAnd access sensing M in each time slotiIdle channels not accessed at that time; this process is repeated until set MiEach channel in (1) is accessed at least one time; for each time slot, updating Ti(t)、Yi(t)、Ti,j(t) and Yi,j(t),j＝1,2,...,M；
Step 4.2: the process is cycled along with time:
step 4.2.1: at time slot t, calculating coefficient
Figure BDA0002113233180000048
And selects a set of channels
Figure BDA0002113233180000049
Perception set
Figure BDA00021132331800000410
A channel of interest;
step 4.2.2: if it is not
Figure BDA00021132331800000411
Computing coefficients for a set of channels perceived as idle, non-space-time
Figure BDA00021132331800000412
Step 4.2.3: selecting a channel
Figure BDA00021132331800000413
Access channel
Figure BDA00021132331800000414
And checking whether the channel access is successful;
step 4.2.4: updating statistical data
Figure BDA00021132331800000415
Step 4.2.5: updating
Figure BDA0002113233180000051
Compared with the prior art, the invention has the remarkable advantages that: (1) establishing a double-layer multi-arm tiger machine statistical mathematical model, and providing a multi-channel sensing and partial access method under unknown information aiming at channel opportunity access; (2) the sensing access under the conditions of an imperfect channel and unknown channel statistical information can be realized, and the method has wide applicability; (3) aiming at two situations of isomorphic perception and heterogeneous perception, effective channel perception and access are achieved, and under the provided performance metric standard, O (lnt) revenue loss performance can be achieved under the conditions of limited time t and time asymptotic t → ∞.
Drawings
Fig. 1 is a diagram of a typical scenario of cognitive network spectrum access.
Fig. 2 is a flow chart of secondary user spectrum dynamic access.
Fig. 3 is a flow chart of the spectrum opportunity access method based on the double-layer multi-arm slot machine statistical model.
Fig. 4 is a schematic flow chart of isomorphic sensing spectrum access in the present invention.
Fig. 5 is a flowchart illustrating a heterogeneous sensing spectrum access according to the present invention.
Fig. 6 is a graph of r (t)/lnt performance of the homogeneous spectrum sensing access method in the embodiment of the present invention.
Fig. 7 is a graph of r (t)/lnt performance of the heterogeneous spectrum sensing access method according to the embodiment of the present invention.
Detailed Description
On the basis of a classical MABP statistical model, aiming at the multi-channel sensing and partial access process of a secondary user and the imperfect channel sensing condition, the invention provides a double-layer MABP statistical model to carry out statistical modeling on physical problems; on the basis of a double-layer MABP statistical model, aiming at an actual operation scene, a channel perception access opportunity method combining isomorphic channel perception and heterogeneous channel perception is provided, and the efficiency of the method is verified through theoretical analysis and simulation.
The following terms are explained for ease of understanding:
imperfect main channel perception: in a practical cognitive network, due to the limitation of the receiving signal-to-noise ratio of a primary user channel, the perception of a secondary user channel is usually imperfect, and channel identification errors can exist.
Opportunistic spectrum access: if the secondary user does not interfere with the primary user, the secondary user may identify the opportunistic spectrum and use it in the system.
Bi-MABP: a double-layer multi-arm tiger machine. Assuming there are multiple booms, selecting any one will generate random revenue; for the player, the instant profit generation process can be regarded as a black box and cannot be obtained by direct observation, and the statistical characteristics need to be obtained through learning and selection is made based on the statistical characteristics to ensure that the experimenter can obtain the highest profit.
And (3) detection probability: when the receiver input does have a signal, two decisions may be made due to interference background, etc.: "signal" or "no signal". When there is a signal, the probability of making a correct decision that "there is a signal" is called the detection probability.
False alarm probability: the probability that "no signal" is detected but "signal" is judged, or the probability that "signal" is judged as "no signal".
Figure BDA0002113233180000061
Meaning rounding up, for example, when N is 3, M is 2,
Figure BDA0002113233180000062
sample mean value: the mean is the sum of all data in a set of data divided by the number of the set of data, and the sample mean is the mean of the sample data in the population.
It is desired that: the probability of each possible outcome in the trial is multiplied by the sum of its outcomes, which reflects the magnitude of the average value of the random variable.
Figure BDA0002113233180000063
The combination is called, that is, M different elements (0. ltoreq. M. ltoreq.N) are taken out of N different elements at a time. The calculation formula is as follows:
Figure BDA0002113233180000064
a main user: users authorized to access the spectrum.
And (4) secondary users: and the users which can access the spectrum opportunity under the condition of the primary user are not influenced.
f (t) to O (lnt): denotes that f (t) there is a constant c such that for time t > t0There is a constant such that f (t). ltoreq.clnt.
The invention is described in further detail below with reference to the figures and the embodiments.
With reference to fig. 3, the spectrum opportunity access method based on the double-layer multi-arm slot machine statistical model of the present invention includes the following steps:
step 1, modeling a channel sensing and accessing process by using a double-layer multi-arm slot machine statistical model, which comprises the following steps:
partial channel sensing scenarios for the secondary user to the grant channel include both homogeneous and heterogeneous channel sensing scenarios, as shown in fig. 1. The secondary user in each time slot can sense a plurality of channels and selects part of the channels to carry out opportunistic access; modeling a channel perception and access process through a double-layer MABP statistical model, and proving that at least access yield loss in an O (lnt) relation with time t exists in any physically feasible method;
setting N authorized channels in a cognitive network, and accessing a channel by a master user according to a time slot; the idle probability of the main user channel satisfies theta1＞θ2＞…＞θN；
In a single time slot, the secondary user can simultaneously sense M channels, M is less than N, and at most, K sensing idle channels are accessed, K is less than or equal to M, and the method is used as shown in figure 1
Figure BDA0002113233180000065
And
Figure BDA0002113233180000066
respectively representing the detection probability and the false alarm probability of the channel i;
aiming at the problem, modeling is carried out by using a double-layer multi-arm slot machine problem, a player needs to select to pull the slot machine arms at the same time, and the income statistical information of the pulled arms is unknown; after the player selects to pull the horn each time, the corresponding income value which possibly has errors can be obtained; based on the observation information, the player needs to further decide to obtain the real benefit of part of the arms; to maximize the revenue from multiple trials, players need to design efficient information estimation and decision methods that minimize the loss compared to the revenue expectation given the statistical information.
Step 2, analyzing the operation scene of the secondary user, and entering step 3 if the secondary user belongs to a cognitive network under a single main user station; if the cognitive network belongs to the cognitive network under the multi-master user station, entering the step 4, specifically as follows:
on the basis of a double-layer multi-arm slot machine statistical model, a solution is provided for the problem that a channel opportunity is accessed to a corresponding specific double-layer multi-arm slot machine, a secondary user operation scene is analyzed, and two different channel perception design methods are designed for two typical scenes, namely a cognitive network under a single main user station and a plurality of main user stations, as shown in fig. 2, and the two different channel perception design methods comprise an isomorphic channel perception method and a heterogeneous channel perception method. Under the condition of a single main user station, the energy of the received signals on each main channel is similar and the spectrum sensing parameters are the same within a certain distance range between a secondary user and the single user, so that the requirements of meeting the requirements of the requirements on the energy of the received signals on the frequency spectrum sensing parameters are met
Figure BDA0002113233180000071
Under the condition of multiple main user stations, the distances between a secondary user and different user stations are different, and the energy of received signals on each main channel is different, so that the heterogeneous channel sensing condition is achieved. In particular, for a certain channel i, there are different perceptual parameters.
Step 3, performing spectrum opportunity access by adopting an isomorphic channel perception access method, which specifically comprises the following steps:
under the sensing access scene of the homogeneous channel, the sensing performance of the secondary user to all the main user channels is the same, and the detection characteristic meets the detection probability
Figure BDA0002113233180000072
And false alarm probability
Figure BDA0002113233180000073
The process flow is shown in FIG. 4;
at time slot T, the secondary user records T (T) ═ T (T)1(t),T2(t),...,TN(t)) and Y (t) ═ Y1(t),Y2(t),...,YN(T)), wherein Ti(t) represents the number of time slots when channel i has been sensed before time slot t, Yi(t) represents the number of time slots before time slot t where the perceived channel i is idle;
with reference to fig. 4, the isomorphic channel sensing access method specifically includes the following steps:
step 3.1: initialization: by using
Figure BDA0002113233180000074
Sensing all N main channels by each time slot, randomly selecting a sensing idle channel to access each time slot, and updating T and Y; wherein
Figure BDA0002113233180000075
Represents rounding up;
step 3.2: the process is cycled along with time:
step 3.2.1: for all grant channels i 1, 2.. times.n, at time slot t, θ, which can be calculated according to the expectation of the sample mean valueiEstimated value
Figure BDA0002113233180000076
And determining a set of perceptual channels M (t) comprising the largest M coefficients
Figure BDA0002113233180000081
A corresponding channel;
step 3.2.2: the channels of the secondary user perception set M (t) use the channel set I (t) representing perception idle, and the recorded data T (t) and Y (t) are updated in sequence;
step 3.2.3: if set I (t) is not empty, access channel set
Figure BDA0002113233180000082
Returning to the step 3.2.1; otherwise, no channel is accessed.
Step 4, performing spectrum opportunity access by adopting a heterogeneous channel perception access method, which specifically comprises the following steps:
under the sensing scene of the homogeneous channel, the detection probability and the false alarm probability of the secondary user to different channels are different, and the secondary user can sense M channels in a single time slot, namely sensing
Figure BDA0002113233180000083
Any one of a set of potential channels, the set of perceived channels being denoted as
Figure BDA0002113233180000084
Wherein
Figure BDA0002113233180000085
Taking out the combination of M different channels from N different channels each time, wherein M is more than or equal to 0 and less than or equal to N, and the calculation formula is
Figure BDA0002113233180000086
Collection
Figure BDA0002113233180000087
In (1), let mij J 1,2, M denotes a set MiThe jth channel in the set if the secondary user perceives the set M at time slot tiUse of
Figure BDA0002113233180000088
Indicating a set of channels that are perceptually idle, using T before time slot Ti(t) denotes perception MiNumber of time slots of, Yi(T) denotes a set of channels that are perceived as idle, Ti,j(t), j ═ 1, 2.. M denotes perception MiAggregation and access channel mijNumber of time slots of, Yij(t) denotes perception MiAggregation and access channel mijIn time slots of the channel mij(ii) the accumulated revenue obtained; secondary user maintenance record Ti(t)、Yi(t)、Ti,j(t) and Yi,j(t) a statistic;
with reference to fig. 5, the specific steps of the heterogeneous channel aware access method are as follows:
step 4.1: initialization: for the
Figure BDA0002113233180000089
Performing sensing M continuously in successive time slotsiAnd access sensing M in each time slotiIdle channels not accessed at that time; this process is repeated until set MiEach channel in (1) is accessed at least one time; for each time slot, updating Ti(t)、Yi(t)、Ti,j(t) and Yi,j(t),j＝1,2,...,M；
Step 4.2: the process is cycled along with time:
step 4.2.1: at time slot t, calculating coefficient
Figure BDA00021132331800000810
And selects a set of channels
Figure BDA0002113233180000091
Perception set
Figure BDA0002113233180000092
A channel of interest;
step 4.2.2: if it is not
Figure BDA0002113233180000093
Computing coefficients for a set of channels perceived as idle, non-space-time
Figure BDA0002113233180000094
Step 4.2.3: selecting a channel
Figure BDA0002113233180000095
Access channel
Figure BDA0002113233180000096
And checking whether the channel access is successful;
step 4.2.4: updating statistical data
Figure BDA0002113233180000097
Step 4.2.5: updating
Figure BDA0002113233180000098
The access method performance metrics are performed as follows:
under the condition that the statistical information is unknown, the access method has the profit loss. The expected revenue for the access method phi is given by the revenue loss metric equation as follows:
Figure BDA0002113233180000099
under the conditions of perfect channel sensing and full channel access sensing, because the secondary user is unknown about the channel information, the yield loss has a lower bound o (lnt), that is, any feasible method at least has the yield loss in a logarithmic relation with time. Compared with the perfect channel sensing situation, sensing the situation that the non-optimal channel is accessed in the idle channel possibly causes more profit loss, wherein the profit refers to the data volume successfully transmitted by the secondary user through the primary user spectrum access and is determined by the receiving end confirmation message. In practical situations, if the message is not received, the message collides with the transmission of the primary user who fails to detect, so that secondary transmission failure is caused, and the corresponding transmission gain is 0.
Through analysis, the secondary user performs dynamic sensing access of the primary user channel according to the method, and the loss of revenue R (t) before the time slot t meets the requirement of R (t) -O (lnt).
Example 1
In a specific embodiment of the invention, a scene of partial channel perception access of a secondary user is considered, statistical modeling is carried out on channel access, and the fact that the statistical information is unknown, the next user at least faces the profit loss in O (lnt) relation with time is analyzed, and with the performance as the target, a spectrum opportunity access method based on a double-layer multi-arm slot machine statistical model is adopted for spectrum access. With reference to fig. 6 and 7, strict analysis and simulation prove that the yield loss of the proposed method satisfies the convergence performance of o (lnt) under the condition of finite time t and time asymptotic t → ∞.
Fig. 6 shows a simulation result of channel sensing access of the next user in an isomorphic channel sensing scenario, that is, a time-varying area curve of the loss-of-revenue performance, when the channel access number K is 2. Fig. 7 shows a simulation result of the K value for the next user channel sensing access in the heterogeneous channel sensing scenario. It can be seen that for homogeneous channel sensing and channel sensing access with independent and same distribution model, the normalized loss-of-revenue value increases when the number of access channels K increases from 1 to the number of sensing channels M. This is because the M optimal sensing channels selected by the method proposed by us and the method assisted by prior knowledge may be different in the partial channel sensing process, resulting in additional channel access revenue loss. If we choose to access more channels, i.e. the K value increases further, the channel selected by the proposed method may differ more from the optimal channel set selected by the prior-known-aided method, resulting in an increased normalized revenue loss value. Unlike the homogeneous sensing scenario, it can be seen from fig. 7 that the normalized revenue loss of channel sensing access under heterogeneous channel sensing decreases as the value of the access channel number K increases. This is because in this scenario we adopt a channel sensing access method based on a channel sensing set, and utilize feedback information of the channel access result.
In conclusion, the method provided by the invention researches the perception access problem under the conditions of imperfect channels and unknown channel statistical information, is closer to the application scene of real cognitive radio, and has wide applicability; providing a statistical model of the problem of the double-layer multi-arm slot machine, and modeling partial perception and selective access processes of secondary users to the authorized channel; the matching problem provides a method performance measurement standard under the condition of unknown statistical information; aiming at the two situations of isomorphic perception and heterogeneous perception, an effective channel perception and access method is provided, and under the provided performance metric standard, O (lnt) revenue loss performance can be realized under the conditions of limited time t and time asymptotic t → ∞.

Claims (2)

1. A spectrum opportunity access method based on a double-layer multi-arm slot machine statistical model is characterized by comprising the following steps:
step 1, modeling a channel sensing and accessing process by using a double-layer multi-arm slot machine statistical model;
step 2, analyzing the operation scene of the secondary user, and entering step 3 if the secondary user belongs to a cognitive network under a single main user station; if the cognitive network belongs to the cognitive network under the multi-master user station, entering the step 4;
step 3, adopting a isomorphic channel perception access method to carry out spectrum opportunity access;
step 4, performing spectrum opportunity access by adopting a heterogeneous channel perception access method;
the channel sensing and accessing process is modeled by using the double-layer multi-arm slot machine statistical model in the step 1, and the channel sensing and accessing process specifically comprises the following steps:
the secondary user can sense a plurality of channels in each time slot and select part of the channels to perform opportunistic access; modeling a channel perception and access process through a double-layer multi-arm slot machine statistical model, wherein at least access yield loss in an O (ln t) relation with time t exists in any physically feasible method, and the O (ln t) represents that the yield loss of an algorithm linearly changes along with an ln t curve;
setting a cognitive network to have N authorized channels, wherein a master user accesses the channels according to time slots, and the idle probability of the master user channel meets theta1＞θ2＞…＞θN；
In a single time slot, the secondary user can simultaneously sense M channels, M is less than N, and at most, K sensing idle channels are accessed, K is less than or equal to M, and the method is used
Figure FDA0002580573670000011
And
Figure FDA0002580573670000012
respectively representing the detection probability and the false alarm probability of the channel i;
the spectrum opportunity access is performed by adopting the isomorphic channel sensing access method in the step 3, which specifically comprises the following steps:
under the sensing scene of homogeneous channels, the sensing performance of the secondary user to all the primary user channels is the same, and the detection characteristic meets the detection probability
Figure FDA0002580573670000013
And false alarm probability
Figure FDA0002580573670000014
At time slot T, the secondary user records T (T) ═ T (T)1(t),T2(t),...,TN(t)) and Y (t) ═ Y1(t),Y2(t),...,YN(T)), wherein Ti(t) represents the number of time slots when channel i has been sensed before time slot t, YiWhen (t) representsThe number of idle time slots of the perceived channel i before the slot t;
the isomorphic channel perception access method comprises the following specific steps:
step 3.1: initialization: by using
Figure FDA0002580573670000015
Sensing all N main channels in each time slot, randomly selecting a sensing idle channel to access in each time slot, and updating Ti(t) and Yi(T) wherein Ti(t) and Yi(t) has the same meaning as above; wherein
Figure FDA0002580573670000016
Represents rounding up;
step 3.2: the process is cycled along with time:
step 3.2.1: at time slot t, for all grant channels i 1,2iIs estimated value of
Figure FDA0002580573670000021
And determining a set of perceptual channels M (t) comprising the largest M coefficients
Figure FDA0002580573670000022
A corresponding channel;
step 3.2.2: the channels of the secondary user perception set M (T) use the channel set I (T) representing perception idle, and the recorded data T is updated in sequencei(t) and Yi(t)；
Step 3.2.3: if set I (t) is not empty, access channel set
Figure FDA0002580573670000023
Returning to the step 3.2.1; otherwise, no channel is accessed;
the spectrum opportunity access is performed by adopting the heterogeneous channel sensing access method in the step 4, which specifically comprises the following steps:
under the isomorphic channel perception scene, the detection probability and the false alarm probability of the secondary user for different channels existIn contrast, a secondary user can perceive M channels in a single time slot, i.e., perceive
Figure FDA0002580573670000024
Any one of a set of potential channels, the set of perceived channels being denoted as
Figure FDA0002580573670000025
Wherein
Figure FDA0002580573670000026
Taking out the combination of M different channels from N different channels each time, wherein M is more than or equal to 0 and less than or equal to N, and the calculation formula is
Figure FDA0002580573670000027
Collection
Figure FDA0002580573670000028
In (1), let mijJ 1,2, M denotes a set MiThe jth channel in the set if the secondary user perceives the set M at time slot tiUse of
Figure FDA0002580573670000029
Indicating a set of channels that are perceptually idle, using T before time slot Ti(t) denotes perception MiNumber of time slots of, Yi(T) denotes a set of channels that are perceived as idle, Ti,j(t), j ═ 1, 2.. M denotes perception MiAggregation and access channel mijNumber of time slots of, Yij(t) denotes perception MiAggregation and access channel mijIn time slots of the channel mij(ii) the accumulated revenue obtained; secondary user maintenance record Ti(t)、Yi(t)、Ti,j(t) and Yi,j(t) a statistic;
the method for sensing and accessing the heterogeneous channel comprises the following specific steps:
step 4.1: initialization: for the
Figure FDA00025805736700000210
Performing sensing M continuously in successive time slotsiAnd access sensing M in each time slotiIdle channels not accessed at that time; this process is repeated until set MiEach channel in (1) is accessed at least one time; for each time slot, updating Ti(t)、Yi(t)、Ti,j(t) and Yi,j(t),j＝1,2,...,M；
Step 4.2: the process is cycled along with time:
step 4.2.1: at time slot t, calculating coefficient
Figure FDA0002580573670000031
And selects a set of channels
Figure FDA0002580573670000032
Perception set
Figure FDA0002580573670000033
A channel of interest;
step 4.2.2: if it is not
Figure FDA0002580573670000034
Computing coefficients for a set of channels perceived as idle, non-space-time
Figure FDA0002580573670000035
Step 4.2.3: selecting a channel
Figure FDA0002580573670000036
Access channel
Figure FDA0002580573670000037
And checking whether the channel access is successful;
step 4.2.4: updating statistical data
Figure FDA0002580573670000038
Step 4.2.5: updating
Figure FDA0002580573670000039
2. The method for spectrum opportunity access based on the double-layer multi-arm slot machine statistical model of claim 1, wherein the analysis of the secondary user operation scenario in step 2 is as follows:
based on a double-layer multi-arm slot machine statistical model, aiming at two typical scenes of a cognitive network under a single-master user station and a multi-master user station, channel sensing is carried out by respectively adopting two different methods, namely an isomorphic channel sensing method and a heterogeneous channel sensing method, under the condition of the single-master user station, a secondary user and a single master user are within a certain distance range, the energy of received signals on each master channel is similar, frequency spectrum sensing parameters are the same, and the requirements of meeting the requirements of the requirement of similar received signal energy and same frequency spectrum
Figure FDA00025805736700000310
Under the condition of multiple main user stations, the distances between a secondary user and different main user stations are different, and the energy of received signals on each main channel is different, so that the situation belongs to the sensing situation of heterogeneous channels; for channel i, there are different perceptual parameters.
CN201910581217.9A 2019-06-29 2019-06-29 Spectrum opportunity access method based on double-layer multi-arm tiger machine statistical model Active CN110351884B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910581217.9A CN110351884B (en) 2019-06-29 2019-06-29 Spectrum opportunity access method based on double-layer multi-arm tiger machine statistical model

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910581217.9A CN110351884B (en) 2019-06-29 2019-06-29 Spectrum opportunity access method based on double-layer multi-arm tiger machine statistical model

Publications (2)

Publication Number Publication Date
CN110351884A CN110351884A (en) 2019-10-18
CN110351884B true CN110351884B (en) 2020-08-28

Family

ID=68177459

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910581217.9A Active CN110351884B (en) 2019-06-29 2019-06-29 Spectrum opportunity access method based on double-layer multi-arm tiger machine statistical model

Country Status (1)

Country Link
CN (1) CN110351884B (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111313994B (en) * 2019-12-30 2021-04-30 中国人民解放军军事科学院国防科技创新研究院 Multi-user spectrum access method based on multi-arm gambling machine model under fairness principle
CN113098641B (en) * 2021-03-26 2022-09-09 天津(滨海)人工智能军民融合创新中心 Opportunistic spectrum access method under energy limitation condition

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103179675A (en) * 2013-01-08 2013-06-26 中国人民解放军理工大学通信工程学院 Epsilon-greed based online sequential perceiving and opportunity accessing method
WO2017023074A1 (en) * 2015-07-31 2017-02-09 주식회사 윌러스표준기술연구소 Multi-carrier signal transmission method, device, and system
CN108029138A (en) * 2015-07-21 2018-05-11 法国国立应用科学学院 Method for the chance access to frequency spectrum

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170278114A1 (en) * 2016-03-24 2017-09-28 Xerox Corporation Adaptive collaborative filtering with extended kalman filters and multi-armed bandits

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103179675A (en) * 2013-01-08 2013-06-26 中国人民解放军理工大学通信工程学院 Epsilon-greed based online sequential perceiving and opportunity accessing method
CN108029138A (en) * 2015-07-21 2018-05-11 法国国立应用科学学院 Method for the chance access to frequency spectrum
WO2017023074A1 (en) * 2015-07-31 2017-02-09 주식회사 윌러스표준기술연구소 Multi-carrier signal transmission method, device, and system

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
Joint Channel Selection and Power Control for NOMA: A Multi-Armed Bandit Approach;Mohamed Ali Adjif, Oussama Habachi, and Jean-Pierre Cances;《2019 Mathematical Tools for IoT Networks Modeling》;20190418;全文 *
Multiple Armed Bandits Based Opportunistic Channel Access with Switching Cost in Block Fading Channel;Zhiqiang Qin，Weijun Liu, Yawu Jiang, Wenhui Li，Jingdi Zhou;《2015 International Conference on Wireless Conmunications &Signal Processing》;20151203;全文 *
基于认知的短波宽带信道选择关键技术研究;秦志强;《中国优秀博士论文全文数据库》;20180615;全文 *

Also Published As

Publication number Publication date
CN110351884A (en) 2019-10-18

Similar Documents

Publication Publication Date Title
US8805428B2 (en) Cooperative spectrum sensing in cognitive radio networks
KR101469659B1 (en) Method for controlling random access for the efficient sensing of the cooperative spectrum in a cognitive radio-based frequency resource sharing system
CN110351884B (en) Spectrum opportunity access method based on double-layer multi-arm tiger machine statistical model
EP3694149A1 (en) Network quality determining method and apparatus, and storage medium
Yousefvand et al. Maximizing network capacity of cognitive radio networks by capacity-aware spectrum allocation
CN110351886A (en) Opportunistic spectrum access method based on sideband observation information multi-arm Slot Machine model
CN114205046B (en) Communication perception integrated network interference coordination method and device
CN110649982B (en) Double-threshold energy detection method based on secondary user node selection
CN111313994B (en) Multi-user spectrum access method based on multi-arm gambling machine model under fairness principle
KR20110094983A (en) Weighted-cooperative spectrum sensing scheme using markov model in cognitive radio systems
CN102394712B (en) Joint optimization method of detection period and detection time in cognitive radio system
CN104394543A (en) Joint frequency spectrum sensing method based on Adaboost algorithm
CN113472463A (en) Method for detecting interference constraint violation behavior of secondary user in cognitive wireless network
Zhang et al. Machine learning based protocol classification in unlicensed 5 GHz bands
CN107509253B (en) Channel frequency band access method and device
CN113395709B (en) Service coverage analysis method facing user base station poisson distribution
CN114866171B (en) Signal statistics method, device, computer equipment and storage medium
CN111294128B (en) Opportunistic spectrum access method based on Markov channel model
CN113068247B (en) Wireless access point power estimation method based on multilayer perceptron
Bezerra et al. Hidden markov model spectrum predictor for poisson distributed traffic
Pu et al. Primary user emulation detection algorithm based on distributed sensor networks
CN114285701B (en) Method, system, equipment and terminal for identifying transmitting power of main user
Yi et al. Secondary user monitoring in unslotted cognitive radio networks with unknown models
Li et al. Research on adaptive energy detection technology based on correlation window
CN114285506B (en) Hybrid two-stage multi-star cooperative spectrum sensing method

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant