WO2011023118A1 - 频谱预测方法、装置和系统 - Google Patents

频谱预测方法、装置和系统 Download PDF

Info

Publication number
WO2011023118A1
WO2011023118A1 PCT/CN2010/076380 CN2010076380W WO2011023118A1 WO 2011023118 A1 WO2011023118 A1 WO 2011023118A1 CN 2010076380 W CN2010076380 W CN 2010076380W WO 2011023118 A1 WO2011023118 A1 WO 2011023118A1
Authority
WO
WIPO (PCT)
Prior art keywords
spectrum
resource occupancy
target
mode
spectrum resource
Prior art date
Application number
PCT/CN2010/076380
Other languages
English (en)
French (fr)
Inventor
张弓
王蛟
项炎平
Original Assignee
华为技术有限公司
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 华为技术有限公司 filed Critical 华为技术有限公司
Priority to BR112012004383-8A priority Critical patent/BR112012004383B1/pt
Priority to EP10811278.0A priority patent/EP2472933B1/en
Publication of WO2011023118A1 publication Critical patent/WO2011023118A1/zh
Priority to US13/407,220 priority patent/US8687516B2/en

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04WWIRELESS COMMUNICATION NETWORKS
    • H04W16/00Network planning, e.g. coverage or traffic planning tools; Network deployment, e.g. resource partitioning or cells structures
    • H04W16/14Spectrum sharing arrangements between different networks
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04WWIRELESS COMMUNICATION NETWORKS
    • H04W48/00Access restriction; Network selection; Access point selection
    • H04W48/16Discovering, processing access restriction or access information
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04WWIRELESS COMMUNICATION NETWORKS
    • H04W72/00Local resource management
    • H04W72/12Wireless traffic scheduling
    • H04W72/1215Wireless traffic scheduling for collaboration of different radio technologies

Definitions

  • the embodiments of the present invention relate to the field of communication technologies, and in particular, to a spectrum prediction method, device, and system. Background technique
  • DSA Dynamic spectrum access
  • SU unauthorized user
  • Frequency measurement technology is one of the basic technologies leading to practical application of cognitive radio technology.
  • the existing spectrum measurement technology mostly uses the first-order Markov process to predict the channel state.
  • the basic idea is: Assume that the channel state of the current time slot is only related to the channel state of the previous time slot; The transition of the channel state occurs at the beginning of each time slot.
  • the channel state at the next moment of the frequency measurement scheme of the first-order Markov model is only related to the current channel state, but is related to the history of each channel
  • the state of the time slot is not very relevant; in the analysis process, the correlation between the channels is not specifically considered. In the case of multiple channels, it is generally assumed that the channels are independent. Therefore, in the spectrum prediction process, the accuracy of predicting the spectrum state of the future time slot is low, the false detection rate and the missed detection rate are high, and the reliability of the spectrum hole selection is low.
  • the embodiments of the present invention provide a spectrum prediction method, device, and system to reduce the false detection rate and the miss detection rate in the spectrum prediction process, and improve the accuracy of the spectrum prediction result.
  • the embodiment of the present invention provides a frequency prediction method, including:.
  • An embodiment of the present invention further provides an apparatus for extracting a spectrum resource occupancy mode, including:
  • a first data acquisition unit configured to acquire first sample data of a target spectrum, where the first sample data includes existing service information, channel information, and channel state information of the target spectrum;
  • a sample matrix generating unit configured to extract channel state information of all channels of the same service in each time slot from the first sampled data to generate a sample matrix
  • the mode analysis unit is configured to analyze the sample matrix, and extract the spectrum resource occupancy mode combination of the target spectrum from the sample matrix.
  • An embodiment of the present invention further provides a spectrum resource prediction device, including:
  • the second data acquisition unit is configured to acquire second sampling data of the target frequency i", the second sampling The data includes real-time service information, channel information, and channel state information of the target spectrum; a target matrix generating unit is used to extract the channel state information of all channels of the same service in each time slot from the second sampled data to generate Target matrix
  • the channel prediction unit is configured to: when data matching the first spectrum resource occupancy mode in the spectrum resource occupancy mode combination exists in the target matrix, according to the second spectrum associated with the first spectrum resource occupancy mode The resource occupancy mode predicts the channel state of the target spectrum in the future time slot.
  • An embodiment of the present invention also provides a spectrum prediction system, including:
  • a spectrum resource occupancy mode extraction device configured to obtain first sampling data of a target spectrum, where the first sampling data includes existing service information, channel information, and channel state information of the target spectrum; from the first sampling data Extracting channel state information of all channels of the same service in each time slot to generate a sample matrix; extracting the target frequency spectrum resource occupancy mode combination from the sample matrix;
  • a device for predicting a frequency word resource configured to obtain second sampling data of the target spectrum, match the spectrum resource occupancy mode combination according to the second sampling data of the target spectrum, and compare the target spectrum in the target spectrum according to the matching result
  • the channel state of the future time slot is predicted.
  • the spectrum resource occupancy mode combination of the target spectrum can be obtained, and the second data of the target spectrum is used to perform the combination of the spectrum resource occupancy mode
  • the channel state of the target spectrum in the future time slot is predicted according to the matching result, which can reduce the false detection rate and the missed detection rate in the spectrum prediction process, and improve the accuracy of the spectrum prediction.
  • Figure 1 is a schematic diagram of a channel state determination method in an embodiment of the present invention
  • Fig. 2 is a channel idle delay (Channe l Vacancy Durat ion, abbreviated as: CVD) schematic diagram of sequence generation;
  • FIG. 3 is a schematic diagram of channel idle occurrence time slot (Channel Vacancy Appearance Interval, CVAI) sequence generation in an embodiment of the present invention
  • Figure 4 is a CVD distribution curve of the G SM900 uplink in an embodiment of the present invention.
  • Fig. 5 is a graph of the uplink S CR statistics of G SM900 and G SM 1800 in an embodiment of the present invention
  • Fig. 6 is a schematic diagram of a partial service SC in an embodiment of the present invention
  • Fig. 7 is a flowchart of a first embodiment of a spectrum prediction method according to the present invention.
  • Fig. 8 is a schematic diagram of a second embodiment of a spectrum prediction method according to the present invention.
  • FIG. 9 is a schematic diagram of a third embodiment of a spectrum prediction method according to the present invention.
  • Fig. 10 is a schematic structural diagram of the first embodiment of an apparatus for extracting spectrum resource occupancy patterns according to the present invention
  • Fig. 11 is a schematic structural diagram of a second embodiment of an apparatus for extracting spectrum resource occupancy patterns according to the present invention. Schematic;
  • Fig. 13 is a schematic structural diagram of an embodiment of a spectrum prediction system according to the present invention. Detailed ways
  • the wireless access standards applicable to the embodiments of the present invention include: Global System for Mobile Communications (Global
  • GSM Global System for Mobile Communications
  • WCDMA Wideband-Code Division Multiple Access
  • TD-SCDMA Time Division-Synchronized Code Division Multiple Access
  • CDMA Code Division Multiple access
  • WIMAX Wireless Local Area Network
  • WLA Wireless Local Area Network
  • LTE Long Term Evolution
  • the embodiment of the present invention performs analysis based on actual sampled data, and obtains analysis data at four locations in southern China (two urban environments and two rural environments).
  • the acquisition duration is 7 days, and the target spectrum range is 20nmz-3GHz.
  • the frequency distribution interval of the sampling points is 0.2MHz, and the time distribution of the sampling points is between The interval is 75s.
  • FIG. 1 is a schematic diagram of the channel state determination method in the embodiment of the present invention.
  • the highest value of the measured interference (noise) power is the highest sampling frequency of 11.
  • the lowest value is the lowest sampling frequency 12
  • the interval between the highest sampling frequency 11 and the lowest sampling frequency 12 in the figure is close to 3dB ⁇ V
  • the threshold power of interference can be defined.
  • the threshold power K In a certain frequency band, take the measured minimum power value of the channel and set it as min ⁇ , and define the interference threshold power K as the formula (1):
  • the channel state of the channel in this time slot can be considered as occupied; otherwise, the channel state of the channel in this time slot can be considered to be Idle state.
  • Fig. 2 is a schematic diagram of CVD sequence generation in an embodiment of the present invention.
  • the channel state is 1 (busy), and The channel status below K is 0 (idle).
  • the CVD of the sequence in FIG. 2 is the number of consecutive "0"s (idle) in the channel state information (Channel state information, abbreviated: CSI) sequence.
  • CSI Channel state information
  • FIG 3 is a schematic diagram of the CVAI sequence generation in the embodiment of the present invention. Taking the CSI sequence in the above figure as an example, the CVAI of the sequence in Figure 3 can be obtained, where CVAI is each consecutive "0" (idle) occurrence in the CSI sequence Interval.
  • Figure 4 shows the CVD distribution curve of the GSM900 uplink in the embodiment of the present invention. As shown in Figure 4, taking the GSM900 uplink as an example, after analyzing all sample data, the statistical CVD distribution curve conforms to the following formula (2 ):
  • y a + be ⁇ cx (2)
  • a 0.000947
  • b l.671621
  • c l.079681
  • r 2 0.993586.
  • the parameter r 2 is the similarity between the data calculated by the formula (2) and the actual data. The closer the value of r 2 is to 1, the more closely the statistical results of the formula are Close to the actual value.
  • the parameter table of abc in the CVD distribution curve of each service can be obtained through statistics, where service can refer to various types of services in wireless communication, for example: Table 1 CVD distribution curve parameters of each service As shown in the table.
  • SCR (t, S) the number of channels occupied by the service at time t/the total number of channels for the service (3)
  • SCR can reflect the congestion law of various services.
  • Figure 5 shows the GSM900 and GSM1800 uplinks in the embodiment of the present invention.
  • the SCR statistical curve chart can be obtained from Figure 5: The distribution of the SCR of the two services of GSM900 and GSM1800 has daily repeatability. Similarly, the SCR of other services also has daily repeatability, and the types of services are not limited here. . Then, analyze the frequency correlation (Spectrum Correlation, SC for short) of the SCR sequence.
  • FIG. 6 is a schematic diagram of a partial service SC in an embodiment of the present invention. It can be seen from Fig.
  • the value of the frequency-frequency correlation (SC) between different channels of the same service in the TV band and the GSM band can reach 90%, especially the GSM900 uplink (Uplink) and the GSM1800 uplink (Uplink).
  • the SC value of some TV bands is as high as 95% or more.
  • the SCR of the same service has strong similarity; different channels of the same service Time has strong spectral correlation in terms of time, space, and frequency. Therefore, the state of the future time slot of the channel can be predicted based on the historical state of the channel in the spectrum resource.
  • the main idea of the embodiment of the present invention is to predict the channel state of each service in the future time slot according to the first data of each channel included in the frequency resource of each service.
  • FIG. 7 is a flowchart of the first embodiment of the frequency borrowing prediction method according to the present invention. As shown in FIG. 7, the spectrum prediction method includes:
  • Step 101 Acquire first sample data of a target spectrum, where the first sample data includes existing service information, channel information, and channel state information of the target spectrum.
  • the method for obtaining first sampling data of the target spectrum specifically includes: collecting first data of the target spectrum to obtain first sampling data, where the first data includes historical data, for example: Existing spectrum resource data in the historical time period.
  • the first data of the target spectrum is collected with a fixed long time period and every certain interval as a time slot, where the fixed long time period and the time slot can be selected according to specific measurement capabilities. Collect and store the results of data collection to form initial first sample data.
  • the first sample data may include service information of all services in the collected target spectrum, and all channels of each service (Channe l) Information and channel state information of each channel, etc.
  • Step 102 From the first sampled data, extract channel state information of all channels of the same service in each time slot to generate a sample matrix.
  • the same service When processing the first sample data, the same service is used as a unit, and the values of each time slot of all channels included in the service are extracted as a two-dimensional sample matrix used for spectrum resource occupancy pattern mining, where "0" can be assumed Is the channel idle state, "is the channel occupancy state, and the obtained two-dimensional sample matrix is a two-dimensional "0/1" matrix of channels and time slots.
  • the process of generating the sample matrix may specifically include:
  • the channel status of the selected channel in the selected time slot is occupied, otherwise The channel state of the selected channel in the selected time slot is an idle state.
  • the channel state information corresponding to the occupancy state has an element bit of 1 in the sample matrix (it can also be other constants), and the channel state information corresponding to the idle state has an element bit of G in the sample matrix (it can also be other constants) .
  • the first sampled data is converted to The method for determining matrix elements in the two-dimensional sample matrix process is specifically as follows: compare the maximum sampling power of the channel in a specific time slot with the threshold power K. If it is greater than K, the channel is considered to be occupied in the time slot. In the sample matrix, the element bit corresponding to the channel in the time slot is " ⁇ ; otherwise, if it is less than K, the channel is considered to be idle in the time slot. In the two-dimensional sample matrix, the element bit corresponding to the channel in the time slot is It is "0". In the two-dimensional sample matrix, "0" is used to indicate the idle state, and "1" is used for the occupied state, but there is no limit to the values of the idle state and the occupied state, and the use of other values is not excluded.
  • Step 103 Extract the target frequency spectrum resource occupancy mode combination from the sample matrix.
  • the spectrum resource occupancy mode extracted in this step is called the first spectrum resource occupancy mode. If the number of occurrences of the first frequency resource occupancy mode in the sample matrix exceeds the predefined threshold, the first frequency resource occupancy mode can be regarded as an effective spectrum resource occupancy mode. Subsequently, obtaining a second spectrum resource occupancy pattern associated with the first spectrum resource occupancy pattern whose number of occurrences exceeds a threshold value from the sample matrix. In the sample matrix, search for other effective spectrum resource occupancy modes associated with the first spectrum resource occupancy mode, and other effective spectrum resource occupancy modes associated with the first spectrum resource occupancy mode are called second spectrum resources Occupation mode.
  • the association relationship between the first spectrum resource occupancy mode and the second spectrum resource occupancy mode is the relationship between the source mode and the target mode, which may include the following examples:
  • Example 1 If the The first spectrum resource occupancy mode is included in the second spectrum resource occupancy mode, then the first spectrum resource occupancy mode is the source mode, and the second spectrum resource occupancy mode is the target mode. That is: if a certain frequency resource occupancy mode of the sample matrix is included in other modes, the included spectrum resource occupancy mode is called the source mode, and the spectrum resource occupancy mode that includes the source mode is called the target mode.
  • Example 2 If the second spectrum resource occupation mode appears every certain number of time slots after the first spectrum resource occupation mode appears, the first spectrum resource occupation mode is the source mode, and the first spectrum resource occupation mode is the source mode.
  • the second spectrum resource occupancy mode is the target mode. That is: if a certain frequency resource occupancy pattern of the sample matrix appears every N time slots after it appears, another pattern associated with it appears, then the spectrum resource occupancy pattern that appears first is called the source pattern, and the other pattern associated with it appears.
  • the spectrum resource occupation mode is called the target mode.
  • the first spectrum resource occupancy mode is the source mode
  • the second spectrum resource occupancy mode is the target mode.
  • the correlation coefficient between the source mode and the target mode may be defined as: in the sample matrix, the source mode and the target The ratio of the number of associated occurrences of the pattern to the total number of occurrences of the source pattern. But it is not limited to this definition, and can be other definitions. If the correlation coefficient between the target mode and the source mode reaches a certain threshold or more, the source mode and the corresponding target mode and correlation coefficients can be used as an effective spectrum resource occupancy mode combination.
  • the data structure in the spectrum resource occupancy mode combination may include the following fields: Source mode field: a two-dimensional array, which stores the detection rules used for spectrum prediction; target mode field: a two-dimensional array, which stores the spectrum resource occupancy mode associated with the source mode. If the channel state matrix of the current time slot matches the corresponding source mode, the target mode can be used to predict the channel state of the next time slot;
  • Correlation coefficient field a real number located between “0” and "1”, indicating the probability that the channel state of the next time slot is the channel state of the corresponding time slot in the target mode.
  • Step 104 Acquire second sampling data of the target spectrum, perform matching on the frequency resource occupancy mode combination according to the second sampling data of the target spectrum, and perform matching on the target spectrum in the future time slot according to the matching result.
  • the channel state is predicted.
  • the second data of the target spectrum is collected to obtain second sample data.
  • the second sample data may include service information, channel information, and channel state information at a certain moment or time period, such as including the target Real-time service information, channel information, and channel state information of the spectrum.
  • the second data of the target spectrum may be spectrum resource data at a certain moment or time period, for example, real-time spectrum resource data of the target spectrum in a current time period.
  • channel state information of all channels of the same service in each time slot is extracted to generate a target matrix.
  • the second frequency resource occupancy mode associated with the first spectrum resource occupancy mode To predict the channel state of the target spectrum in the future time slot. For example: Use the first spectrum resource occupancy mode (source mode) to match the target matrix. If a certain first spectrum resource occupancy mode can match a certain part of the value in the target matrix, it is considered that the first spectrum resource occupancy mode is matched successfully. Stop the analysis process and output the first spectrum resource occupancy mode.
  • the first spectrum resource occupancy mode that is successfully matched is used as input, and the second spectrum resource occupancy mode (target mode) associated with the first spectrum resource occupancy mode is used to predict the channel state of the corresponding position; output prediction
  • the pre-judgment result is the channel state of a certain channel or several channels in the next time slot, that is, the channel state consistent with the corresponding position of the second spectrum resource occupancy mode, and the coincidence probability is a percentage of the correlation coefficient.
  • the unlicensed user can select an available channel for selective access based on the pre-judgment result. If the state of a channel of the target spectrum in the future time slot is idle, then in the future The time slot selects the channel for access.
  • a combination of spectrum resource occupancy modes can be obtained from the first data of frequency resources, so as to perform frequency transmission prediction based on the correlation of different channels of the same service, which increases the access opportunities of unlicensed users to the spectrum and reduces
  • the false detection rate and the missed detection rate in the spectrum prediction process are improved, and the accuracy of frequency borrowing prediction and the reliability of spectrum hole selection are improved.
  • the spectrum hole refers to the temporarily available frequency band that is not occupied in the spectrum of the currently authorized user.
  • FIG. 8 is a schematic diagram of the second embodiment of the spectrum prediction method according to the present invention. Based on the first embodiment of the spectrum prediction method according to the present invention, it is assumed that the spectrum resource occupancy mode in the spectrum resource occupancy mode combination is in an inclusive relationship. In the collection process of the spectrum resource occupancy pattern, the spectrum resource occupancy patterns P1 and P2 are obtained. As shown in FIG. 8, the association relationship between P1 and P2 is an inclusive relationship, that is, P1 is a subset of P2. Among them, P1 is the source mode, and P2 is the target mode. The source mode P1 is a 2 x 3 matrix, and the target mode P2 is a 2 x 4 matrix.
  • the target of the spectrum prediction in this embodiment is the channel state at the moment, and the specific process of the spectrum prediction method includes: a spectrum resource occupancy mode extraction process and a spectrum resource pre-judgment process.
  • the extraction process of the frequency language resource occupancy mode is as follows:
  • the first data transmitted frequently by the target is collected; taking the same service as a unit, extract the value of each time slot of all channels included in the service to obtain a sample matrix, structure sample matrix such as a column matrix is shown in Figure 8 to a two-dimensional column t 5 "0 / ⁇ matrix, from the two-dimensional" extract 0/1 "source pattern matrix Pl, P2 target mode, and PI and P2
  • the phase The number of relationships and association relationships, etc. constitute a spectrum resource occupancy mode combination. For details, reference may be made to the specific description in the first embodiment of the frequency prediction method of the present invention, which is not repeated here.
  • the pre-judgment process of spectrum resources is as follows:
  • the method for judging the channel state refers to the related description and formula (1) of the first embodiment of the frequency transmission prediction method of the present invention.
  • the structure of the target matrix is, for example, a two-dimensional "0/ matrix from t n — 3 columns to columns of the matrix in Figure 8. Assuming that the future time slot is t n+1 time slot, the purpose of this spectrum prediction is to predict t n+1 time The channel status of the slot.
  • the source pattern P1 matches the matrix formed by the ch4 and ch5 rows from the column from time slot t-2 to slot tj ⁇ (which may be a complete match or a certain precision match).
  • the source mode P1 and the channel state of the current time slot are successfully matched, and the state of the channels ch4 and ch5 of the next time slot is pre-judged according to the target mode P2 associated with the source mode P1.
  • the prediction result can be obtained: the probability that the channel state of ch4 in time slot t n+1 is 0 (idle) is equal to 95%, and the probability that the channel state of ch5 in time slot t n+1 is 1 (busy) Equal to 95%.
  • channel ch4 can be selected as the channel for the unauthorized user to dynamically access in the time slot t +1 .
  • a combination of spectrum resource occupancy modes can be obtained from the first data of spectrum resources, so as to perform frequency prediction based on the correlation of different channels of the same service, which increases the access opportunities of unlicensed users to the spectrum;
  • the availability prediction process has high accuracy. For example, the accuracy of the prediction for the GSM frequency band and the TV frequency band is particularly prominent, which reduces the false detection rate and the missed detection rate in the frequency prediction process.
  • FIG. 9 is a schematic diagram of a third embodiment of a spectrum prediction method according to the present invention.
  • the frequency resource occupancy mode in the spectrum resource occupancy mode combination is Continuous relationship.
  • the frequent resource occupancy modes P3 and P4 are obtained.
  • the association relationship between P3 and P4 is a continuous relationship, that is, P4 appears every N time slots after each occurrence of P3.
  • N 2.
  • P3 is the source mode
  • P4 is the target mode.
  • Source The pattern P3 is a 2 ⁇ 3 matrix
  • the target pattern P4 is a 2 ⁇ 4 matrix.
  • the correlation coefficient between P3 and P4 is 90%, that is, every time P3 appears 100 times, P2 appears every two time slots after each occurrence of P3.
  • the number of times is 90 times.
  • the target of the spectrum prediction in this embodiment is the channel state at time t n+1
  • the specific process of the spectrum prediction method includes: a spectrum resource occupancy mode extraction process and a spectrum resource pre-judgment process.
  • the extraction process of the spectrum resource occupancy mode is as follows:
  • the structure of the matrix is, for example, a column-to-column two-dimensional "0/1" matrix of the matrix shown in FIG. 9, and the correlation coefficients and correlations of the source pattern P3, the target pattern P4, P3, and P4 are extracted from the two-dimensional "0/1" matrix.
  • the relationship, etc. form a combination of frequency resource occupancy modes. For details, reference may be made to the specific description in the first embodiment of the frequency general prediction method of the present invention, which will not be repeated here.
  • the source mode P3 and the target mode P4 can be obtained as a continuous relationship.
  • the target mode P4 appears every two time slots after the source mode P3 appears, the source mode P3 is a 2 X 3 matrix, and the target mode P4 is 2 X 3 Matrix.
  • the correlation coefficient between the source mode P3 and the target mode P4 is 90% and other information.
  • the pre-judgment process of spectrum resources is as follows:
  • a target matrix Store the scan results and analyze the scan results to generate a target matrix.
  • the structure of the target matrix is shown in Fig. 9 as a two-dimensional "0/1" matrix from column t n - 3 to column t n of the matrix.
  • the purpose of frequency prediction is to predict ⁇ +1 The channel status of the time slot.
  • the source pattern P 3 matches the matrix of ch3 and cM rows of the column and cM rows from t...-4 time slots to t n - 2 time slots (it can be a complete match or a certain precision match)
  • the source mode P3 successfully matches the channel state of the current time slot, and the state of the channels ch3 and ch4 of the next time slot is pre-judged according to the target mode P4 associated with the source mode P3.
  • the probability that the channel states of channels ch 3 and ch4 in time slot t nl to time slot t n+3 are completely consistent with P4 at this time is 90 %. Therefore, the prediction result can be obtained: (:! The probability that the channel states from slot ⁇ to slot 3 are 0 (idle), 0 (idle), and 1 (busy) respectively, is 90%, and ch4 is at t The channel states from time slot n+1 to time slot t n+3 are 1 (busy), 1 (busy), and 0 (idle) respectively The probability is 90%.
  • Unlicensed users can access the channel in the idle state (idle) in the time slot ⁇ +1 . Therefore, according to the prediction result, the channel ch3 can be selected as the channel for the unauthorized user to dynamically access in the time slot tn+1.
  • a combination of spectrum resource occupancy modes can be obtained from the first data of spectrum resources, so as to perform frequency transmission prediction according to the correlation of different channels of the same service, which increases the access opportunities of unlicensed users to the spectrum;
  • the continuous occurrence of the relationship between the mode and the target mode can predict the channel status of the next multiple future time slots, so as to better provide a dynamic access scheme and provide more choices for channel allocation for unlicensed users;
  • the false detection rate and the missed detection rate in the spectrum prediction process are improved, and the accuracy in the channel availability prediction process is improved.
  • FIG. 10 is a schematic structural diagram of a first embodiment of an apparatus for extracting a frequency resource occupancy pattern according to the present invention.
  • the apparatus for extracting a spectrum resource occupancy pattern includes: a first data collection unit 31, a sample matrix generation unit 33, and a pattern Analysis unit 35.
  • the first data collection unit 31 is configured to obtain first sampling data of the target spectrum, where the first sampling data includes existing service information, channel information, and channel state information of the target spectrum.
  • the sample matrix generating unit 33 is configured to extract the channel state information of all channels of the same service in each time slot from the first sampled data to generate a sample matrix.
  • the mode analysis unit 35 is configured to analyze the sample matrix, and extract the spectrum resource occupation mode combination of the target spectrum from the sample matrix.
  • the sample matrix generation unit 33 stores the collected first sample data, and performs a specific rule on the first data.
  • One sample data is processed to generate a sample matrix for mining the spectrum resource occupancy pattern.
  • the sample matrix is a two-dimensional zero-one matrix composed of time slots and channels.
  • the mode analysis unit 35 analyzes the sample matrix output from the sample matrix generation unit to extract effective spectrum resource occupancy mode combinations.
  • the mode output unit may output available spectrum resource occupancy mode combinations, including a source mode field, a target mode field, and a correlation coefficient field.
  • the correlation coefficient is a real number between 0 and 1, which represents the probability that the channel state of the next time slot is the channel state of the corresponding time slot in the target mode.
  • the spectrum resource occupancy mode extraction device extracts the combination of the frequency word resource occupancy mode
  • the sample matrix generating unit extracts the channel state information of all channels of the same service in each time slot to generate a sample matrix
  • the mode analysis unit analyzes the sample matrix to obtain the spectrum resource occupancy mode combination of the target spectrum, thereby performing spectrum prediction according to the correlation of different channels of the same service, increasing the access opportunities of unlicensed users to the spectrum, and reducing the spectrum prediction
  • the false detection rate and the missed detection rate in the process improve the accuracy of the frequency word prediction.
  • the apparatus for extracting the spectrum resource occupancy mode in the embodiment of the present invention may specifically be embodied as a circuit, an integrated circuit, or a chip.
  • the various units in the embodiments of the present invention may be integrated into one body, or may be deployed separately.
  • the above-mentioned units can be combined into one unit or further divided into multiple sub-units.
  • FIG. 11 is a schematic structural diagram of the second embodiment of the apparatus for extracting a spectrum resource occupancy pattern according to the present invention.
  • the sample matrix generating unit 33 may include : Threshold power subunit 331, channel state subunit 333, and sample matrix subunit 335.
  • the threshold power subunit 331 is configured to obtain the lowest value of the detection power of the selected channel of the service in the selected time slot from the first sampled data, and set the selection according to the lowest value of the detection power. The threshold power of a given channel.
  • the channel state subunit 333 is configured to: if the maximum sampling power of the selected channel of the service in the selected time slot is greater than the threshold power, the selected channel is in the selected time slot The channel state is an occupied state, otherwise the channel state of the selected channel in the selected time slot is an idle state.
  • the sample matrix subunit 335 is used to integrate the channel state information of all channels of the same service in each time slot into a two-dimensional sample matrix, and the channel state information corresponding to the occupancy state has an element bit of the two-dimensional sample matrix 1.
  • the channel state information corresponding to the idle state has an element bit of 0 in the two-dimensional sample matrix.
  • the pattern analysis unit 35 may include: an extracting subunit 351, an obtaining subunit 353, and a combining subunit 355.
  • the extraction subunit 351 is configured to extract the first spectrum resource occupancy mode whose number of occurrences exceeds the threshold value from the sample matrix.
  • the obtaining subunit 353 is configured to obtain, from the sample matrix, a second frequency-acquainted resource occupancy pattern associated with the first spectrum resource occupancy mode whose number of occurrences exceeds a threshold value.
  • the combining subunit 355 is configured to, if the correlation coefficient between the second frequency resource occupancy mode and the first frequency resource occupancy mode is greater than a set threshold, combine the first spectrum resource occupancy mode and the second spectrum resource occupancy mode And the correlation coefficient are combined into the spectrum resource occupancy mode Combination.
  • the device for extracting the frequency spectrum resource occupancy mode may further include: a mode output unit 37, configured to output the spectrum resource occupancy mode combination.
  • the sample matrix generation unit 33 stores the collected first sample data, and the threshold of the sample matrix generation unit 33
  • the power subunit 331 obtains the lowest value of the detected power of the selected channel of the service in the selected time slot from the first sampled data, and sets the threshold power of the selected channel according to the lowest value of the detected power If the maximum sampling power of the selected channel of the service in the selected time slot is greater than the threshold power, the channel state subunit 333 obtains the channel state of the selected channel in the selected time slot It is an occupied state, otherwise the channel state of the selected channel in the selected time slot is acquired as an idle state. Then the sample matrix subunit 335 can integrate the channel state information of all channels of the same service in each time slot into a two-dimensional sample matrix.
  • the acquiring sub-unit 353 acquires the first frequency resource occupancy pattern associated with the first spectrum resource from the sample matrix.
  • the second spectrum resource occupancy mode whose number of occurrences exceeds the threshold, if the correlation coefficient between the second spectrum resource occupancy mode and the first spectrum resource occupancy mode is greater than the set threshold, finally, the combining subunit 355 will The combination of the first spectrum resource occupancy mode, the second spectrum resource occupancy mode, and the correlation coefficient is the spectrum resource occupancy mode combination.
  • the mode output unit 37 can then output the combination of the frequency language resource occupation mode.
  • each subunit of the sample matrix generation unit extracts the channel state information of all channels of the same service in each time slot, A sample matrix is generated, and each subunit of the mode analysis unit analyzes the sample matrix to obtain a combination of spectrum resource occupancy modes, thereby performing spectrum prediction based on the correlation of different channels of the same service, increasing the access opportunities of unlicensed users to the spectrum Therefore, the false detection rate and the missed detection rate in the frequency error prediction process are reduced, and the accuracy of the frequency prediction is improved.
  • the apparatus for extracting the spectrum resource occupancy mode in the embodiment of the present invention may specifically be embodied as a circuit, an integrated circuit, or a chip.
  • the various units in the embodiments of the present invention may be integrated into one body, or may be deployed separately.
  • the above-mentioned units can be combined into one unit or further divided into multiple sub-units.
  • FIG. 12 is a schematic structural diagram of an embodiment of an apparatus for predicting frequent resources according to the present invention.
  • the The spectrum resource prediction device includes: a second data collection unit 41, a target matrix generation unit 43, and a channel prediction unit 45.
  • the second data acquisition unit 41 is configured to acquire second sampling data of the target spectrum, where the second sampling data includes real-time service information, channel information, and channel state information of the target spectrum.
  • the target matrix generating unit 43 is configured to extract the channel state information of all channels of the same service in each time slot from the second sample data to generate a target matrix.
  • the channel prediction unit 45 is configured to, when there is data in the target matrix that matches the first spectrum resource occupancy pattern in the combination of the spectrum resource occupancy pattern, according to the first spectrum resource occupancy pattern associated with the first spectrum resource occupancy pattern
  • the second spectrum resource occupancy mode is to perform sub-frame measurement on the channel state of the target spectrum in the future time slot.
  • the device for predicting spectrum resources may further include: a mode management unit 47, configured to obtain the combination of spectrum resource occupancy modes from the device 51 for extracting spectrum resource occupancy modes, and send the combination of spectrum resource occupancy modes to the channel prediction Unit 45.
  • a mode management unit 47 configured to obtain the combination of spectrum resource occupancy modes from the device 51 for extracting spectrum resource occupancy modes, and send the combination of spectrum resource occupancy modes to the channel prediction Unit 45.
  • the device for predicting frequency resources may further include: a decision output unit 49, configured to output the predicted channel state of the target spectrum in a future time slot.
  • the target matrix generation unit 43 stores the collected second sampling data of the target spectrum, and extracts all channels of the same service from the second sampling data. In the channel state information of each time slot, a target matrix is generated.
  • the mode management unit 47 may take the result of the mode output unit of the spectrum resource occupancy mode extraction device 51 as input, manage and store the combination of the frequency resource occupancy mode, and at the same time use it as the input of the channel prediction unit 45 for prediction, where the spectrum resource occupancy mode combination includes Source mode field, target mode field, correlation coefficient field, etc.
  • the channel prediction unit 45 uses the second spectrum resource occupancy mode associated with the first spectrum resource occupancy mode. Mode to predict the channel state of the target spectrum in the future time slot, for example: the next time slot. Finally, the decision output unit 49 outputs the result of the frequency word prediction: the channel state of the predicted target frequency ⁇ pu in the future time slot is shown as an idle state or an occupied state, where the prediction accuracy rate is the corresponding correlation coefficient field in the mode management unit.
  • the second data collection unit of this embodiment collects the second data of the target frequency i"
  • the target moment The array generation unit extracts the channel state information of all channels of the same service in each time slot from the second sampled data to generate a target matrix, and the channel prediction unit obtains the same service according to the spectrum resource occupancy mode combination extracted in the spectrum resource occupancy mode extraction device
  • the correlation between different channels of the target spectrum can predict the channel state of the target spectrum in the future time slot, which increases the access opportunities of unlicensed users to the spectrum, reduces the false detection rate and the missed detection rate in the frequency spectrum prediction process, and improves The accuracy of spectrum prediction is improved.
  • the spectrum resource prediction device of the embodiment of the present invention may be specifically represented as a circuit, an integrated circuit, or a chip.
  • the various units in the embodiments of the present invention may be integrated into one body, or may be deployed separately.
  • the above-mentioned units can be combined into one unit or further divided into multiple sub-units.
  • FIG. 13 is a schematic structural diagram of an embodiment of a frequency word frame detection system according to the present invention.
  • the spectrum prediction system includes: a frequency resource occupancy pattern extraction device 51 and a spectrum resource prediction device 53.
  • the spectrum resource occupancy mode extraction device 51 is configured to obtain first sampling data of a target spectrum, where the first sampling data includes existing service information, channel information, and channel state information of the target spectrum; In one sample of data, channel state information of all channels of the same service in each time slot is extracted to generate a sample matrix; and a combination of the frequency resource occupancy mode of the target spectrum is extracted from the sample matrix.
  • the spectrum resource prediction device 53 is configured to obtain second sampling data of the target spectrum, match the spectrum resource occupancy mode combination according to the second sampling data of the target spectrum, and compare the target frequency language according to the matching result. The channel state in the future time slot is predicted.
  • the spectrum resource prediction device is specifically configured to: collect second data of the target spectrum to obtain second sampling data, where the second sampling data includes real-time service information, channel information, and channel information of the target spectrum State information; from the second sampled data, extract the channel state information of all channels of the same service in each time slot to generate a target matrix; if the target matrix contains the first spectrum in combination with the spectrum resource occupancy mode For data matching the resource occupancy pattern, the channel state of the target frequency word in the future time slot is predicted according to the second spectrum resource occupancy pattern associated with the first frequency resource occupancy pattern.
  • the spectrum resource occupancy mode extraction device 51 collects the first data of the target frequency language, and after obtaining the first sample data, extracts all channels of the same service in each time slot from the first sample data. Generate a sample matrix; extract the spectrum resource occupation mode combination of the target spectrum from the sample matrix. Then the spectrum resource prediction device 53 collects the second data of the target spectrum to obtain second sampling data, and extracts from the second sampling data The channel state information of all channels of the same service in each time slot is used to generate a target matrix; if there is data in the target matrix that matches the first spectrum resource occupancy mode in the frequency resource occupancy mode combination, then the target matrix is The second spectrum resource occupancy mode associated with the first spectrum resource occupancy mode predicts the channel state of the target spectrum in the future time slot.
  • the spectrum resource occupancy mode extraction device 51 in this embodiment can use any of the spectrum resource occupancy mode extraction devices in the above-mentioned spectrum resource occupancy mode extraction device embodiments, and the spectrum resource prediction device 53 can be implemented by the aforementioned spectrum resource prediction device. Any kind of spectrum resource prediction device in the example.
  • the telephony sub-frame detection system in this embodiment may be a central telephony controller in a CR network, and it may be set on equipment such as a base station.
  • the apparatus for extracting the spectrum resource occupancy mode of this embodiment can obtain the combination of frequency resource occupancy mode from the first data of the spectrum resource, and the spectrum resource predicting apparatus combines the frequency resource occupancy mode combination according to the second data of the target frequency. Matching is performed to predict the channel state of the target spectrum in the future time slot, which increases the access opportunities of unlicensed users to the spectrum, reduces the false detection rate and the missed detection rate in the frequency borrowing prediction process, and improves The accuracy of frequency prediction.
  • a person of ordinary skill in the art can understand that all or part of the steps in the above method embodiments can be implemented by a program instructing relevant hardware.
  • the foregoing program can be stored in a computer readable storage medium. When the program is executed, the program is executed. Including the steps of the foregoing method embodiment; and the foregoing storage medium includes: various media capable of storing program codes, such as ROM, RAM, magnetic disk, or optical disk.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Computer Security & Cryptography (AREA)
  • Mobile Radio Communication Systems (AREA)

Description

频谱预测方法、 装置和系统
本申请要求于 2009年 8月 28 日 提交中 国 专利局、 申请号为 200910171727. 5、 发明名称为 "频谱预测方法、 装置和系统" 的中国专利申 请的优先权, 其全部内容通过引用结合在本申请中。 技术领域
本发明实施例涉及通信技术领域, 特別涉及一种频谱预测方法、 装置和 系统。 背景技术
随着无线通信技术的迅猛发展, 作为无线通信的最宝贵资源, 频谱的需 求日趋紧张。 目前的频谙资源多是由国家统一分配给各频谙需求机构。 多项 研究表明, 在已分配出去的频谱资源中, 大多数的频段利用率均低于 25%, 有 部分频段资源的利用率甚至低于 10%。 但是一些新兴的无线业务和无线设备对 频谱的需求却无法得到满足。 如何有效利用频谱资源已经成为无线通信领域 的热点问题。基于软件无线电技术发展而来的认知无线电(Cogni t ive Rad io , 简称: CR )技术应运而生。 认知无线电技术充分考虑现有频谱资源的低利用 率以及无线通信技术智能化的演进路线, 通过对周围环境的感知, 根据特定 的学习和决策算法, 自适应的改变应用参数, 动态检测及选择可以有效利用 的空闲频谱。 动态频谱接入(Dynamic Spec t rum Acces s , 简称: DSA )是 CR 技术领域中的一个重要方向, DSA技术允许多个系统共享一个频段, 在保证不 影响其它系统通信的前提下, 允许后接入的系统占用频率。 DS A技术的前提条 件是频谱感知, 即寻找有效的空闲频段, 频谱感知是未授权用户 (Secondary User , 简称: SU )接入服务的先决条件。 目前频谱感知技术由于感知时间要 求较短、 需要感知的频谱带宽较大等因素, 存在着较大困难。
采用频谱测量技术及预测方法可以解决频谱感知技术的问题。 频谙测量 技术是认知无线电技术通向实际应用的基础技术之一。
现有的频谱测量技术多采用一阶马尔可夫过程对信道状态进行预判, 其 基本思想是: 假设当前时隙的信道状态只与上一个时隙的信道状态有关; 信 道状态的转换发生在每个时隙的最初时刻。
发明人在实现本发明的过程中至少发现现有技术至少存在如下问题: 一阶马尔可夫模型的频傅子员测方案的下一时刻的信道状态只与当前信道 状态相关, 而与各个信道的历史时隙的状态相关性不大; 在分析过程中并没 有特别考虑信道间的相关性,多信道的情况一般假设各个信道之间独立。 因 此, 在频谱预测过程中对未来时隙的频谱状态预判的准确率较低、 误检率和 漏检率较高、 频谱空洞选择的低可靠性。 发明内容
本发明实施例提供一种频谱预测方法、 装置和系统, 用以降低频谱预测 过程中的误检率和漏检率, 提高频谱预测结果的准确度。
本发明实施例提供一种频旙预测方法, 包括: 。
获取目标频谱的第一釆样数据, 所述第一采样数据包括所述目标频谱的 已有的服务信息、 信道信息和信道状态信息;
从所述第一采样数据中, 提取同一服务全部信道在各个时隙的信道状态 信息, 生成样本矩阵;
从所述样本矩阵中提取所述目标频谱的频谱资源占用模式组合; 获取所述目标频谱的第二釆样数据 , 根据所述目标频谱的第二采样数据 , 对所述频谱资源占用模式组合进行匹配, 根据匹配结果对所述目标频谙在未 来时隙的信道状态进行预测。
本发明实施例又提供一种频谱资源占用模式提取装置, 包括:
第一数据采集单元, 用于获取目标频谱的第一采样数据, 所述第一釆样 数据包括所述目标频谱的已有的服务信息、 信道信息和信道状态信息;
样本矩阵生成单元, 用于从所述第一采样数据中, 提取同一服务全部信 道在各个时隙的信道状态信息, 生成样本矩阵;
模式分析单元, 用于对所述样本矩阵进行分析, 从所述样本矩阵中提取 所述目标频谱的频谱资源占用模式组合。
本发明实施例再提供一种频谱资源预测装置, 包括:
第二数据采集单元, 用于获取目标频 i "的第二采样数据, 所述第二采样 数据包括所述目标频谱的实时的服务信息、 信道信息和信道状态信息; 目标矩阵生成单元, 用于从所述第二采样数据中, 提取同一服务全部信 道在各个时隙的信道状态信息, 生成目标矩阵;
信道预测单元, 用于当所述目标矩阵中存在与所述频谱资源占用模式组 合中的第一频谱资源占用模式匹配的数据时, 根据与所述第一频谱资源占用 模式相关联的第二频谱资源占用模式, 对所述目标频谱在未来时隙的信道状 态进行预测。
本发明实施例还提供一种频谱预测系统, 包括:
频谱资源占用模式提取装置, 用于获取目标频谱的第一采样数据, 所述 第一采样数据包括所述目标频谱的已有的服务信息、 信道信息和信道状态信 息; 从所述第一采样数据中, 提取同一服务全部信道在各个时隙的信道状态 信息, 生成样本矩阵; 从所述样本矩阵中提取所述目标频谙的频谱资源占用 模式组合;
频语资源预测装置, 用于获取所述目标频谱的第二采样数据, 根据所述 目标频谱的第二采样数据, 对所述频谱资源占用模式组合进行匹配, 根据匹 配结果对所述目标频谱在未来时隙的信道状态进行预测。
本发明实施例提供的频借预测方法、 装置和系统, 从目标频谱的第一数 据中, 可以获取目标频谱的频谱资源占用模式组合, 利用目标频谱的第二数 据, 对频谱资源占用模式组合进行匹配后, 根据匹配结果对目标频谱在未来 时隙的信道状态进行预测, 可以降低频谱预测过程中的误检率和漏检率, 提 高频谱预测的准确度。 附图说明
为了更清楚地说明本发明实施例中的技术方案, 下面将对实施例或现有 技术描述中所需要使用的附图作筒单地介绍, 显而易见地, 下面描述中的附 图是本发明的一些实施例, 对于本领域普通技术人员来讲, 在不付出创造性 劳动性的前提下, 还可以根据这些附图获得其他的附图。
图 1为本发明实施例中信道状态判定方法的示意图;
图 2为本发明实施例中信道空闲延时(Channe l Vacancy Dura t ion, 简称: CVD)序列生成的示意图;
图 3为本发明实施例中信道空闲发生时隙 (Channel Vacancy Appearance Interval, 简称: CVAI )序列生成的示意图;
图 4为本发明实施例中 G SM900上行链路的 CVD分布曲线;
图 5为本发明实施例中 G SM900和 G SM 1800上行链路 S CR统计曲线图; 图 6为本发明实施例中部分服务的 SC的示意图;
图 7为本发明频谱预测方法笫一实施例的流程图;
图 8为本发明频谱预测方法第二实施例的示意图;
图 9为本发明频谱预测方法第三实施例的示意图;
图 10为本发明频谱资源占用模式提取装置第一实施例的结构示意图; 图 11为本发明频谱资源占用模式提取装置第二实施例的结构示意图; 图 12为本发明频谱资源预测装置实施例的结构示意图;
图 13为本发明频谱预测系统实施例的结构示意图。 具体实施方式
下面将结合本发明实施例中的附图, 对本发明实施例中的技术方案进行 清楚、 完整地描述, 显然, 所描述的实施例是本发明一部分实施例, 而不是 全部的实施例。 基于本发明中的实施例, 本领域普通技术人员在没有作出创 造性劳动前提下所获得的所有其他实施例, 都属于本发明保护的范围。
本发明实施例适用的无线接入制式包括: 全球移动通信系统 (Global
System for Mobile Communications, GSM ) 、 宽带码分多址 ( Wideband-Code Division Multiple Access, WCDMA )、 时分同步码分多址接入(Time Divis ion - Synchronized Code Division Multiple Access , TD-SCDMA) , 码分多址 (Code-Division Multiple Access , CDMA )、全球 皮互联接入( Wor ldwide Interoperability for Microwave Access , WIMAX )、 无线局域网 ( Wireless Local Area Network, WLA ) 、 长期演进 (Long Term Evolution , LTE)等。
本发明实施例基于实际采样数据进行分析, 在中国南方的四个地点 (两 个城市环境, 两个农村环境)获取分析数据, 采集持续时间为 7天, 目标频谱 范围为 20匪 z-3GHz , 采样点的频 ΐ普分布间隔为 0.2MHz, 采样点的时间分布间 隔即时隙为 75s。
首先, 通过对每一时隙的采样频率进行提取分析, 可以获得当前时隙的 信道状态。 信道状态定义为空闲状态 ( idle )和占用状态 (busy) , 其中空 闲状态用 "0" 表示, 占用状态用 "1" 表示。 图 1为本发明实施例中信道状态 判定方法的示意图,如图 1所示,在无有效信号的情况下,测得的干扰(noise ) 功率的最高值为最高采样频率 11, 千扰功率的最低值为最低采样频率 12, 图 中最高采样频率 11与最低采样频率 12的间隔接近 3dB μ V, 而在某一信道的采 样频率最低值均为干扰, 因此可定义干扰的门限功率。 例如: 在某一频段范 围内, 取该信道实测的最小功率值, 设为 min Ρ, 定义干扰的门限功率 K为公 式( 1 ):
Figure imgf000007_0001
如果某一时隙内对该信道的采样功率最大值超过了该门限功率 Κ, 则可认 为该信道在这一时隙的信道状态为占用状态; 否则, 可认为该信道在这一时 隙的信道状态是空闲状态。
图 2为本发明实施例中 CVD序列生成的示意图, 如图 2所述, 信道在每个时 隙的采样功率的最大值在设置的功率门限 Κ以上的, 信道状态为 1 ( busy) , 在 K以下的信道状态为 0 ( idle ) 。 图 2中序列的 CVD为信道状态信息(Channel state information, 简称: CSI )序列中连续 "0" ( idle )的数目。 其中 CVD 的值越大, 表明该信道被占用的时间越少, 该信道的可利用价值越大。
图 3为本发明实施例中 CVAI序列生成的示意图, 已上图中的 CSI序列为例, 可以得出图 3中序列的 CVAI, 其中 CVAI为 CSI序列中每一次连续 "0" ( idle ) 出现的间隔。
图 4本发明实施例中 GSM900上行链路的 CVD分布曲线, 如图 4所示, 以 GSM900上行链路为例, 对全部样本数据进行分析后, 统计 CVD的分布曲线, 符 合如下的公式( 2 ) :
y = a + be~cx ( 2 ) 在图 4中的 GSM900上行链路的 CVD的分布曲线中, a=0.000947 , b=l.671621, c=l.079681, r2=0.993586。 其中参数 r2是公式( 2 )计算所得数 据与实际数据之间的相似度, r2的值越接近于 1 , 表明该公式统计的结果越接 近于实际值。
在公式(2 ) 中各服务的 CVD分布曲线中 a b c的参数表可以通过统计得 到, 其中, 服务可以是指无线通信中的各种类型的业务, 例如: 可以表 1各服 务的 CVD分布曲线参数表中所示。
表 1 各服务的 CVD分布曲线参数表
Figure imgf000008_0001
随后, 对为服务拥塞率 (Servi ce Conges t ion Rate , 简称: SCR )进行 分析, 某服务在 t时刻的 SCR的计算方法如下公式(3 ) :
SCR (t, S) =该服务 t时刻处于占用状态的信道数 /该服务全部信道数 ( 3 ) SCR可以反映各种服务的拥堵规律, 图 5为本发明实施例中 GSM900和 GSM1800上行链路 SCR统计曲线图, 从图 5中可以得出: 该 GSM900 GSM1800两 个服务的 SCR的分布具有每日重复性, 同样, 其他服务的 SCR同样具有每曰重 复性, 在这里并不限制服务的种类。 然后, 分析 SCR序列的频镨相关性(Spectrum Corre lat ion, 简称: SC ) 图 6为本发明实施例中部分服务的 SC的示意图。 由图 6可知, TV频段和 GSM频段 的同一服务的不同信道之间频傅相关性(SC ) 的值可达 90%, 尤其是 GSM900上 行链路(Upl ink ) 、 GSM1800上行链路 ( Upl ink )及部分 TV频段的 SC的值更是 高达 95%以上。
综上可以得出: 同一服务的 CVD随时间的变化呈规则曲线分布, 曲线方程 为: _y = + be— 以一天为周期, 同一服务的 SCR具有较强的相似性; 同一 服务的不同信道之间在时、 空、 频等方面具有较强的频谱相关性。 因此, 可 以根据频谱资源中信道的历史状态对该信道的未来时隙的状态进行预测。 本 发明实施例的主要思想是: 根据各个服务的频谙资源中包括的各个信道的第 一数据, 对各个服务未来时隙的信道状态进行预测。
图 7为本发明频借预测方法第一实施例的流程图, 如图 7所示, 该频谱预 测方法包括:
步骤 101、 获取目标频谱的第一采样数据, 所述第一釆样数据包括所述目 标频谱的已有的服^言息、 信道信息和信道状态信息。
获取目标频谱的第一采样数据的方法具体包括: 对目标频谱的第一数据 进行采集, 得到第一采样数据, 所述第一数据包括历史数据, 例如: 第一数 据为该目标频谱在某一个历史时间段内已有的频谱资源数据。 本发明实施例 中以定长时间周期, 每隔一定间隔为一个时隙, 对目标频谱的第一数据进行 采集, 其中定长时间周期、 时隙可以根据具体的测量能力选定。 对数据采集 的结果进行收集并存储, 形成初始的第一采样数据, 该第一采样数据可包括 釆集的目标频谱中的所有服务的服务(Service )信息、 每个服务的所有的信 道(Channe l )信息及每个信道的信道状态信息等。
步骤 102、 从所述第一采样数据中, 提取同一服务全部信道在各个时隙的 信道状态信息, 生成样本矩阵。
对第一釆样数据进行处理时, 以同一服务为单位, 提取该服务包含的全 部信道的各个时隙的值, 作为用于频谱资源占用模式挖掘的二维样本矩阵, 其中可以假设 "0" 为信道空闲状态, " 为信道占用状态, 得到的二维样 本矩阵则是关于信道和时隙的二维 "0/1 " 矩阵。 其中, 生成样本矩阵的过程具体可以包括:
从所述第一采样数据中, 获取所述服务的选定信道在选定时隙的检测功 率的最低值, 根据所述检测功率的最低值设置所述选定信道的门限功率。
若所述服务的所述选定信道在所述选定时隙的采样功率最大值大于所述 门限功率, 则所述选定信道在所述选定时隙的信道状态为占用状态, 否则所 述选定信道在所述选定时隙的信道状态为空闲状态。
将同一服务全部信道在各个时隙的信道状态信息, 整合成样本矩阵。 其 中所述占用状态对应的信道状态信息在样本矩阵的元素位为 1 (也可以为其他 常数) , 所述空闲状态对应的信道状态信息在样本矩阵的元素位为 G (也可以 为其他常数) 。
例如: 假设选定时隙的检测功率的最低值为 min Ρ, 则选定信道的门限功 率 Κ的定义方法, 可以参照上述的公式(1 ) K= min P +3dBuV0 第一采样数据 转换为二维样本矩阵过程中矩阵元素的判定方法, 具体为: 将信道在特定时 隙的采样功率最大值与门限功率 K相比较, 如果大于 K, 则认为信道在该时隙 被占用, 在二维样本矩阵中, 该信道在该时隙对应的元素位为 "Γ ; 否则, 如果小于 Κ , 则认为信道在该时隙空闲, 在二维样本矩阵中, 该信道在该时隙 对应的元素位为 " 0" 。 在二维样本矩阵中, 采用 " 0" 表示空闲状态, 采用 "1" 占用状态, 但并不限制表示空闲状态、 占用状态的数值, 不排除采用其 他数值的情况。
步骤 103、 从所述样本矩阵中提取所述目标频谙的频谱资源占用模式组 合。
对生成的样本矩阵进行分析, 可以提取用于预测的频谱资源占用模式组 合, 具体包括:
首先, 从所述样本矩阵中提取出现的次数超过门限值的第一频谱资源占 用模式。 样本矩阵中的数据相当于一个二维数组, 在样本矩阵中提取一个频 谱资源占用模式, 就是从样本矩阵对应二维数组中提取一个子集。 为了方便 叙述, 该步骤中提取出的频谱资源占用模式称为第一频谱资源占用模式。 如 果第一频谙资源占用模式在样本矩阵中出现的次数超过预先定义的门限值, 则该第一频豫资源占用模式可以看作一个有效的频谱资源占用模式。 随后, 从所述样本矩阵中获取与所述第一频谱资源占用模式相关联的出 现的次数超过门限值的第二频谱资源占用模式。 在样本矩阵中, 查找与该第 一频谱资源占用模式相关联的其他有效的频语资源占用模式, 与该第一频谱 资源占用模式相关联的其他有效的频谱资源占用模式称为第二频谱资源占用 模式。
在本实施例中, 所述第一频谱资源占用模式与所述第二频谱资源占用模 式之间的关联关系为源模式与 标模式之间的关系, 可以包括以下示例: 示例一、 如果所述第一频谱资源占用模式包含在所述第二频谱资源占用 模式中, 则所述第一频谱资源占用模式为源模式, 所述第二频谱资源占用模 式为目标模式。 也就是: 如果样本矩阵的某个频讲资源占用模式被包含在其 他模式中, 则被包含的频谱资源占用模式称为源模式, 包含源模式的频谱资 源占用模式称为目标模式。
示例二、 如果所述第一频谱资源占用模式出现后, 每隔一定个数的时隙 出现所述第二频谱资源占用模式, 则所述第一频语资源占用模式为源模式, 所述第二频谱资源占用模式为目标模式。 也就是: 如果样本矩阵的某个频谙 资源占用模式出现后的每隔 Ν个时隙, 与其关联的另外一个模式出现, 则先出 现的频谱资源占用模式称为源模式, 与其关联的另一个频谱资源占用模式称 为目标模式。
若所述第二频谱资源占用模式与第一频谱资源占用模式的相关系数大于 设定阐值, 则将所述第一频谱资源占用模式、 第二频谱资源占用模式和相关 系数组合为所述频谱资源占用模式组合。 其中, 第一频谱资源占用模式为源 模式, 第二频谱资源占用模式为目标模式, 可以将所述源模式与所述目标模 式的相关系数定义为: 在样本矩阵中, 所述源模式和目标模式关联出现的次 数与所述源模式出现的总次数的比值。 但不限于该定义, 可以是其他的定义 方式。 如果目标模式与源模式的相关系数达到某一特定阔值以上, 则可以将 源模式和对应的目标模式及相关系数等信息作为一个有效的频谱资源占用模 式组合。
保存上述的频谱资源占用模式組合提取的结果。 其中频譜资源占用模式 组合中的数据结构, 可以包括以下字段: 源模式字段: 二维数组, 存储用于频谱预测的检测规则; 目标模式字段: 二维数组, 存储与源模式相关联的频谱资源占用模式。 如杲当前时隙的信道状态矩阵与其对应的源模式相匹配, 则可用该目标模式 对下一时隙的信道状态进行预测;
相关系数字段: 位于 "0" 和 " 1" 之间的实数, 表示下一时隙的信道状 态是目标模式中对应时隙的信道状态的概率。 计算方法为: 相关系数=源模式 出现后的目标模式的出现次数 /源模式在样本中的出现次数。
步骤 104、 获取所述目标频谱的第二采样数据, 根据所述目标频谱的第二 采样数据, 对所述频镨资源占用模式組合进行匹配, 根据匹配结杲对所述目 标频谱在未来时隙的信道状态进行预测。
具体包括:
首先、 对所述目标频谱的第二数据进行采集, 得到第二采样数据, 所述 第二采样数据可以包括某个时刻或时间段的服务信息、 信道信息和信道状态 信息, 比如包括所述目标频谱的实时的服务信息、 信道信息和信道状态信息。 其中目标频谱的第二数据可以是某个时刻或时间段的频谱资源数据, 比如该 目标频谱在当前的一个时间段内的实时的频谱资源数据。 对需要预测的目标 频谱的第二数据进行采样, 可以获得目标频谱当前时隙的各个信道的信道状 态信息, 对数据采集的结果进行收集并存储, 形成初始的第二采样数据, 该 第二采样数据包括服务(Service )信息、 信道(Channel )信息、 信道状态 信息等。
然后、 从所述第二采样数据中 , 提取同一服务全部信道在各个时隙的信 道状态信息, 生成目标矩阵。
对第二采样数据进行处理时, 以同一服务为单位, 提取该服务包含的全 部信道的当前各时隙的值作为二维目标矩阵, 该矩阵作为分析并预测下一时 隙信道状态的样本空间; 可以個殳 "0" 为信道空闲状态, 1为信道占用状态, 得到的二维样本矩阵则是关于信道和时隙的二维 "0/ 矩阵。
最后、 若所述目标矩阵中存在与所述频谱资源占用模式组合中的第一频 谱资源占用模式匹配的数据, 则根据与所述第一频谱资源占用模式相关联的 第二频谘资源占用模式, 对所述目标频谱在未来时隙的信道状态进行预测。 例如: 利用第一频谱资源占用模式(源模式)对目标矩阵进行匹配, 如果某 个第一频镨资源占用模式可以匹配目标矩阵中的某部分数值, 视为第一频谱 资源占用模式匹配成功, 停止分析过程并输出该第一频谱资源占用模式。 然 后, 将匹配成功的第一频谱资源占用模式作为输入, 利用与该第一频谱资源 占用模式相关联的第二频谱资源占用模式(目标模式)对相应位置的信道状 态进行预判; 输出预判结果, 该预判结果即某一个或者几个信道在接下来的 时隙中的信道状态, 即与第二频谱资源占用模式对应位置相一致的信道状态, 相一致的几率为相关系数个百分比。
频谱资源预判过程结束后, 未授权用户可以根据预判的结果选择可用的 信道进行选择性接入, 若所述目标频谱在未来时隙的一个信道的状态为空闲 状态, 则在所述未来时隙选择所述信道接入。
本实施例从频普资源的第一数据中, 可以获取频谱资源占用模式组合, 从而根据同一服务的不同信道的相关性进行频傳预测, 增大了未授权用户对 频谱的接入机会, 降低了频谱预测过程中的误检率和漏检率, 提高了频借预 测的准确度和频谱空洞选择的可靠性, 其中频谱空洞是指目前授权用户的频 谱中没有被占用的临时可用频段。
图 8为本发明频谱预测方法第二实施例的示意图, 在本发明频语预测方法 第一实施例的基础上, 假设频谱资源占用模式组合中的频谱资源占用模式是 包含关系。 在频谱资源占用模式的采集过程中, 得到频谱资源占用模式 P1和 Ρ2 , 如图 8所示, P1和 Ρ2的关联关系为包含关系, 即 P1是 Ρ2的子集。 其中 P1为 源模式, Ρ2为目标模式。 源模式 P1为 2 x 3矩阵, 目标模式 Ρ2为 2 X 4矩阵。 假 设 P1与 Ρ2的相关系数为 95% , 即 P1每出现 100次, 同时 Ρ2出现的次数为 95次。 本实施例中频谱预测的目标是 时刻的信道状态,该频谙预测方法具体过程 包括: 频谱资源占用模式提取过程和频谱资源预判过程。
其中, 频语资源占用模式提取过程如下:
以定长时间周期, 每隔一定间隔为一个时隙, 对目标频傳的第一数据进 行采集;以同一服务为单位, 提取该服务包含的全部信道的各个时隙的值, 得 到样本矩阵, 样本矩阵的结构例如为图 8所示的矩阵的 列至 t5列二维 "0/Γ 矩阵, 从该二维 "0/1" 矩阵中提取源模式 Pl、 目标模式 P2, 以及 PI与 P2的相 关系数和关联关系等组成频谱资源占用模式组合。 具体可以参照本发明频 预测方法第一实施例一中的具体描述, 在此不再赘述。
频谱资源预判过程如下:
先扫描目标频谱, 获取当前时隙的各个信道的信道状态信息。 信道状态 的判断方法参照本发明频傳预测方法第一实施例的相关描述与公式(1 ) 。
存储扫描结果并对扫描结果进行分析, 生成目标矩阵。 目标矩阵的结构 例如为图 8中矩阵的 tn3列至 列的二维 " 0/ 矩阵, 假设未来时隙为 tn+1时隙, 该频谱预测的目的是预测 tn+1时隙的各信道状态。
然后利用源模式 P1中的字段作为检测规则, 匹配目标矩阵。 在本实施例 的图 8中, 源模式 P1在 t„-2时隙至 tj†隙的列与的 ch4和 ch5行组成的矩阵匹配 (可以是完全匹配, 也可以是一定精度的匹配) 。 此时源模式 P1与当前时隙 的信道状态匹配成功 , 根据与源模式 P1相关联的目标模式 P2对下一个时隙的 信道 ch4和 ch5的状态进行预判。 由于 P2的第四列的信息分别为 0和 1, 即由 tn2 至 tn+1列与 ch4和 ch5行组成的矩阵的第四列与目标模式 P2的第四列完全一致的 可能性有 95°/。。 因此, 可得到预测结果: ch4在 tn+1时隙的信道状态为 0 ( idle ) 的可能性等于相关系数 95%, ch5在 tn+1时隙的信道状态为 1 ( busy )的可能性等 于 95%。
在 tn+1时隙, 未授权用户可以接入空闲状态 (i dle ) 的信道, 因此根据上 述的预测结果, 可以选择信道 ch4作为未授权用户在 t„+1时隙动态接入的信道。
本实施例从频谱资源的第一数据中, 可以获取频谱资源占用模式组合, 从而根据同一服务的不同信道的相关性进行频语预测, 增大了未授权用户对 频谱的接入机会; 在信道可用性预测过程中具有较高的准确度, 例如: 对于 GSM频段和 TV频段的预测的准确度尤为突出, 降低了频豫预测过程中的误检率 和漏检率。
图 9为本发明频谱预测方法第三实施例的示意图, 如图 9所示, 在本发明 频谱预测方法第一实施例的基础上, 假设频谱资源占用模式组合中的频 ΐ普资 源占用模式是连续关系。 在频谱资源占用模式的采集过程中, 得到频谙资源 占用模式 Ρ3和 Ρ4, 如图 9所示, Ρ3和 Ρ4的关联关系为连续关系, 即每次 Ρ3出现 后隔 Ν个时隙 Ρ4出现, 本实施例中 Ν=2。 其中 Ρ3为源模式, Ρ4为目标模式。 源 模式 P3为 2 χ 3矩阵, 目标模式 P4为 2 X 4矩阵 £ P3与 P4的相关系数为 90%, 即 P3每出现 100次之中, 在每次出现 P3之后隔两个时隙出现 P2的次数为 90次。 。 本实施例中频谱预测的目标是 tn+1时刻的信道状态,该频谱预测方法具体过程 包括: 频谱资源占用模式提取过程和频谱资源预判过程。
其中, 频谱资源占用模式提取过程如下:
以定长时间周期, 每隔一定间隔为一个时隙, 对目标频谱的第一数据进 行采集; 以同一服务为单位, 提取该服务包含的全部信道的各个时隙的值, 得到样本矩阵,样本矩阵的结构例如为图 9所示的矩阵的 列至 列二维 "0/1" 矩阵, 从该二维 "0/ 矩阵中提取源模式 P3、 目标模式 P4、 P3与 P4的相关系 数和关联关系等组成频 i普资源占用模式组合。 具体可以参照本发明频 i普预测 方法第一实施例一中的具体描述, 在此不再赘述。
从频谙资源占用模式组合可以获得源模式 P3和目标模式 P4为连续关系, 源模式 P3出现后每两个时隙目标模式 P4出现, 源模式 P3为 2 X 3矩阵, 目标模 式 P4为 2 X 3矩阵。 源模式 P3与目标模式 P4的相关系数为 90%等信息。
频谱资源预判过程如下:
先扫描目标频镨, 获取当前时隙的各个信道的信道状态信息。 信道状态 的判断方法参照本发明频语预测方法第一实施例的相关描述与公式(1 ) 。
存储扫描结果并对扫描结果进行分析, 生成目标矩阵。 该目标矩阵的结 构如图 9中矩阵的 tn-3列至 tn列的二维 "0/1" 矩阵, 假设未来时隙为 tntl时隙, 频语预测的目的是预测 ^+1时隙的各信道状态。
然后利用源模式 P3中的字段作为检测规则, 匹配目标矩阵。 在本实施例 中, 源模式 P 3与 t„-4时隙至 tn-2时隙的列与的 ch3和 cM行组成的矩阵匹配(可以 是完全匹配, 也可以是一定精度的匹配) 。 此时源模式 P3与当前时隙的信道 状态匹配成功, 根据与源模式 P3相关联的目标模式 P4对下一个时隙的信道 ch3 和 ch4的状态进行预判。 由于源模式 P3出现后每隔两个时隙目标模式 P4出现的 可能性等于相关系数 90%, 则此时 tn l时隙至 tn+3时隙中信道 ch 3和 ch4的信道状 态与 P4完全一致的可能性为 90%。 因此, 可得到预测结果: (:!^在^时隙至 3 时隙的信道状态分別为 0 ( idle) 、 0 ( idle) 、 1 (busy) 的可能性为 90%, ch4在 tn+1时隙至 tn+3时隙的信道状态分别为 1 (busy) 、 1 (busy) 、 0 ( idle) 的可能性为 90%。
在^+1时隙未授权用户可以接入空闲状态 (idl e ) 的信道, 因此根据预测 结杲, 可以选择信道 ch3作为未授权用户在 tn+1时隙动态接入的信道。
本实施例从频谱资源的第一数据中, 可以获取频谱资源占用模式组合, 从而根据同一服务的不同信道的相关性进行频傳预测, 增大了未授权用户对 频谱的接入机会; 通过源模式和目标模式之间的连续出现关系, 可以预测接 下来的多个未来时隙的信道状态, 从而可以更好的提供动态接入方案, 为未 授权用户的信道分配提供更多的选择; 降低了频谱预测过程中的误检率和漏 检率, 提高了在信道可用性预测过程中的准确度。
图 10为本发明频 i普资源占用模式提取装置第一实施例的结构示意图, 如 图 10所示, 该频谱资源占用模式提取装置包括: 第一数据采集单元 31、 样本 矩阵生成单元 33和模式分析单元 35。
其中, 第一数据采集单元 31, 用于获取目标频谱的第一采样数据, 所述 第一采样数据包括所述目标频谱的已有的服务信息、 信道信息和信道状态信 息。 样本矩阵生成单元 33, 用于从所述第一采样数据中, 提取同一服务全部 信道在各个时隙的信道状态信息, 生成样本矩阵。
模式分析单元 35, 用于对所述样本矩阵进行分析, 从所述样本矩阵中提 取所述目标频谱的频谱资源占用模式组合。
具体地, 第一数据采集单元 31对无线链路上的一段选定的目标频谙的第 一数据进行采集后, 样本矩阵生成单元 33存储采集到的第一采样数据, 通过 特定的规则对第一采样数据进行处理, 生成用于频谱资源占用模式挖掘的样 本矩阵, 该样本矩阵是以时隙和信道组成的二维零一矩阵。 然后模式分析单 元 35通过对样本矩阵生成单元中输出的样本矩阵进行分析, 提取有效的频谱 资源占用模式组合。 模式输出单元可以输出可用的频谱资源占用模式组合, 包括源模式字段, 目标模式字段及相关系数字段。 其中, 相关系数是位于 0和 1之间的实数, 表示下一时隙的信道状态是目标模式中对应时隙的信道状态的 概率。 计算方法为: 相关系数=源模式与目标模式在样本矩阵中关联出现的次 数 /源模式在样本矩阵中的出现次数。
在本实施例中, 频谱资源占用模式提取装置提取频语资源占用模式组合 的方法, 具体可以参照本发明频 预测方法第一、 第二实施例中的相关描述。 本实施例从频谱资源的第一数据中, 第一数据采集单元对目标频谱的第 一数据进行采集后, 样本矩阵生成单元提取同一服务全部信道在各个时隙的 信道状态信息, 生成样本矩阵, 模式分析单元对样本矩阵进行分析可以获取 目标频谱的频谱资源占用模式组合, 从而根据同一服务的不同信道的相关性 进行频谱预测, 增大了未授权用户对频谱的接入机会, 降低了频谱预测过程 中的误检率和漏检率, 提高了频语预测的准确度。
本发明实施例的频谱资源占用模式提取装置具体可以表现为电路、 集成 电路或芯片等。 本发明实施例的各个单元可以集成于一体, 也可以分离部署。 上述单元可以合并为一个单元, 也可以进一步拆分成多个子单元。
图 11为本发明频谱资源占用模式提取装置笫二实施例的结构示意图, 如 图 11所示, 在本发明频谙资源占用模式提取装置第一实施例的基础上, 样本 矩阵生成单元 33可以包括: 门限功率子单元 331、 信道状态子单元 333和样本 矩阵子单元 335。 其中门限功率子单元 331 , 用于从所述第一采样数据中, 获 取所述服务的选定信道在选定时隙的检测功率的最低值, 根据所述检测功率 的最低值设置所述选定信道的门限功率。 信道状态子单元 333 , 用于若所述服 务的所述选定信道在所述选定时隙的采样功率最大值大于所述门限功率, 则 所述选定信道在所述选定时隙的信道状态为占用状态, 否则所述选定信道在 所述选定时隙的信道状态为空闲状态。 样本矩阵子单元 335, 用于将同一服务 全部信道在各个时隙的信道状态信息, 整合成一个二维样本矩阵, 所述占用 状态对应的信道状态信息在所述二维样本矩阵的元素位为 1 , 所述空闲状态对 应的信道状态信息在所述二维样本矩阵的元素位为 0。
进一步地, 模式分析单元 35可以包括: 提取子单元 351、 获取子单元 353 和组合子单元 355。 其中提取子单元 351, 用于从所述样本矩阵中提取出现的 次数超过门限值的第一频谱资源占用模式。 获取子单元 353 , 用于从所述样本 矩阵中获取与所述第一频谱资源占用模式相关联的出现的次数超过门限值的 第二频谙资源占用模式。 组合子单元 355, 用于若所述第二频谙资源占用模式 与第一频谙资源占用模式的相关系数大于设定阈值, 则将所述第一频谱资源 占用模式、 第二频谱资源占用模式和相关系数組合为所述频谱资源占用模式 組合。
进一步地, 该频 i瞥资源占用模式提取装置还可以包括: 模式输出单元 37, 用于输出所述频谱资源占用模式组合。
具体地, 第一数据采集单元 31对无线链路上的一段选定的目标频谱的第 一数据进行采集后, 样本矩阵生成单元 33存储采集到的第一采样数据, 样本 矩阵生成单元 33的门限功率子单元 331从所述第一采样数据中, 获取所述服务 的选定信道在选定时隙的检测功率的最低值, 根据所述检测功率的最低值设 置所述选定信道的门限功率, 若所述服务的所述选定信道在所述选定时隙的 采样功率最大值大于所述门限功率, 信道状态子单元 333获取所述选定信道在 所述选定时隙的信道状态为占用状态, 否则获取所述选定信道在所述选定时 隙的信道状态为空闲状态。 然后样本矩阵子单元 335可以将同一服务全部信道 在各个时隙的信道状态信息, 整合成一个二维样本矩阵。
然后, 模式分析单元 35的提取子单元 351从样本矩阵中提取出现的次数超 过门限值的第一频 资源占用模式后, 获取子单元 353从样本矩阵中获取与第 一频谱资源占用模式相关联的、 出现的次数超过门限值的第二频谱资源占用 模式, 若所述第二频镨资源占用模式与第一频谙资源占用模式的相关系数大 于设定阈值, 最后, 组合子单元 355将所述第一频谱资源占用模式、 第二频谱 资源占用模式和相关系数组合为所述频语资源占用模式组合。 模式输出单元 37则可以输出所述频语资源占用模式组合。
本实施例从频谱资源的第一数据中 , 第一数据采集单元对目标频谱的第 一数据进行采集后, 样本矩阵生成单元的各个子单元提取同一服务全部信道 在各个时隙的信道状态信息, 生成样本矩阵, 模式分析单元的各个子单元对 样本矩阵进行分析可以获取频谱资源占用模式组合, 从而根据同一服务的不 同信道的相关性进行频谱预测, 增大了未授权用户对频谱的接入机会, 降低 了频錯预测过程中的误检率和漏检率, 提高了频 ϊ普预测的准确度。
本发明实施例的频谱资源占用模式提取装置具体可以表现为电路、 集成 电路或芯片等。 本发明实施例的各个单元可以集成于一体, 也可以分离部署。 上述单元可以合并为一个单元, 也可以进一步拆分成多个子单元。
图 12为本发明频谙资源预测装置实施例的结构示意图, 如图 12所示, 该 频谱资源预测装置包括: 第二数据采集单元 41、 目标矩阵生成单元 43和信道 预测单元 45。
其中第二数据采集单元 41 , 用于获取目标频谱的第二采样数据, 所述第 二采样数据包括所述目标频谱的实时的服务信息、 信道信息和信道状态信息。 目标矩阵生成单元 43 , 用于从所述第二釆样数据中, 提取同一服务全部信道 在各个时隙的信道状态信息, 生成目标矩阵。 信道预测单元 45, 用于当所述 目标矩阵中存在与所述频语资源占用模式组合中的第一频谱资源占用模式匹 配的数据时, 根据与所述笫一频谱资源占用模式相关联的第二频谱资源占用 模式, 对所述目标频谱在未来时隙的信道状态进行子贞测。
进一步地, 该频 i普资源预测装置还可以包括: 模式管理单元 47 , 用于从 频谱资源占用模式提取装置 51获取所述频谱资源占用模式组合, 将所述频谱 资源占用模式组合发送给信道预测单元 45。
再进一步地, 该频普资源预测装置还可以包括: 判决输出单元 49, 用于 输出预测的所述目标频谱在未来时隙的信道状态。
具体地, 第二数据采集单元 41对目标频谱的第二数据进行采集后, 目标 矩阵生成单元 43存储采集到的目标频谱的第二采样数据, 并从第二采样数据 中, 提取同一服务全部信道在各个时隙的信道状态信息, 生成目标矩阵。 模 式管理单元 47可以将频谱资源占用模式提取装置 51的模式输出单元的结果作 为输入, 管理并存储频谙资源占用模式组合, 同时作为信道预测单元 45的输 入进行预测, 其中频谱资源占用模式组合包括源模式字段、 目标模式字段及 相关系数字段等。 在所述目标矩阵中存在与所述频谱资源占用模式组合中的 第一频谱资源占用模式匹配的数据时, 信道预测单元 45根据与所述第一频谱 资源占用模式相关联的第二频谱资源占用模式, 对所述目标频谱在未来时隙 例如: 下一时隙的信道状态进行预测。 最后, 判决输出单元 49输出频语预测 的结果: 预测的目标频 Ϊ普在未来时隙的信道状态, 表现为空闲状态或者占用 状态, 其中预测准确率是模式管理单元中对应的相关系数字段。
在本实施例中, 频谱资源预测装置进行频 预测的方法, 具体可以参照 本发明频普预测方法第一、 第三实施例中的相关描述。
本实施例第二数据采集单元对目标频 i "的第二数据进行采集后, 目标矩 阵生成单元从第二采样数据中, 提取同一服务全部信道在各个时隙的信道状 态信息, 生成目标矩阵, 信道预测单元根据频谱资源占用模式提取装置中提 取的频谱资源占用模式组合, 获取同一服务的不同信道的相关性, 对目标频 谱在未来时隙的信道状态进行预测, 增大了未授权用户对频谱的接入机会, 降低了频普预测过程中的误检率和漏检率, 提高了频谱预测的准确度。
本发明实施例的频谱资源预测装置具体可以表现为电路、 集成电路或芯 片等。 本发明实施例的各个单元可以集成于一体, 也可以分离部署。 上述单 元可以合并为一个单元, 也可以进一步拆分成多个子单元。
图 13为本发明频语子贞测系统实施例的结构示意图, 如图 1 3所示, 该频谱 预测系统包括: 频谙资源占用模式提取装置 51和频谱资源预测装置 53。
其中, 频谱资源占用模式提取装置 51, 用于获取目标频谱的第一采样数 据, 所述第一采样数据包括所述目标频谱的已有的服务信息、 信道信息和信 道状态信息; 从所述第一采样数据中, 提取同一服务全部信道在各个时隙的 信道状态信息, 生成样本矩阵; 从所述样本矩阵中提取所述目标频谱的频谙 资源占用模式组合。 频谱资源预测装置 53, 用于获取所述目标频谱的第二采 样数据, 根据所述目标频谱的第二采样数据, 对所述频谱资源占用模式组合 进行匹配, 根据匹配结果对所述目标频语在未来时隙的信道状态进行预测。
进一步地, 频谱资源预测装置具体用于: 对所述目标频谱的第二数据进 行采集, 得到第二采样数据, 所述第二采样数据包括所述目标频谱的实时的 服务信息、 信道信息和信道状态信息; 从所述第二采样数据中, 提取同一服 务全部信道在各个时隙的信道状态信息, 生成目标矩阵; 若所述目标矩阵中 存在与所述频谱资源占用模式组合中的第一频谱资源占用模式匹配的数据, 则根据与所述第一频谙资源占用模式相关联的第二频谱资源占用模式, 对所 述目标频语在未来时隙的信道状态进行预测。
具体地, 频谱资源占用模式提取装置 51对所述目标频语的第一数据进行 釆集, 得到第一采样数据后, 从所述第一釆样数据中, 提取同一服务全部信 道在各个时隙的信道状态信息, 生成样本矩阵; 从所述样本矩阵中提取所述 目标频谱的频谱资源占用模式组合。 然后频谱资源预测装置 53对所述目标频 谱的第二数据进行采集, 得到第二采样数据, 从所述第二采样数据中, 提取 同一服务全部信道在各个时隙的信道状态信息, 生成目标矩阵; 若所述目标 矩阵中存在与所述频 i普资源占用模式组合中的第一频谱资源占用模式匹配的 数据, 则根据与所迷第一频谱资源占用模式相关联的第二频谱资源占用模式, 对所述目标频谱在未来时隙的信道状态进行预测。 具体方法可以参见本发明 频谱预测方法第一、 第二、 第三实施例中的相关描述。 本实施例中的频谱资 源占用模式提取装置 51可以采用上述的频谱资源占用模式提取装置实施例中 的任意一种频谱资源占用模式提取装置, 频谱资源预测装置 53可以采用上述 的频谱资源预测装置实施例中的任意一种频谱资源预测装置。
本实施例中的频语子贞测系统可以为 CR网络中的中心频语控制器, 可以设 置在基站等设备上。
本实施例本实施例频谱资源占用模式提取装置从频谱资源的笫一数据 中 , 可以获取频谘资源占用模式组合, 频谱资源预测装置根据目标频谙的第 二数据, 对频语资源占用模式组合进行匹配, 以对所述目标频谱在未来时隙 的信道状态进行预测, 增大了未授权用户对频谱的接入机会, 降低了频借预 测过程中的误检率和漏检率, 提高了频语预测的准确度。
本领域普通技术人员可以理解: 实现上述方法实施例的全部或部分步骤 可以通过程序指令相关的硬件来完成, 前述的程序可以存储于一计算机可读 取存储介质中, 该程序在执行时, 执行包括上述方法实施例的步骤; 而前述 的存储介质包括: R0M、 RAM、 磁碟或者光盘等各种可以存储程序代码的介质。
最后应说明的是: 以上实施例仅用以说明本发明的技术方案, 而非对其 限制; 尽管参照前述实施例对本发明进行了详细的说明, 本领域的普通技术 人员应当理解: 其依然可以对前述各实施例所记载的技术方案进行修改, 或 者对其中部分技术特征进行等同替换; 而这些修改或者替换, 并不使相应技 术方案的本质脱离本发明各实施例技术方案的范围。

Claims

权 利 要 求
1、 一种频语预测方法, 其特征在于, 包括:
获取目标频谱的第一釆样数据, 所述第一釆样数据包括所述目标频谱的 已有的服务信息、 信道信息和信道状态信息;
从所述第一采样数据中, 提取同一服务全部信道在各个时隙的信道状态 信息, 生成样本矩阵;
从所述样本矩阵中提取所述目标频谱的频谱资源占用模式组合; 获取所述目标频谱的第二釆样数据, 根据所述目标频谱的第二采样数据, 对所述频镨资源占用模式组合进行匹配, 根据匹配结果对所述目标频谱在未 来时隙的信道状态进行预测。
2、 根据权利要求 1所述的频普预测方法, 其特征在于, 所述获取目标频 谱的第一采样数据包括:
对目标频谱的第一数据进行釆集, 得到第一釆样数据, 所述第一数据包 括历史数据。
3、 根据权利要求 1所述的频 Ϊ普预测方法, 其特征在于, 所述从所述第一 采样数据中, 提取同一服务全部信道在各个时隙的信道状态信息, 生成样本 矩阵包括:
从所述第一采样数据中, 获取所述服务的选定信道在选定时隙的检测功 率的最低值, 根据所述检测功率的最低值设置所述选定信道的门限功率; 若所述服务的所述选定信道在所述选定时隙的采样功率最大值大于所述 门限功率, 则所述选定信道在所述选定时隙的信道状态为占用状态, 否则所 述选定信道在所述选定时隙的信道状态为空闲状态;
将同一服务全部信道在各个时隙的信道状态信息, 整合成样本矩阵。
4、 根据权利要求 1所述的频 Ϊ普预测方法, 其特征在于, 所述从所述样本 矩阵中提取所述目标频谱的频谱资源占用模式组合包括:
从所述样本矩阵中提取出现的次数超过门限值的第一频谱资源占用模 式;
从所述样本矩阵中获取与所述第一频谱资源占用模式相关联的出现的次 数超过门限值的第二频潘资源占用模式;
若所述第二频谱资源占用模式与第一频谱资源占用模式的相关系数大于 设定阔值, 则将所述第一频谱资源占用模式、 第二频谱资源占用模式和相关 系数组合为所述频豫资源占用模式组合。
5、 根据权利要求 1所述的频傳预测方法, 其特征在于, 所述获取所述目 标频谱的第二采样数据, 根据所述目标频谱的第二采样数据, 对所述频 ΐ普资 源占用模式组合进行匹配, 根据匹配结果对所迷目标频谱在未来时隙的信道 状态进行预测包括:
对所述目标频谱的第二数据进行采集, 得到第二采样数据, 所述第二采 样数据包括所述目标频谱的实时的服务信息、 信道信息和信道状态信息; 从所述第二采样数据中, 提取同一服务全部信道在各个时隙的信道状态 信息, 生成目标矩阵;
若所述目标矩阵中存在与所述频谱资源占用模式组合中的第一频谱资源 占用模式匹配的数据, 则根据与所述第一频谱资源占用模式相关联的第二频 谱资源占用模式, 对所述目标频谱在未来时隙的信道状态进行预测。
6、 根据权利要求 4或 5所述的频讲预测方法, 其特征在于,
所述第一频谱资源占用模式与所述第二频谙资源占用模式之间的关联关 系为源模式与目标模式之间的关系; 如果所述第一频谱资源占用模式包含在所述第二频谱资源占用模式中, 则所述第一频谱资源占用模式为源模式, 所述第二频谱资源占用模式为目标 模式; 或
如果所述第一频谙资源占用模式出现后, 每隔一定个数的时隙出现所述 第二频谱资源占用模式, 则所述第一频谱资源占用模式为源模式, 所述第二 频谱资源占用模式为目标模式。
7、 根据权利要求 1-5任一所述的频 i普预测方法, 其特征在于, 还包括: 若所述目标频谱在未来时隙的一个信道的状态为空闲状态, 则在所述未 来时隙选择所述信道接入。
8、 一种频谱资源占用模式提取装置, 其特征在于, 包括:
第一数据采集单元, 用于获取目标频谱的第一采样数据, 所述第一釆样 数据包括所述目标频谱的已有的服务信息、 信道信息和信道状态信息;
样本矩阵生成单元, 用于从所述第一采样数据中, 提取同一服务全部信 道在各个时隙的信道状态信息, 生成样本矩阵;
模式分析单元, 用于对所述样本矩阵进行分析, 从所述样本矩阵中提取 所述目标频语的频谱资源占用模式组合。
9、 根据权利要求 8所述的频谙资源占用模式提取装置, 其特征在于, 所 述样本矩阵生成单元包括:
门限功率子单元, 用于从所述第一釆样数据中, 获取所述服务的选定信 道在选定时隙的检测功率的最低值, 根据所述检测功率的最低值设置所述选 定信道的门限功率;
信道状态子单元, 用于若所述服务的所述选定信道在所述选定时隙的采 样功率最大值大于所述门限功率, 则所述选定信道在所述选定时隙的信道状 态为占用状态, 否则所述选定信道在所述选定时隙的信道状态为空闲状态; 样本矩阵子单元, 用于将同一服务全部信道在各个时隙的信道状态信息, 整合成样本矩阵。
10、 根据权利要求 8所述的频谱资源占用模式提取装置, 其特征在于, 所 述模式分析单元包括:
提取子单元, 用于从所述样本矩阵中提取出现的次数超过门限值的第一 频谱资源占用模式;
获取子单元, 用于从所述样本矩阵中获取与所述第一频谱资源占用模式 相关联的出现的次数超过门限值的第二频谱资源占用模式;
组合子单元, 用于若所述第二频谱资源占用模式与第一频谱资源占用模 式的相关系数大于设定阈值, 则将所述第一频谱资源占用模式、 第二频 ΐ普资 源占用模式和相关系数组合为所述频谱资源占用模式组合。
11、 一种频谙资源预测装置, 其特征在于, 包括:
第二数据采集单元, 用于获取目标频谱的第二采样数据, 所述第二采样 数据包括所述目标频谱的实时的服务信息、 信道信息和信道状态信息;
目标矩阵生成单元, 用于从所述第二釆样数据中, 提取同一服务全部信 道在各个时隙的信道状态信息, 生成目标矩阵;
信道预测单元, 用于当所述目标矩阵中存在与所述频谱资源占用模式组 合中的第一频谱资源占用模式匹配的数据时, 根据与所述第一频谱资源占用 模式相关联的第二频镨资源占用模式, 对所述目标频谱在未来时隙的信道状 态进行预测。
12、 根据权利要求 11所述的频谱资源预测装置, 其特征在于, 还包括: 模式管理单元, 用于从频谱资源占用模式提取装置获取所述频谱资源占 用模式组合, 将所述频谱资源占用模式组合发送给所述信道预测单元。
13、 一种频博预测系统, 其特征在于, 包括:
频谱资源占用模式提取装置, 用于获取目标频谱的第一采样数据, 所述 第一采样数据包括所述目标频谱的已有的服务信息、 信道信息和信道状态信 息; 从所述第一采样数据中, 提取同一服务全部信道在各个时隙的信道状态 信息, 生成样本矩阵; 从所述样本矩阵中提取所迷目标频谙的频谱资源占用 模式组合;
频谱资源预测装置, 用于获取所述目标频谱的第二采样数据, 根据所述目标 频谱的第二采样数据 , 对所述频谱资源占用模式组合进行匹配, 根据匹配结 果对所述目标频谱在未来时隙的信道状态进行预测。
PCT/CN2010/076380 2009-08-28 2010-08-26 频谱预测方法、装置和系统 WO2011023118A1 (zh)

Priority Applications (3)

Application Number Priority Date Filing Date Title
BR112012004383-8A BR112012004383B1 (pt) 2009-08-28 2010-08-26 método para predição de espectro, aparelho para extração de modo de ocupação de recurso de espectro e sistema para predição de espectro
EP10811278.0A EP2472933B1 (en) 2009-08-28 2010-08-26 Spectrum prediction method, apparatus and system
US13/407,220 US8687516B2 (en) 2009-08-28 2012-02-28 Method, apparatus and system for spectrum prediction

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN200910171727.5A CN102006124B (zh) 2009-08-28 2009-08-28 频谱预测方法、装置和系统
CN200910171727.5 2009-08-28

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US13/407,220 Continuation US8687516B2 (en) 2009-08-28 2012-02-28 Method, apparatus and system for spectrum prediction

Publications (1)

Publication Number Publication Date
WO2011023118A1 true WO2011023118A1 (zh) 2011-03-03

Family

ID=43627273

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2010/076380 WO2011023118A1 (zh) 2009-08-28 2010-08-26 频谱预测方法、装置和系统

Country Status (5)

Country Link
US (1) US8687516B2 (zh)
EP (1) EP2472933B1 (zh)
CN (1) CN102006124B (zh)
BR (1) BR112012004383B1 (zh)
WO (1) WO2011023118A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102595570A (zh) * 2012-01-11 2012-07-18 北京邮电大学 认知无线电系统中基于隐马尔可夫模型的频谱接入方法
US8687516B2 (en) 2009-08-28 2014-04-01 Huawei Technologies Co., Ltd. Method, apparatus and system for spectrum prediction

Families Citing this family (18)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9215704B2 (en) * 2011-01-26 2015-12-15 Lg Electronics Inc. Channel state information transmitting method and user equipment, channel state information receiving method and base station
CN102833760B (zh) * 2011-06-15 2015-08-12 电信科学技术研究院 异系统间频谱共享情况下的干扰抑制方法和设备
CN102857938A (zh) * 2011-06-30 2013-01-02 冯奇 认知无线电中频谱环境学习与预测
CN102625319B (zh) * 2012-04-06 2015-04-22 电信科学技术研究院 无线认知传感网的实现方法和设备
PT2849522E (pt) * 2013-09-16 2016-03-15 Kapsch Trafficcom Ag Método para enviar mensagens em redes ad-hoc
US9819459B2 (en) 2014-12-22 2017-11-14 Microsoft Technology Licensing, Llc Channel sensing enhancement
EP3481124B1 (en) 2016-08-11 2022-10-05 Huawei Technologies Co., Ltd. Method and device for determining resource energy
CN106788823B (zh) * 2017-03-16 2018-11-09 石春 一种无线信道状态信息测量方法
CN107396370A (zh) * 2017-06-30 2017-11-24 杭州清本科技有限公司 频谱规划装置
CN107343281A (zh) * 2017-06-30 2017-11-10 杭州清本科技有限公司 频谱规划方法及计算机存储介质
JP7166584B2 (ja) 2017-12-04 2022-11-08 株式会社国際電気通信基礎技術研究所 チャネル状態予測装置、無線通信装置、チャネル状態予測方法、および、プログラム
CN112751792B (zh) * 2019-10-31 2022-06-10 华为技术有限公司 一种信道估计方法及装置
CN112688746B (zh) * 2020-12-14 2021-11-30 中山大学 一种基于时空数据的频谱预测方法
CN112860768B (zh) * 2021-02-17 2022-02-25 中国电子科技集团公司第二十二研究所 一种电磁频谱可用频率推荐方法
CN113179144A (zh) * 2021-04-16 2021-07-27 中国科学院微电子研究所 频谱预测方法和装置
CN113347638B (zh) * 2021-05-14 2024-04-30 东北大学 一种认知工业物联网动态协作频谱预测与感知方法
CN114050856B (zh) * 2021-09-27 2024-02-20 南京邮电大学 一种卫星频谱占用状态拟合与预测方法
CN113852970B (zh) * 2021-09-30 2022-07-26 中山大学 基于图神经网络的多维频谱预测方法、系统、装置及介质

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20070014746A (ko) * 2005-07-29 2007-02-01 한국과학기술원 무선 이동 통신 시스템에서의 귀환 신호를 이용한 하향무선 채널 환경 예측 장치 및 방법
US20080005219A1 (en) * 2006-01-10 2008-01-03 Marvell International Ltd. Transmission scheduling for receiver feedback
CN101277278A (zh) * 2008-05-21 2008-10-01 清华大学 一种针对有轨交通的无线信道预测方法
WO2009030035A1 (en) * 2007-09-07 2009-03-12 Wi-Lan Inc. Quantized channel state information prediction in multiple antenna systems

Family Cites Families (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7366471B1 (en) * 2000-08-31 2008-04-29 Intel Corporation Mitigating interference between wireless systems
US7139285B2 (en) * 2001-12-10 2006-11-21 Telcordia Technologies, Inc. Frequency hop collision prediction in a multi-channel, bluetooth-enabled packet transmission system
EP1424865A1 (en) * 2002-11-29 2004-06-02 Siemens Mobile Communications S.p.A. Method of and device for resource allocation in digital cellular communication systems
CN1889752A (zh) * 2006-06-26 2007-01-03 西安交通大学 认知无线电系统中授权用户检测的频率分配方法
US20080069079A1 (en) * 2006-09-19 2008-03-20 Star-H Corporation System and method for spectrum sharing
KR100824602B1 (ko) * 2007-08-23 2008-04-24 한국전자통신연구원 인식 무선 기반의 채널 상태 예측장치 및 그 방법
CN101262701A (zh) * 2008-04-22 2008-09-10 华东师范大学 一种基于遗传算法的动态信道分配方法
CN101442807B (zh) * 2008-12-30 2012-09-05 北京邮电大学 通信系统中资源分配的方法及系统
CN102006124B (zh) 2009-08-28 2014-05-07 华为技术有限公司 频谱预测方法、装置和系统

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20070014746A (ko) * 2005-07-29 2007-02-01 한국과학기술원 무선 이동 통신 시스템에서의 귀환 신호를 이용한 하향무선 채널 환경 예측 장치 및 방법
US20080005219A1 (en) * 2006-01-10 2008-01-03 Marvell International Ltd. Transmission scheduling for receiver feedback
WO2009030035A1 (en) * 2007-09-07 2009-03-12 Wi-Lan Inc. Quantized channel state information prediction in multiple antenna systems
CN101277278A (zh) * 2008-05-21 2008-10-01 清华大学 一种针对有轨交通的无线信道预测方法

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8687516B2 (en) 2009-08-28 2014-04-01 Huawei Technologies Co., Ltd. Method, apparatus and system for spectrum prediction
CN102595570A (zh) * 2012-01-11 2012-07-18 北京邮电大学 认知无线电系统中基于隐马尔可夫模型的频谱接入方法
CN102595570B (zh) * 2012-01-11 2014-09-24 北京邮电大学 认知无线电系统中基于隐马尔可夫模型的频谱接入方法

Also Published As

Publication number Publication date
CN102006124B (zh) 2014-05-07
EP2472933A1 (en) 2012-07-04
BR112012004383B1 (pt) 2019-10-29
US20120155318A1 (en) 2012-06-21
CN102006124A (zh) 2011-04-06
EP2472933B1 (en) 2014-10-01
EP2472933A4 (en) 2012-07-04
US8687516B2 (en) 2014-04-01
BR112012004383A2 (pt) 2016-03-22

Similar Documents

Publication Publication Date Title
WO2011023118A1 (zh) 频谱预测方法、装置和系统
CN104012158B (zh) 灵活带宽系统中的动态带宽调整
US9877270B2 (en) Methods and communication devices for performing cell search and detection
US8467787B2 (en) Method and apparatus for searching for modes and frequencies
US11811576B2 (en) Methods and devices for narrowband communications
Liu et al. Information theoretic criterion-based spectrum sensing for cognitive radio
Askari et al. A channel assignment algorithm for cognitive radio wireless sensor networks
Shukla et al. A survey on energy detection schemes in cognitive radios
JP2021535686A (ja) 干渉源識別方法、関連デバイス、およびコンピュータ記憶媒体
US20170054538A1 (en) Mobile terminal devices and methods of detecting reference signals
CN105813089B (zh) 一种对抗噪声不确定性的匹配滤波频谱感知方法
CN108064053B (zh) 网络性能分析方法及装置
Syed et al. On the usage of history for energy efficient spectrum sensing
Li et al. MTF 2 N: Multi-Channel Temporal-Frequency Fusion Network for Spectrum Prediction
CN110996354B (zh) 移动通信中具有干扰避免的小区重选方法及其装置
US9313011B2 (en) Method and apparatus for scheduling terminals in system with multiple antenna elements combining into common cell
Zhang et al. Spectrum opportunity exploration based on joint spectrum sensing and the spectrum prediction enabled spectrum database
Liu et al. Optimal cooperative sensing scheme under time-varying channel for cognitive radio networks
Arif et al. Sensing time minimization using pipelining in two stage spectrum sensing
CN101415239A (zh) 一种终端接入方法和装置
Pan et al. Efficient spectrum handoff scheme in cognitive radio
JP5747577B2 (ja) コグニティブ無線通信システム
KR20190119364A (ko) 에너지 효율 와이파이 인원 계수 시스템 및 방법
Takeuchi et al. Radio environment prediction for cognitive radio
CN112969143B (zh) 一种终端室内外区分方法、系统、服务器和存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 10811278

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

WWE Wipo information: entry into national phase

Ref document number: 513/KOLNP/2012

Country of ref document: IN

WWE Wipo information: entry into national phase

Ref document number: 2010811278

Country of ref document: EP

REG Reference to national code

Ref country code: BR

Ref legal event code: B01A

Ref document number: 112012004383

Country of ref document: BR

ENP Entry into the national phase

Ref document number: 112012004383

Country of ref document: BR

Kind code of ref document: A2

Effective date: 20120228