JP2009213093A - Radio controller and radio communication apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To predict a radio environment during a communication by a radio module in which a radio communication system is mounted, and to determine a radio communication system that a radio communication apparatus uses, on the basis of the prediction. <P>SOLUTION: A radio controller includes: a prediction expression selecting means for acquiring radio environment information representing propriety of the radio environment for each of a plurality of radio modules and for selecting a prediction expression suitable for the radio environment on the basis of the obtained radio environment information from among a plurality of prediction expressions for predicting future radio environment information in the respective radio modules; and a radio module selecting means which uses the prediction expression selected by the prediction expression selecting means to obtain predictive values of future radio environment information in the respective radio modules, and which selects a radio module to be used for radio communications on the basis of these predictive values. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a wireless control device that determines a wireless communication method according to the state of a wireless environment, and a wireless communication device that performs wireless communication using the determined wireless communication method.

  Conventionally, a cognitive radio communication technique has been proposed as a technique for realizing efficient frequency utilization. The cognitive radio communication technology is a technology for dynamically selecting a radio communication method and a radio channel according to a surrounding radio environment and performing radio communication using this. In realizing the cognitive radio communication technology, there is a problem of how to select a radio module or a radio channel in which each radio communication method is implemented.

With respect to this problem, a technique for selecting a wireless module using fuzzy control has been proposed (see Non-Patent Document 1). In the technique disclosed in Non-Patent Document 1, a membership function is defined for each wireless system for each wireless environment information (reception power, wireless channel occupation time, packet loss rate, etc.) used for selecting a wireless module. The And a cognitive radio communication system absorbs the feature between different radio systems by using each membership function, and selects an optimal radio module.
Further, a specific wireless environment recognition technique using reception power as wireless environment information has been proposed (see Non-Patent Document 2).

  In addition, although not limited to a cognitive radio communication technique, a technique related to a radio channel allocation method in a radio network has also been proposed (see Patent Document 1). This patent document 1 discloses a technique for determining the number of assigned radio channels in accordance with network traffic fluctuations and the transmission quality of each radio link when a mesh network is constructed by radio communication. According to this technology, in a network constructed in a mesh shape, the number of radio channels allocated to the detour path is increased in response to the problem of traffic concentration on the detour path when a certain link becomes unavailable. It becomes possible to respond.

  In addition, by estimating the link reliability between nodes using a prediction mechanism and a variable weighting filter, it is possible to determine an optimum route between a source node and a destination node in a communication network such as wireless ad hoc communication. Technology has also been proposed (see Patent Document 2). The prediction mechanism disclosed in Patent Document 2 is a technology that performs prediction using information in the physical layer such as signal strength and signal-to-noise ratio as link quality. The weighting filter also corrects the value predicted based on the traffic situation experienced by the node whose link quality is estimated. The traffic situation is information indicating, for example, whether the traffic is in an idle state, light traffic, medium traffic, or heavy traffic.

JP 2002-353974 A JP-T-2006-527525 R. Agusti, O. Sallent, J. Perez-Romero, L. Giupponi, “A Fuzzy-Neural Based Approach for Joint Radio Resource Management in a Beyond 3G Framework”, QSHINE’04, 2004 N.Hoven, A. Sahai, “Power scaling for cognitive radio”, Proc. International conference on Wireless Networks, Communications and Mobile Computing, Hawaii, USA, June, 2005, pp. 250-255

However, the wireless environment changes from time to time. Furthermore, the manner of change is not always constant. For this reason, the current wireless environment does not always accurately represent the subsequent wireless environment, and the selected wireless module may not be optimal at the time of transmission.
The present invention has been made to solve the above-described problem, and an object of the present invention is to provide an apparatus capable of determining a wireless module used by a wireless communication apparatus in accordance with a wireless environment that changes every moment. To do.

A first aspect of the present invention is a wireless control device, which includes a prediction formula selection unit and a wireless module selection unit. The prediction formula selection means acquires radio environment information representing the quality of the radio environment for each of the plurality of radio modules, and is acquired from a plurality of prediction formulas for predicting future radio environment information in each radio module. Based on the wireless environment information, a prediction formula suitable for the wireless environment is selected. The wireless module selection means acquires a predicted value of future wireless environment information in each wireless module using the prediction formula selected by the prediction formula selection means, and based on the predicted value, selects a wireless module used for wireless communication. select.
According to the radio control device configured as described above, a prediction formula suitable for the radio environment is selected from a plurality of prediction formulas based on the radio environment information. And a predicted value is calculated using this prediction formula. Therefore, it is possible to select a prediction formula corresponding to each wireless environment in which various situations are assumed, and it is possible to obtain a predicted value with a small error. Therefore, it becomes possible to select a wireless module according to such a change in the wireless environment.

In the wireless control device according to the first aspect of the present invention, the prediction formula selection unit uses the past actual measurement value of the wireless environment information in each wireless module and the past prediction value in each prediction formula to predict with little prediction error. It may be configured to select an expression.
In the wireless control device of the present invention, the prediction formula selection means selects a prediction formula based on a model that matches the time series information based on the time series information of the past actual measurement values of the radio environment information in each radio module. It may be configured.

In the wireless control device according to the first aspect of the present invention, the prediction formula selection means is configured to update the value of the coefficient used for the currently selected prediction formula more frequently than the frequency of selecting the prediction formula. May be.
In general, the process of selecting a prediction formula takes a lot of time compared to the process of updating the coefficient values used in the currently selected prediction formula. Even if the prediction formula does not change, the accuracy of the prediction is improved by updating the value of the coefficient used in the prediction formula. Therefore, with this configuration, it is possible to improve the accuracy of prediction while suppressing the time required for processing.

In the radio control apparatus according to the first aspect of the present invention, the prediction formula selection means determines whether or not the currently selected prediction formula is valid, and selects the prediction formula only when it is determined that the prediction formula is not valid. It may be configured.
In general, the process of selecting a prediction formula requires a longer time than the process of determining whether or not the prediction formula is valid. If the prediction formula is valid, the accuracy of the prediction can be maintained high without changing the prediction formula. Therefore, with this configuration, it is possible to maintain high prediction accuracy while suppressing the time required for processing.

  A second aspect of the present invention is a wireless communication device that performs wireless communication with another wireless communication device, and includes a wireless control device having any one of the above-described configurations, a plurality of wireless modules, and a wireless module selection unit. Packet switch means for passing the packet to the selected wireless module.

  The present invention may be specified as a computer program for causing a computer to operate as the above-described wireless control device. Further, the present invention may be specified as a wireless control method performed by a computer that functions as the above-described wireless control device.

  According to the present invention, it is possible to determine a wireless module used by a wireless communication apparatus according to a wireless environment that changes from moment to moment.

The wireless communication device 100 according to the present invention performs wireless communication with other wireless communication devices. The other wireless communication device may be a mobile communication terminal device such as a mobile phone, PHS, or a personal computer having a wireless communication function, or may be a relay device such as a wireless communication base station or a wireless router. Further, the other wireless communication device may be a wireless communication device according to the present invention.
The wireless communication device 100 according to the present invention includes a plurality of wireless modules each having a different wireless communication method. Each wireless module performs wireless communication based on the wireless communication system implemented in each. Examples of the wireless communication system include “IEEE802.11j”, “IEEE802.11b”, and “IEEE802.11g” which are standards for wireless LAN systems. Examples of the wireless communication method include a wireless access method called “Mobile WiMAX” and a CDMA method called “cdma2000 lxEV-DO” compliant with “IEEE802.16e” which is a standard of a mobile communication system. . Note that the “different wireless communication systems” described above include the same wireless communication system and different channels. That is, for example, even if they are common in terms of “IEEE802.11j”, if the channels are different, the respective wireless communication systems are treated as different.
The wireless control unit 40 selects an optimal wireless communication method according to the wireless environment from the plurality of wireless communication methods, and performs wireless communication using a wireless module in which the selected wireless communication method is mounted. Do. For example, the selection of the wireless communication method and the execution of wireless communication using the selected wireless module can be implemented based on, for example, cognitive wireless technology.

  Cognitive wireless technology recognizes the surrounding wireless environment, selects the optimal wireless communication method from a plurality of wireless communication methods according to the recognized wireless environment, and uses the selected wireless communication method. This is a technology for performing wireless communication. The cognitive radio technology can also select a more optimal radio channel in accordance with the recognized radio environment in the selected radio communication method, and perform radio communication using the selected radio channel. Hereinafter, the wireless communication device 100 according to the present invention will be described.

FIG. 1 is a diagram illustrating functional blocks of the wireless communication apparatus 100. The wireless communication device 100 includes a plurality of wireless modules 10-1, 10-2,..., 10-n (hereinafter, items common to each wireless module are described as “wireless module 10”), a packet switch unit. 20, a wireless environment recognition unit 30, and a wireless control unit 40 (corresponding to a “wireless control device” in the present invention).
The wireless communication device 100 can be configured using an information processing device that is connected to a computing device or a storage device via a bus and operates based on a program stored in the storage device. Each functional unit or part of the wireless communication device 100 may be configured using dedicated hardware. Hereinafter, each functional unit included in the wireless communication device 100 will be described with reference to FIG.

The wireless module 10 performs processing of the PHY layer and the data link layer in wireless communication performed between the wireless communication device 100 and another wireless communication device. The wireless module 10 is configured by applying the existing wireless communication method as described above. The wireless module 10 passes a packet received from another wireless communication device to the packet switch unit 20. In addition, when the wireless module 10 receives a packet from the packet switch unit 20, the wireless module 10 transmits the packet by using a wireless communication method implemented in itself.
The radio module 10 measures radio environment information indicating whether the radio environment is good or bad for each radio channel. Examples of the radio environment information include received power, interference power, the number of packet retransmissions, the number of CRC (Cyclic Redundancy Check) error packets, and the radio channel occupation time. When the radio channel occupation time is used as the radio environment information, the radio module 10 may be configured to use a value obtained by subtracting the time used for communication of the local station from the radio channel occupation time. The wireless module 10 passes the wireless environment information acquired in this way to the wireless environment recognition unit 30.

  The packet switch unit 20 is used when transmitting (when the wireless communication device 100 transmits data to another wireless communication device) and when receiving (when the wireless communication device 100 receives data from another wireless communication device). Behave differently. At the time of transmission, the packet switch unit 20 performs distribution to the radio module 10 instructed for each packet in accordance with an instruction from the radio control unit 40. At the time of reception, the packet switch unit 20 controls the order of packets received by each wireless module 10 and passes the packet to a higher-layer control unit (not shown) that processes the packet.

  The wireless environment recognition unit 30 receives wireless environment information for each wireless channel from the wireless module 10. The wireless environment recognizing unit 30 totals the received wireless environment information, and passes the collection results to the wireless control unit 40 at a constant cycle.

The wireless control unit 40 receives wireless environment information from the wireless environment recognition unit 30. In addition, the wireless control unit 40 receives information related to the destination of the packet to be transmitted from an upper layer control unit (for example, a network control unit that performs routing).
Based on the wireless environment information, the wireless control unit 40 determines which wireless module 10 is to be used from among the wireless modules 10 that are connected to the destination wireless communication partner for each packet. To do. Further, the wireless control unit 40 may be configured to determine a wireless channel to be used in the wireless module 10 in which a connection is established with the wireless communication partner to which the packet is destined based on the wireless environment information. good. The wireless communication partner that is the destination of each packet can be identified from the destination address of the packet.

  The wireless control unit 40 includes a prediction formula selection unit 41 and a wireless module selection unit 42 in order to realize the above-described processing. The prediction formula selection unit 41 has a plurality of prediction formulas for predicting future radio environment information (at the time of packet transmission), and selects a prediction formula used for prediction based on the radio environment information received from the radio environment recognition unit 30. . The wireless module selection unit 42 determines the wireless module 10 to be used for each packet (or for each destination of the packet) using the prediction formula selected by the prediction formula selection unit 41 and instructs the packet switch unit 20. The time series information storage unit 43 stores the wireless environment information at each time point notified from the wireless environment recognition unit 30 as time series information. Hereinafter, specific operations of the prediction formula selection unit 41 and the wireless module selection unit 42 will be described.

[Prediction formula]
As described above, the prediction formula selection unit 41 has a plurality of prediction formulas for predicting future radio environment information (during packet transmission). The prediction formula is a formula for predicting a future radio environment based on current or past radio environment information. Specific examples of prediction formulas include formulas based on AR models (Auto-Regressive models), MA models (Moving Average models), ARMA models (Auto-Regressive Moving Average models), exponential smoothing methods, Brownian methods, and the like. is there. The prediction formula selection unit 41 selects one prediction formula from such a plurality of prediction formulas. At this time, the prediction formula selection unit 41 may perform selection that the prediction formula is not used. That is, when it is determined that the accuracy of any prediction formula is low, the prediction formula selection unit 41 selects the wireless module 10 to be used based on the current wireless environment information without using the prediction formula. decide. Note that the type and number of prediction formulas installed in the prediction formula selection unit 41 may be appropriately determined by the designer. The prediction formula selection unit 41 also calculates a coefficient used for each prediction formula.

As a specific example of the prediction formula, a prediction formula based on the AR model will be described. Equation 1 represents a prediction equation based on the AR model. In Equation 1, the value on the left side represents the predicted value of the radio environment information at time t + 1 (during packet transmission). X _t represents the actual value of the radio environment information at time t. a ₀ , a ₁ ... a _p represent coefficients of the regression equation, and p represents the order. The value of p is determined in advance as a parameter. Each coefficient of the regression equation can be calculated using past data. For example, each coefficient can be calculated by using the least square method. This calculation is performed by the prediction formula selection unit 51 at a predetermined timing. This timing will be described later. The number of past data used at this time is referred to as “learning period” in the following description. The larger the value of p, the larger the learning period value needs to be set. The past data can be acquired by reading from the time-series information storage unit 43.

Note that when the prediction formula is not used as described above, the value of the wireless environment information at time t + 1 (at the time of packet transmission) is expressed by Formula 2. That is, when the prediction formula selection unit 41 determines not to use the prediction formula, the wireless module selection unit 42 performs selection of the wireless module based on the current wireless environment information without using the predicted value.

[Selection method of prediction formula]
Next, a method in which the prediction formula selection unit 41 selects a prediction formula will be described. The prediction formula selection unit 41 selects a prediction formula with a small prediction error based on the prediction error in each prediction formula. In other words, the prediction formula selection unit 41 acquires the reliability in each prediction formula and selects a prediction formula with high reliability. The degree of reliability is a value representing the goodness of time series application in the wireless environment information up to the present time for each prediction formula, and is a predicted value obtained in the past (hereinafter referred to as “past predicted value”) and its prediction. It is obtained by using an actual value corresponding to the value (hereinafter referred to as “past actual measurement value”). Note that the number of past data used when acquiring the reliability may be determined as appropriate by the designer.

As the reliability, the prediction formula selection unit 41, based on the past prediction value and the past actual measurement value, statistical information (distribution, standard deviation, kurtosis, etc.) of the distribution of the prediction error, determination count, AIC (Akaike's Information Criterion) : Akaike information criterion) can be used.
The determination coefficient is a value represented by Equation 3, and the greater the value, the higher the reliability.

AIC is a value represented by Equation 4, and the smaller the value, the higher the reliability.

  The prediction formula selection unit 41 may select a prediction formula by using an autocorrelation function and a partial autocorrelation function. FIG. 2 is a graph showing an autocorrelation function (FIG. 2A) and a partial autocorrelation function (FIG. 2B) when the time series is an AR model. When the time series in the current wireless environment information is an AR model, the autocorrelation function gradually decreases as the value of lag increases, and the partial autocorrelation function decreases rapidly and is cut off.

  Based on such characteristics, the prediction formula selection unit 41 can determine whether or not to use the prediction formula of the AR model based only on the past actual measurement value without using the past prediction value. The prediction formula selection unit 41 determines that the prediction formula of the AR model should be used when the autocorrelation function and the partial autocorrelation function of the past actual measurement values indicate the characteristics of the AR model described above. On the other hand, the prediction formula selection unit 41 selects another prediction formula when the past actual measurement value does not indicate the feature of the AR model. This determination is based on the values of a plurality of sampling points of the autocorrelation function and the partial autocorrelation function obtained from the past actual measurement values, and the autocorrelation function and the partial autocorrelation function obtained from the data series that are previously stored as the AR model values. And the corresponding sampling point value can be compared and determined according to the distance or the like. It can also be determined by pattern matching between the graph pattern of the auto-correlation function and partial auto-correlation function of the AR model that is held in advance and the graph pattern of the actual measurement value. In this case, the prediction formula selection unit 41 may have a plurality of patterns used for pattern matching. The prediction formula selection unit 41 can improve the accuracy of determination based on the autocorrelation function and the partial autocorrelation function by having more patterns.

  Further, the prediction formula selection unit 41 may acquire the slope of the autocorrelation function and determine whether or not the feature of the AR model is generated based on the position where the slope and the partial autocorrelation function are cut off. For example, when the slope in the autocorrelation function is gentler than the slope threshold value set in advance as a parameter, the prediction formula selection unit 41 can determine that a feature of the AR model has occurred. Specifically, the determination can be made based on whether the ratio of the slope values exceeds a threshold value. For the partial autocorrelation function, the prediction formula selection unit 41 can also regard the partial autocorrelation function as disconnected when the vibration falls within a predetermined confidence interval (for example, a plus or minus 5% interval). In this case, first, the prediction formula selection unit 41 obtains a partial autocorrelation function up to a predetermined horizontal axis value (lag value). When the vibration (vibration of the value on the vertical axis) falls within a predetermined confidence interval after a certain lag value, the prediction formula selection unit 41 determines that this partial autocorrelation function has been cut at the lag value. . And the prediction formula selection part 41 judges that the characteristic of the partial autocorrelation function in the case of AR model has arisen. Further, the prediction formula selection unit 41 may not use the value of the degree in advance, and may use the value of the lag in which the cut has occurred as the value of the degree (p). These determinations may be realized by using other known methods.

Also, in these determination processes, when values corresponding to past measured values and similarities of the autocorrelation function and partial autocorrelation function of the AR model are calculated, the prediction formula selection unit 41 uses the prediction formula of the AR model. Instead of determining whether or not to use this value, this value may be acquired as the reliability, and a prediction formula may be selected based on the reliability by comparison with the reliability of other prediction formulas. .
Further, the above-described method specialized for the AR model may be applied in the selection of the MA model or the prediction formula of the ARMA model. In the case of the MA model, features of the autocorrelation function in the AR model appear in the partial autocorrelation function. In the case of the MA model, the characteristic of the partial autocorrelation function in the AR model appears in the autocorrelation function. In the case of the ARMA model, the feature of the autocorrelation function in the AR model appears in both the autocorrelation function and the partial autocorrelation function. Based on such characteristics, the prediction formula selection unit 41 can perform a selection method specialized for the MA model or the ARMA model. Further, by applying each of the features and methods described above, the prediction formula selection unit 41 obtains an autocorrelation function and a partial autocorrelation function for the time series in the current wireless environment information, and is the most among the AR model, MA model, and ARMA model. A prediction formula of a matching model may be used.

[Execution timing of prediction formula selection processing]
Next, the execution timing of processing in which the prediction formula selection unit 41 selects a prediction formula will be described. 3A to 3F show radio environment information acquisition, prediction formula selection processing, coefficient update processing (processing for updating the coefficient of each prediction formula), validity determination processing (validity of the currently selected prediction formula). It is a figure which shows the variation of the execution timing of the process which judges sex. The prediction formula selection unit 41 may be configured by adopting any variation shown in FIG.

  The execution timing of FIG. 3A will be described. In this case, the prediction formula selection unit 41 executes a prediction formula selection process every time wireless environment information is notified from the wireless environment recognition unit 30. At this time, the prediction formula selection unit 41 also updates the coefficient used in each prediction formula.

  The execution timing of FIG. 3B will be described. In this case, the prediction formula selection unit 41 executes the prediction formula selection process every time a predetermined period elapses. At this time, the prediction formula selection unit 41 also updates the coefficient used in each prediction formula. Note that the predetermined period is set longer than the period in which the wireless environment information is notified from the wireless environment recognition unit 30.

  The execution timing of FIG.3 (c) is demonstrated. In this case, the prediction formula selection unit 41 updates each coefficient every time a predetermined period elapses, and executes a prediction formula selection process every time a predetermined period longer than this period elapses. In addition, the prediction formula selection part 41 updates a coefficient only about the prediction formula currently utilized, when performing only a coefficient update process, without performing a prediction formula selection process. The same applies to the following description.

The execution timing of FIG. 3 (d) will be described. In this case, the prediction formula selection unit 41 determines the validity of the currently used prediction formula, and executes the prediction formula selection process when it is determined that the prediction formula lacks validity. The prediction formula selection unit 41 may make a determination on the validity of the prediction formula based on any criterion. For example, the prediction formula selection unit 41 acquires the difference between the actual measurement value and the prediction value, determines that the validity is insufficient when the difference exceeds a predetermined threshold value, and is valid if the difference is within the threshold value. It may be judged that the sex is satisfied. Such a threshold may be freely determined by the designer. For example, statistical information of prediction errors may be collected for each prediction formula, and a value based on the standard deviation of the prediction error (for example, a value that is three times the standard deviation) may be set in advance as a threshold value. Note that the prediction formula selection unit 41 also updates a coefficient used in each prediction formula when performing the prediction formula selection process.
In addition, a threshold for prediction formula selection processing and a threshold for coefficient update processing may be set separately. That is, when the above-described difference value exceeds a threshold value set for each, processing corresponding to the threshold value may be executed.
Further, as shown in FIG. 3C, the coefficient update process may be periodically executed regardless of the validity determination process.

The execution timing of FIG.3 (e) is demonstrated. In this case, the prediction formula selection unit 41 determines whether or not the difference between the actually measured value and the predicted value exceeds the threshold, as in FIG. Then, when the number of times that the difference exceeds the threshold exceeds a predetermined number (three in FIG. 3E), the prediction formula selection process is executed. The number of times that the difference exceeds the threshold may be the total number or the continuous number (the total number in FIG. 3E).
Note that the prediction formula selection unit 41 also updates a coefficient used in each prediction formula when performing the prediction formula selection process. In this case, the prediction formula selection unit 41 may update the coefficient every time the difference exceeds the threshold (corresponding to FIG. 3F).
Note that the execution timing of the above-described prediction formula selection processing or the like is merely an example, and the prediction formula selection unit 41 may be designed by adopting another form of timing.
The prediction formula selection unit 41 selects the prediction formula with the highest reliability based on the reliability acquired in the execution of the prediction formula selection process. Then, the prediction formula selection unit 41 notifies the wireless module selection unit 42 of the selected prediction formula together with the coefficient. When the radio module selection unit 42 selects a radio module based on a plurality of radio environment information, the prediction formula selection unit 41 selects a prediction formula for each radio environment information used by the radio module selection unit 42. And the coefficient is acquired.

  The wireless module selection unit 42 calculates a predicted value for the wireless environment information for each wireless module, using the prediction formula selected by the prediction formula selection unit 41 and the calculated coefficient. The wireless module selection unit 42 selects the wireless module 10 with the best predicted value based on the calculated predicted value. The process for realizing the selection of the wireless module with the best predicted value can be applied by replacing the selection process in the conventional cognitive radio technology with the predicted value. Then, the wireless module selection unit 42 notifies the packet switch unit 20 that the selected wireless module 10 is used.

  FIG. 4 is a flowchart illustrating an operation example of the wireless control unit 40 of the wireless communication apparatus 100. An operation example of the wireless control unit 40 will be described with reference to FIG. The prediction formula selection unit 41 acquires wireless environment information from the wireless environment recognition unit 30 (S01). The prediction formula selection unit 41 determines whether it is time to update the coefficient of the prediction formula (S02). This timing differs depending on which execution timing is adopted as described above. If it is not time to update the coefficient of the prediction formula (S02-NO), the prediction formula selection unit 41 returns to the processing of S01.

  On the other hand, when it is time to update the coefficient of the prediction formula (S02-YES), the prediction formula selection unit 41 determines whether it is the timing of the prediction formula selection process (S03). This timing also differs depending on which execution timing is adopted as described above.

  When it is not the timing of the prediction formula selection process (S03-NO), the prediction formula selection unit 41 is currently using the prediction formula currently used in the wireless module selection unit 42 (that is, the prediction currently selected by the prediction formula selection unit 41). The coefficient of the formula is updated (S04), and the process returns to S01. On the other hand, when it is the timing of a prediction formula selection process (S03-YES), the prediction formula selection part 41 updates the coefficient in each prediction formula (S05). Next, the prediction formula selection unit 41 acquires the reliability of each prediction formula (S06). Then, the prediction formula selection unit 41 selects the prediction formula with the highest reliability based on the acquired reliability (S07), and notifies the wireless module selection unit 42 together with the coefficient of the prediction formula. Note that in the processes of S06 and S07, the prediction formula selection unit 41 may execute selection of a prediction formula based on the AR model, the MA model, and the characteristics of the ARMA model as described above. After this process, the prediction formula selection unit 41 performs the process of S01 again.

FIG. 5 is a flowchart illustrating an operation example of the wireless control unit 40 of the wireless communication apparatus 100. The wireless control unit 40 executes the process shown in the flowchart of FIG. 4 and the process shown in the flowchart of FIG. 5 at independent cycles. Hereinafter, the process shown in the flowchart of FIG. 5 will be described.
The wireless module selection unit 42 calculates a predicted value for the wireless environment information using the prediction formula and coefficient selected by the prediction formula selection unit 41 (S08). Then, the wireless module selection unit 42 selects a wireless module based on the calculated predicted value (S09). The wireless module selection unit 42 notifies the packet switch unit 20 of the selected wireless module.

  According to the wireless communication device 100 of the present invention, the wireless module 10 used for packet transmission is optimized based on the predicted value of the future wireless environment information (at the time of packet transmission) calculated based on the wireless environment information. Is selected. Therefore, compared with the case where the optimality of each wireless module is determined based on the current or past wireless environment information, it is possible to realize more accurate selection corresponding to the changing wireless environment from moment to moment. It becomes. The accurate selection is to select a wireless module with good communication efficiency when communication efficiency is determined based on the wireless environment information considered in the wireless module selection unit 42.

  In addition, according to the wireless communication device 100 of the present invention, a plurality of calculation methods (prediction formulas) for obtaining the above-described prediction value are prepared, and the prediction formula most suitable for the time series model of the wireless environment information from that to the present ( That is, the prediction formula having the highest reliability is selected. And a predicted value is calculated using this prediction formula. Therefore, it is possible to obtain a predicted value that makes the prediction error smaller, and it is possible to realize more accurate selection of a wireless module.

In addition, according to the wireless communication device 100 of the present invention in which the execution timing of FIG. 3A is adopted, the coefficient update process and the prediction formula selection process are executed each time wireless environment information is acquired. Therefore, it is possible to acquire a predicted value with higher accuracy, and it is possible to accurately select a wireless module.
Also, according to the wireless communication device 100 of the present invention in which the execution timing of FIG. 3B is adopted, the execution frequency of the coefficient update process and the prediction formula selection process is reduced compared to the case of FIG. . Therefore, it is possible to reduce the load in the coefficient update process and the prediction formula selection process.

Further, according to the wireless communication device 100 of the present invention in which the execution timing of FIG. 3C is adopted, the execution frequency of the prediction formula selection process is reduced compared to the case of FIG. And the coefficient update process at the timing when the prediction formula selection process is not executed is executed only for the prediction formula currently used. Therefore, it is possible to reduce the time required for the coefficient update process and the prediction formula selection process and increase the processing speed of the wireless communication apparatus 100.
In addition, according to the wireless communication device 100 of the present invention in which the execution timing of FIG. 3D is adopted, it is determined whether or not the currently used prediction formula is valid, and when it is determined that it is not valid. Only the prediction formula selection process is executed. The processing time required for validity determination can be made shorter than the processing time required for the prediction formula selection process. Therefore, it is possible to reduce the time required for the coefficient update process and the prediction formula selection process and increase the processing speed of the wireless communication apparatus 100.

In addition, according to the wireless communication device 100 of the present invention in which the execution timing of FIG. 3E is adopted, the prediction formula selection process is executed only when it is determined that the validity determination process is not valid a plurality of times. . Therefore, when it is determined that it is not appropriate to happen, it is possible to prevent the prediction formula selection process that requires a long processing time from being performed wastefully, and to increase the processing speed of the wireless communication apparatus 100.
In addition, according to the wireless communication device 100 of the present invention in which the execution timing of FIG. 3 (f) is adopted, the effects in FIG. 3 (c) and FIG. 3 (e) are exhibited together.

<Modification>
In addition, the wireless communication device 100 is configured to further include a wireless module for wireless environment recognition specialized in providing information on the wireless environment to the wireless environment recognition unit 30 separately from the wireless module 10 for communication. May be.
Further, in the processing of the prediction formula selection unit 41, the selection priority order not to use the prediction formula may be configured to be the lowest.
Further, the wireless control unit 40 may receive wireless environment information directly from each wireless module 10 without using the wireless environment recognition unit 30. In this case, the wireless communication device 100 may not include the wireless environment recognition unit 30.

  In addition, you may make it implement | achieve the function of a part of wireless communication apparatus 100 in embodiment mentioned above, for example, the prediction type selection part 41, and the wireless module selection part 42 with a computer. In that case, a program for realizing the information management function may be recorded on a computer-readable recording medium, and the program recorded on the recording medium may be read into a computer system and executed. Here, the “computer system” includes an OS and hardware such as peripheral devices. The “computer-readable recording medium” refers to a storage device such as a flexible medium, a magneto-optical disk, a portable medium such as a ROM and a CD-ROM, and a hard disk incorporated in a computer system. Further, the “computer-readable recording medium” dynamically holds a program for a short time, like a communication line when transmitting a program via a network such as the Internet or a communication line such as a telephone line. It is also possible to include those that hold a program for a certain time, such as a volatile memory inside a computer system serving as a server or client in that case. The program may be a program for realizing a part of the functions described above, and may be a program capable of realizing the functions described above in combination with a program already recorded in a computer system.

  The embodiment of the present invention has been described in detail with reference to the drawings. However, the specific configuration is not limited to this embodiment, and includes designs and the like that do not depart from the gist of the present invention.

  The present invention can be used in a scene where a plurality of wireless modules are provided and communication is performed using any one of them. In particular, it has a remarkable effect in a scene where wireless communication is performed using a cognitive radio technology.

It is a figure which shows the functional block of the radio | wireless control apparatus by this invention. It is a graph showing the autocorrelation function and partial autocorrelation function of AR model. It is a figure which shows the variation of the execution timing of radio | wireless environment information acquisition, a prediction type selection process, a coefficient update process, and a validity judgment process. It is a flowchart which shows the operation example of a radio | wireless control part. It is a flowchart which shows the operation example of a radio | wireless control part.

Explanation of symbols

10 (10-1, 10-2,..., 10-n) ... wireless module, 20 ... packet switch unit, 30 ... wireless environment recognition unit, 40 ... wireless control unit, 41 ... prediction formula selection unit, 42 ... wireless Module selection unit, 43 ... time-series information storage unit, 100 ... wireless communication device

Claims (6)

Based on the acquired wireless environment information from among multiple prediction formulas for acquiring wireless environment information representing the quality of the wireless environment for each of the plurality of wireless modules and predicting future wireless environment information in each wireless module A prediction formula selection means for selecting a prediction formula suitable for the wireless environment;
Using the prediction formula selected by the prediction formula selection means, a predicted value of future wireless environment information in each wireless module is obtained, and a wireless module selection for selecting a wireless module to be used for wireless communication based on the predicted value A wireless control device.   2. The radio according to claim 1, wherein the prediction formula selection unit selects a prediction formula with a small prediction error using a past actual measurement value of wireless environment information in each radio module and a past prediction value in each prediction formula. Control device.   The radio control apparatus according to claim 1, wherein the prediction formula selection unit selects a prediction formula based on a model that matches the time series information based on time series information of past actual measurement values of radio environment information in each radio module. .   The radio control apparatus according to any one of claims 1 to 3, wherein the prediction formula selection unit updates a value of a coefficient used for a currently selected prediction formula more frequently than a frequency of selecting a prediction formula. .   The radio according to any one of claims 1 to 4, wherein the prediction formula selection unit determines whether or not a currently selected prediction formula is valid, and selects a prediction formula only when it is determined that the prediction formula is not valid. Control device. A wireless control device according to any one of claims 1 to 5;
Multiple wireless modules;
Packet switch means for passing a packet to the wireless module selected by the wireless module selection means;
And a wireless communication device that performs wireless communication with other wireless communication devices.

Images