Communications device

Abstract

The present disclosure is a communication device that performs radio communication with a base station, said communication device comprising: a storage unit that stores radio quality information which indicates the radio quality at the time of communication between the communication device and the base station; a correction value calculation unit that calculates a correction value on the basis of a first prediction value, which has been obtained by predicting a change in the radio quality at a first time on the basis of past radio quality information stored in the storage unit, and prescribed radio quality at the time of communication between the communication device and the base station at a time immediately before the first time; and a radio quality prediction unit that calculates a third prediction value on the basis of the correction value and a second prediction value, which has been obtained by predicting a change in the radio quality at a second time after the first time on the basis of the past radio quality information stored in the storage unit.

Description

Communication device

[0001] This disclosure relates to a technology for predicting changes (fluctuations) in wireless quality when a terminal (communication device) communicates wirelessly with a base station.

[0002] In wireless communication systems targeting terminals (communication devices) such as LTE (Long Term Evolution) and 5G (5th Generation), changes (fluctuations) in wireless quality occur as the terminal moves. Therefore, if wireless quality is predicted and control is performed based on the prediction results, and the system attempts to generate communication traffic at the current location (or time) that exceeds the communication quality expected at the next location (or time), even though the communication quality at the next location (or time) will be poor, the terminal's application will become unstable at the next location (or time). For example, in a video viewing application, even if the image quality is set to 4K, if there is some event in the vicinity and a large number of people with terminals are gathered, the application will become unstable, and as a result the video will become unstable. Therefore, various methods have been proposed to predict wireless quality at the destination (see Non-Patent Document 1).

[0003] Here, the technology disclosed in Non-Patent Document 1 will be explained using Figure 5. As shown in Figure 5, the conventional terminal 120 has a communication unit 121, a prediction parameter setting unit 122, a wireless quality prediction unit 123, and a measured value DB 129.

[0004] In conventional terminals (communication devices) 120, the communication unit 121 measures the wireless quality by wirelessly communicating with the base station at each location via the antenna 121a, and stores the measured values ​​in the measured value DB 129. The wireless quality prediction unit 123 then uses prediction parameters obtained by machine learning from the prediction parameter setting unit 122 to predict the wireless quality at the next location based on multiple past measured values ​​obtained from the measured value DB 129.

[0005] Wakao, Kawamura, Moriyama, "Quality Prediction Technology for Optimal Use of Multiple Wireless Access Networks," NTT Technical Journal, vol. 32, no. 4, pp. 11-13, April 2020.

[0006] However, when predicting wireless quality based on pre-measured results at each location, if the current situation differs significantly from the conditions under which past wireless quality measurements were taken, such as when a large number of people gather at an event, there was a possibility that the prediction of wireless quality would be significantly inaccurate.

[0007] This disclosure has been made in view of the above-mentioned issues and aims to suppress a decrease in the accuracy of wireless quality prediction, even when the current situation differs to a certain extent from the situation in which wireless quality was measured in the past.

[0008] To solve the above problems, the present disclosure provides a communication device for wireless communication with a base station, comprising: a storage unit for storing wireless quality information indicating the wireless quality when the communication device communicates with the base station; a correction value calculation unit for calculating a correction value based on a first predicted value obtained by predicting a change in wireless quality at a first time based on past wireless quality information stored in the storage unit and a predetermined wireless quality when communicating with the base station at a time immediately preceding the first time; and a wireless quality prediction unit for calculating a third predicted value based on a second predicted value obtained by predicting a change in wireless quality at a second time after the first time based on past wireless quality information stored in the storage unit and the correction value.

[0009] According to this disclosure, even if the current situation differs to a certain extent from the situation in which wireless quality was measured in the past, it is possible to suppress the decline in the accuracy of wireless quality prediction.

[0010] This figure shows an example of the overall configuration of a wireless communication system according to the embodiment. This figure shows an example of the functional configuration of a terminal according to the embodiment. This flowchart shows an example of wireless quality prediction processing according to the embodiment. This figure illustrates an example of information stored in the measured value DB according to the embodiment. This is figure (1) illustrating an example of handover destination prediction processing according to the embodiment. This figure shows an example where there are multiple measured values ​​used to calculate the correction value. This is a conceptual diagram showing the case where the final predicted value of wireless quality is calculated using new correction values ​​triggered by a handover. This is a hardware configuration diagram of an electrical terminal (communication device) according to the embodiment. This flowchart shows an example of wireless quality prediction processing according to the embodiment. This figure shows an example of the functional configuration of a conventional terminal.

[0011] Embodiments of the present invention will be described below with reference to the drawings. However, the present invention is not limited to the embodiments shown below, and various modifications are possible without departing from the technical spirit of the invention.

[0012] [Overall System Configuration] Figure 1 shows an example of the overall configuration of a wireless communication system according to this embodiment. The wireless communication system 1 includes an arbitrary base station 10 and a terminal (communication device) 20 that performs wireless communication with the base station 10. The wireless communication system 1 includes, but is not limited to, wireless communication systems such as LTE (Long Term Evolution) and / or 5G (5th Generation).

[0013] The terminal 20 is a smartphone or mobile phone, and is an example of a communication device. The communication device may be any device that has a communication function to communicate wirelessly with the base station 10. For example, the communication device may be the terminal 20 or a device equipped with the functions of the terminal 20 (vehicle, mobility scooter, robot, or drone, etc.).

[0014] Furthermore, Figure 1 shows the path taken by the same terminal 20 by the user. Specifically, it shows terminal 20 at the current location p (or current time t), the location immediately preceding p-1 (or the time immediately preceding t-1), and the location immediately following p+1 (or the time immediately following t+1).

[0015] Note that position p (or time t) is an example of a first position (or first time). Position p+1 (or time t+1) is an example of a second position (or second time). Position p-1 (or time t-1) is an example of a position immediately before moving to the first position (or immediately before the first time). Each position (or each time) is updated at each timing (for example, every second) when the communication unit 21 measures the radio quality in the wireless communication with the base station 10.

[0016] [Terminal Functional Configuration] Figure 2 shows an example of the functional configuration of a terminal according to this embodiment.

[0017] As shown in Figure 2, the terminal (communication device) 20 includes an antenna 21a, a communication unit 21, a correction value calculation unit 22, and a wireless quality prediction unit 23. Each of these units is a function or means implemented by instructions from the processor 1001 in Figure 8 based on a program.

[0018] Furthermore, the terminal 20 has a measured value DB (Data Base) 29 built in the memory 1002 or auxiliary storage device 1003 shown in Figure 8. Note that the measured value DB 29 is an example of a measured value management unit or storage unit.

[0019] <Measured Value DB> The measured value DB 29 stores and manages measured values, including wireless quality, measured by the communication unit 21. The measured values ​​include location information and time information indicating the position of the terminal 20, obtained from the GPS device 1005 installed in the terminal 20 (described later). In some cases, time information may not be included.

[0020] <Communication Unit> The communication unit 21 communicates wirelessly with the base station 10 via the antenna 21a, obtains identification information of the base station 10 during this communication, and measures the wireless quality of the current wireless communication. Alternatively, the base station 10 may measure the wireless quality and transmit wireless quality information indicating this wireless quality to the terminal 20.

[0021] For example, the communication unit 21 acquires the current time and location information of the terminal 20 from the GPS device 1005, which will be described later and installed on the terminal 20. The communication unit 21 also stores the measured values, including the time and location information and the radio quality (communication status) of the radio waves from the base station 10, in the correction value calculation unit 22 and the measured value DB 29, and transmits radio quality information, which indicates at least the radio quality, to the base station 10.

[0022] <Correction Value Calculation Unit> The correction value calculation unit 22 obtains multiple past measured values ​​from the measured value DB 29 at the calculation start position (p-1) of the correction value C. Then, the correction value calculation unit 22 predicts the change in wireless quality from the changes in the multiple past measured values ​​at the calculation start position (p-1) and calculates the predicted value P at the current position p, which will be described later. １ The following is calculated. Note that instead of the current position p, multiple past measured values ​​at the calculation start position (p-1) are used because there is a time lag before the correction value C is calculated. If multiple past measured values ​​at the current position p are used, the wireless quality prediction unit 23 will use the correction value C to calculate the predicted value P. ２ Because it won't be possible to correct it in time.

[0023] Here, using Figures 3 to 5, the predicted value P １ The calculation method will be explained below. The first to third calculation methods for the predicted value shown below are calculation methods based on the location of terminal 20, but the location may be converted to time and the predicted value may be calculated from multiple past measured values ​​at the calculation start time (p-1).

[0024] (First method for calculating predicted values) Figure 3 is a conceptual diagram showing the first method for calculating predicted values ​​of wireless quality.

[0025] The first calculation method predicts changes in wireless quality from the minimum (worst-case) value of multiple past measured values ​​listed in the measured value DB29. This method assumes the use of applications on terminal 20 that require stable communication.

[0026] For example, the correction value calculation unit 22 uses multiple past measured values ​​at the calculation start position (p-1) obtained from the measured value DB 29 to create a graph of the number of measured throughputs as an example of wireless quality, as shown in Figure 3, and uses the minimum value (worst-case value) as the predicted value.

[0027] Note that it is not essential for the wireless quality prediction unit 23 to create the graph shown in FIG. 3, and the minimum value (worst value) may be obtained without using the graph.

[0028] (Second calculation method of predicted value) FIG. 4 is a conceptual diagram showing a second calculation method of the predicted value of wireless quality.

[0029] The second calculation method is a method of predicting the change in wireless quality from the average value of a plurality of past measured values described in the measured value DB29. This method assumes the use of normal applications of the terminal 20.

[0030] For example, the correction value calculation unit 22 creates a graph of the number of measured throughput as an example of wireless quality from a plurality of past measured values at the calculation start position (p-1) obtained from the measured value DB29, as shown in FIG. 4, and uses this average value as the predicted value.

[0031] Note that it is not essential for the correction value calculation unit 22 to create the graph shown in FIG. 4, and the average value may be obtained without using the graph.

[0032] (Third calculation method of predicted value) FIG. 5 is a conceptual diagram showing a third calculation method of the predicted value of wireless quality.

[0033] The third calculation method is a method of predicting the change in wireless quality from the quantile (percentile value) of a plurality of past measured values described in the measured value DB29. This method assumes the use of normal applications of the terminal 20.

[0034] For example, the correction value calculation unit 22 creates a graph of the number of measured throughput as an example of wireless quality from a plurality of past measured values at the calculation start position (p-1) obtained from the measured value DB29, as shown in FIG. 5, and uses a predetermined percentile value (for example, 20) as the predicted value.

[0035] Note that it is not essential for the correction value calculation unit 22 to create the graph shown in FIG. 5, and the predetermined percentile value may be obtained without using the graph.

[0036] (Other methods for calculating predicted values) In the first to third calculation methods described above, predicted values ​​may also be calculated by creating a graph that shows the percentage of the change in throughput as a CDF (Cumulative Distribution Function).

[0037] Furthermore, the first to third calculation methods described above may be selected by the application running on terminal 20, or set by the user of terminal 20. Alternatively, terminal 20 may have an application that executes any one of the first to third calculation methods described above.

[0038] Returning to Figure 2, the correction value calculation unit 22 calculates the predicted value P １ Based on the measured value M of a predetermined wireless quality obtained from the communication unit 21, a correction value C is calculated. Here, the method for calculating the correction value C will be explained. The first and second methods for calculating the correction value C are calculation methods based on the position of the terminal 20, but the position may be converted to time and the correction value C may be calculated at the start time (t-1) of the calculation of the correction value C.

[0039] (First method for calculating the correction value) The correction value calculation unit 22 calculates the correction value as shown in (Equation 1), using a predetermined measured value M of wireless quality obtained from the communication unit 21 and a predicted value P １ The difference, and the variation σ of multiple past measured values ​​obtained from the measured value DB29. １ Based on this, the correction value C at the current position p is calculated.

[0040] Here, the measured value M of the predetermined wireless quality, and the multiple past measured values ​​obtained from the measured value DB29, are the values ​​at the starting position (p-1) of the calculation of the correction value C. Variation σ １ This could be, for example, the standard deviation or variance of multiple past measured values ​​at the calculation start position (p-1).

[0041] (Second method for calculating the correction value) The correction value calculation unit 22 calculates the correction value as shown in (Equation 2), using a predetermined measured value M of wireless quality obtained from the communication unit 21 and a predicted value P １Based on the difference, a correction value C at the current position p is calculated. This assumes a case where there is a large discrepancy between the measured value M of a predetermined wireless quality and multiple past measured values ​​obtained from the measured value DB29, resulting in a large correction value C in (the first method for calculating the correction value).

[0042] Next, we will explain an example of a measured value M for a given wireless quality.

[0043] The first and second examples of the measured value M are examples relating to the position of terminal 20, but the position may be converted to time and examples relating to time may also be given.

[0044] (First example of measured value M) The correction value calculation unit 22 uses the measured value obtained from the communication unit 21 at the calculation start position (p-1) as the measured value M. This makes it possible to immediately reflect changes (fluctuations) in the measured value obtained from the communication unit 21 in the correction value C. In this case, if the change is large, the change in the correction value C will also be large. Note that if time is considered instead of position, the measured value M is the measured value obtained at the calculation disclosure time (t-1) of the correction value C.

[0045] (Second example of measured value M) The second example will be explained using Figure 6. Figure 6 is a diagram showing an example where there are multiple measured values ​​used to calculate the correction value. The correction value calculation unit 22 uses the median or average value of multiple radio quality values ​​obtained when communicating with the base station 10 over a predetermined distance L (a predetermined distance L immediately before the current position p) from a predetermined position (pn) traced back from the current position p to the position immediately before position p (p-1), as the measured value M. This is because the correction value C is less likely to change even if the change (fluctuation) of the measured value obtained from the communication unit 21 is large, thus reducing the influence of instantaneous changes such as fading. In this case, the time required to reflect the measured value M in the correction value is longer than in the first example. When considering time instead of position, the measured value M is the median or average value of multiple measured values ​​obtained over a predetermined time T (a predetermined time T immediately before the current time t) from a predetermined time (tn) traced back from the current time t to the time immediately before the current time t (t-1).

[0046] <Wireless Quality Prediction Unit>The wireless quality prediction unit 23 acquires, at the current position p, a plurality of past measured values at the predicted movement position (p + 1) from the measured value DB 29. Then, at the current position p, the wireless quality prediction unit 23 predicts the change in wireless quality from the changes in the plurality of past measured values, and calculates the predicted value P at the predicted movement position (p + 1). ２ is calculated. Note that the method for calculating the predicted value ２ is the same as in the case of the predicted value P １ , and the difference is that the calculation start position (p - 1) is replaced with the predicted movement position (p + 1).

[0047] Further, at the current position p, the wireless quality prediction unit 23 corrects the predicted value P using the correction value C acquired from the correction value calculation unit 22 according to (Equation 3), thereby calculating and outputting the final predicted value P of the wireless quality. ２ [[ID=ll]]to calculate and output the final predicted value P of the wireless quality. ３

[0048] Note that the variation σ ２ is, for example, the standard deviation or variance of a plurality of past measured values at the predicted movement position (p + 1), etc.

[0049] Next, an example of the correction value C applied to (Equation 3) will be described. Note that the first to third examples of the correction value C are examples related to the position of the terminal 20, but they may be examples related to time by converting the position to time.

[0050] (First Example of Correction Value C)As the correction value C applied to (Equation 3), the wireless quality prediction unit 23 calculates the predicted value P using the correction value C acquired from the correction value calculation unit 22 at any time. This can immediately reflect the change (variation) of the correction value C. In this case, if the change in the correction value is large, the change in the predicted value P ３ is also large. ３ is also large.

[0051] (Second Example of Correction Value C) As the correction value C applied to (Equation 3), the wireless quality prediction unit 23 uses, as the correction value C, the correction value C that has already been calculated by the correction value calculation unit 22 before the calculation start position (p - 1) and is being used in the calculation of the predicted value P ３ and the new correction value C calculated by the correction value calculation unit 22 at the calculation start position (p - 1) １２ Compare the two values, and if there is a difference greater than a certain amount, the correction value C used will be used. １ Replace it with a new correction value C ２ Based on this, the predicted value P ３ This calculates the predicted value P when the change (fluctuation) of the correction value C is small. ３ The changes are also small.

[0052] (Third example of correction value C) The third example will be explained using Figure 7. Figure 7 is a conceptual diagram showing the case in which a new correction value is used to calculate the final predicted value of wireless quality in the event of a handover. The wireless quality prediction unit 23 uses the correction value C calculated by the correction value calculation unit 22 based on a predetermined measured value M of wireless quality when a handover occurs and wireless communication with a new base station is started, as the correction value C to be applied to (Equation 3), and then calculates the predicted value P ３ The wireless quality prediction unit 23 calculates the following. For example, as shown in Figure 7, if a handover occurs when terminal 20 moves from the communication area 11a of base station 10a to the communication area of ​​base station 10b, the wireless quality prediction unit 23 applies the correction value C obtained by starting the calculation at the location (p-1) where the handover occurred.

[0053] Alternatively, in the event of a handover, the correction value calculation unit 22, rather than the wireless quality prediction unit 23, may start calculating a new correction value C and output this new correction value C to the wireless quality prediction unit 23.

[0054] [Hardware Configuration] Next, the hardware configuration of terminal 20 will be described using Figure 8. Figure 8 is a hardware configuration diagram of an electrical terminal (communication device) according to the embodiment. Note that terminal 20 does not necessarily have to be configured as shown in Figure 8.

[0055] As shown in Figure 8, terminal 20 includes a processor 1001, memory 1002, auxiliary storage device 1003, communication device 1004, and GPS (Global Positioning System) device 1005. Terminal 20 also includes an audio input device 1006, an audio output device 1007, a display device 1008, an imaging device 1009, a connection device 1010, and a short-range wireless communication device 1011. The various hardware components of terminal 20 are interconnected via a bus 1020.

[0056] The processor 1001 acts as a control unit that controls the entire terminal 20 and has various computing devices such as a CPU (Central Processing Unit). The processor 1001 reads various programs into memory 1002 and executes them. Note that the processor 1001 may include not only a CPU but also a GPU (Graphics Processing Unit).

[0057] The memory 1002 has main memory devices such as ROM (Read Only Memory) and RAM (Random Access Memory). The processor 1001 and the memory 1002 form a so-called computer, and the computer realizes various functions by executing various programs read into the memory 1002 by the processor 1001.

[0058] The auxiliary storage device 1003 stores various programs and various information used when these programs are executed by the processor 1001.

[0059] The communication device 1004 is a communication device for sending and receiving various types of information with other devices (including equipment, servers, and systems). The GPS device 1005 receives time information from GPS and other satellites and detects the location information of the terminal 20.

[0060] The voice input device 1006 detects voice information such as the user's voice and ambient sounds. The voice output device 1007 is a device that outputs various information received from other devices as voice.

[0061] The display device 1008 is, for example, a device that displays various information received from other devices as images.

[0062] The imaging device 1009 captures images of the user and their surroundings and generates image information.

[0063] The connection device 1010 is a connection device used to connect various sensors, external memory, etc., to the terminal 20.

[0064] The short-range wireless communication device 1011 is a wireless device for performing short-range wireless communication with other devices near the terminal 20.

[0065] [Processing of the Embodiment] Next, an example of the wireless quality prediction process according to the embodiment will be explained using Figure 9. Figure 9 is a flowchart showing an example of the wireless quality prediction process according to the embodiment. Note that Figure 9 shows just one example from the various examples described above (the first to third and other calculation methods for the predicted value, the first to second calculation methods for the correction value, the first to second examples of the measured value M, and the first to third examples of the correction value C), and is not limited to the example shown in Figure 9. Furthermore, the following processes S11 to S15 are executed repeatedly at predetermined intervals (for example, every second).

[0066] S11: The communication unit 21 receives radio information from the base station 10, measures the radio quality in the radio communication, and stores the measured values, including the radio quality and location information indicating the location where the radio information was received, in the measured value DB 29.

[0067] S12: The correction value calculation unit 22 obtains multiple past measured values ​​from the measured value DB 29 at the calculation start position (p-1) of the correction value C, and predicts a value P based on these multiple past measured values. １ Calculate.

[0068] S13: The correction value calculation unit 22 obtains the measured value M at the calculation start position (p-1) from the communication unit 21, and the predicted value P calculated in processing 12. １ Based on this, the correction value C at the current position p is calculated.

[0069] S14: The wireless quality prediction unit 23 obtains multiple past measured values ​​at the predicted movement position (p+1) from the measured value DB 29 at the current position p, and predicts the value P at the predicted movement position (p+1).２ Calculate.

[0070] S15: The wireless quality prediction unit 23 determines the correction value C and the predicted value P ２ Based on this, the final predicted value P ３ The predicted value P is calculated and output. As a result, the application on terminal 20 can see the predicted value P. ３ Based on this, the image quality and other settings will be modified.

[0071] [Main Effects of the Embodiment] According to this embodiment, at the current position p, by using the correction value C calculated using the measured values ​​at the calculation start position (p-1) (measured value M from the communication unit 21 and multiple past measured values ​​from the measured value DB 29), the predicted value P of wireless quality at the predicted movement position (p+1) can be calculated. ２ Correct it.

[0072] Similarly, according to this embodiment, at the current time t, a correction value C calculated using the measured values ​​at the calculation start time (t-1) (measured value M from the communication unit 21 and multiple past measured values ​​from the measured value DB 29) is used to predict the wireless quality P at the predicted travel time (t+1). ２ Correct it.

[0073] This has the effect of suppressing a decrease in the accuracy of predicting wireless quality, even if the current situation differs by a certain amount from the situation in which wireless quality was measured in the past.

[0074] [Supplement] (1) In the above embodiment, the measured value DB29 was used, but a pre-trained machine learning model may also be used. The machine learning model is a pre-trained model that has been trained using input data showing measured values ​​of wireless quality such as throughput at the previous location (p-1) and ground truth data showing measured values ​​of wireless quality such as throughput at the current location p. Note that when outputting predicted values ​​using a machine learning model, location information does not need to be used. That is, the machine learning model may be a model that has been trained using input data showing measured values ​​of wireless quality such as throughput at the previous time (t-1) and ground truth data showing measured values ​​of wireless quality such as throughput at the current time t.

[0075] (2) The terminal (communication device) 20 and base station 10 in this embodiment are not limited to being implemented by dedicated devices, but may also be implemented by a general-purpose computer. In that case, the program for implementing this function may be recorded on a computer-readable recording medium, and the program recorded on this recording medium may be loaded into a computer system and executed. The term "computer system" as used herein includes hardware such as an OS and peripheral devices.

[0076] (3) Furthermore, "computer-readable recording media" includes various storage devices such as flexible disks, magneto-optical disks, ROMs, CD-ROMs and other portable media, and storage devices built into computer systems. In addition, "computer-readable recording media" may also include those that dynamically hold programs for a short period of time, such as communication lines used when transmitting programs over networks such as the Internet or communication lines such as telephone lines, and those that hold programs for a certain period of time, such as volatile memory inside computer systems that act as servers or clients in such cases.

[0077] (4) The above-mentioned program may also be for the purpose of realizing some of the functions described above, or it may be for the purpose of realizing the above-mentioned functions in combination with a program already recorded in the computer system, or it may be realized using hardware such as a PLD or FPGA.

[0078] 1 Wireless communication system 10, 10a, 10b Base station 20 Terminal (example of communication device) 21 Communication unit 22 Correction value calculation unit 23 Wireless quality prediction unit 29 Measured value DB (example of storage unit, example of measured value management unit) P １ Predicted value (an example of the first predicted value) P ２ Predicted value (an example of a second predicted value) P ３ Predicted value (an example of a third predicted value)

Claims

1. A communication device for wireless communication with a base station, comprising: a storage unit for storing wireless quality information indicating the wireless quality when the communication device communicates with the base station; a correction value calculation unit for calculating a correction value based on a first predicted value obtained by predicting a change in wireless quality at a first time based on past wireless quality information stored in the storage unit, and a predetermined wireless quality when communicating with the base station at a time immediately preceding the first time; and a wireless quality prediction unit for calculating a third predicted value based on a second predicted value obtained by predicting a change in wireless quality at a second time after the first time based on past wireless quality information stored in the storage unit, and the correction value.

2. A communication device for wireless communication with a base station, comprising: a storage unit for storing wireless quality information indicating the wireless quality when the communication device communicates with the base station; a correction value calculation unit for calculating a correction value based on a first predicted value obtained by predicting a change in wireless quality at a first time based on past wireless quality information stored in the storage unit, and a predetermined wireless quality when communicating with the base station at a predetermined time immediately preceding the first time; and a wireless quality prediction unit for calculating a third predicted value based on a second predicted value obtained by predicting a change in wireless quality at a second time after the first time based on past wireless quality information stored in the storage unit, and the correction value.

3. The communication device according to claim 1 or 2, wherein the wireless quality prediction unit compares the correction value calculated by the correction value calculation unit and used in calculating the third prediction value with a new correction value calculated by the correction value calculation unit, and if there is a difference of a certain amount or more, calculates the third prediction value based on the new correction value instead of the correction value currently in use.

4. The communication device according to claim 1 or 2, wherein the wireless quality prediction unit calculates the third predicted value based on the correction value calculated by the correction value calculation unit based on the predetermined wireless quality when a handover occurs and wireless communication with a new base station is started.

Images