JP2008245011A - Wireless unit and radio network equipped therewith - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a wireless unit for distributing a packet in consideration of a use efficiency of a radio device. <P>SOLUTION: A monitor module 14 detects effective throughput in radio apparatuses 11 to 13 and outputs it to a distribution module 15. The distribution module 15 sets all the combinations of three packets PKT1 to PKT3, read from a queue 16 and the radio apparatuses 11 to 13. The distribution module 15 acquires the total time required for sending the packet in the three radio apparatuses 11 to 13 based on the effective throughput in the radio apparatuses 11 to 13 for all the combinations, and selects the combination of the packets PKT1 to PKT3 and the radio apparatuses 11 to 13, when the acquired total time required for sending the packet is minimized. Then the distribution module 15 distributes the packets PKT1 to PKT3 to the radio apparatuses 11 to 13, in accordance with the combination selected. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a wireless device and a wireless network including the wireless device, and more particularly to a wireless device that performs wireless communication using a plurality of wireless links and a wireless network including the wireless device.

  2. Description of the Related Art Conventionally, a cognitive radio network including a radio apparatus that performs radio communication simultaneously using a plurality of radio communication schemes in order to effectively use radio resources is known.

  In this cognitive radio network, a radio apparatus that can simultaneously use a plurality of radio devices estimates a different delay time for each radio device, and based on the estimated delay time, an EDF (Early Delivery Fast) and a WFQ (Weighted Fairy) Packet distribution is determined in combination with (Queuing) (Non-patent Document 1).

Thereby, it becomes possible to efficiently allocate a bandwidth to a plurality of application flows, and the throughput of the entire network can be improved.
K. Chebrolu and R. Rao, “Communication using Multiple Wire-less Interfaces”, Proc. Of IEEE WCNG '02, pp.327-333, 2002.

  However, in conventional cognitive radio networks, each wireless device determines packet distribution based on transmission time in units of packets, so it is difficult to distribute packets in consideration of the utilization efficiency of wireless devices. There was a problem that there was.

  Accordingly, the present invention has been made to solve such a problem, and an object of the present invention is to provide a wireless device capable of distributing packets in consideration of the utilization efficiency of the wireless device.

  Another object of the present invention is to provide a wireless network including a wireless device capable of distributing packets in consideration of utilization efficiency of wireless devices.

  According to the present invention, the wireless device includes n (n is an integer of 2 or more) wireless devices and packet allocation means. Each of the n radios transmits a packet assigned thereto. The packet allocating means is m so that the total required transmission time of packets in m radios (m is an integer satisfying 1 ≦ m ≦ n) among n radios is minimized. Allocate packets to other radios.

  Preferably, the required transmission time is a time length from a first timing at which one packet is assigned to one radio to a second timing at which an acknowledgment indicating that one packet has been received is received from the transmission destination. It is.

  Preferably, the packet allocating means obtains an actual transmission required time, which is an actual result of the required transmission time in one radio device, as an estimated transmission required time when one radio device attempts to transmit a packet, and the obtained estimated transmission. Packets are allocated to m radio units so that the total required time is minimized.

  Preferably, the wireless device further includes a buffer. The buffer holds the packet. Then, the packet allocating unit reads m packets from the buffer, and allocates the read m packets to m radio devices so that the total estimated transmission time is minimized.

  Preferably, the packet allocation unit includes a calculation unit, a determination unit, and an allocation unit. The calculating means calculates m estimated transmission required times in the m radio devices. The determining means determines a combination of m packets and m radio devices so that the total of m estimated transmission required times calculated by the calculating means is minimized. The assigning means assigns m packets to m radio devices according to the combination determined by the determining means.

  Preferably, the calculation means calculates an effective throughput when one radio device transmits a packet as an estimated throughput when one radio device attempts to transmit a packet, and calculates a packet size by the calculated estimated throughput. A calculation process for obtaining an estimated transmission time in one radio device is performed for m radio devices to calculate m estimated transmission time times.

  Preferably, when the number of packets stored in the buffer is smaller than the number of m radio units, the packet allocating unit reads all the packets stored in the buffer and stores the read packets in the buffer. Assign to the same number of radios as the number of packets.

  Preferably, the packet allocating unit provides each of the m radio devices with a transmittable time during which each radio device can execute a packet transmission operation, and the radio device having a positive transmittable time among the m radio devices. Assign m packets to.

  Preferably, when one packet is assigned to one radio device, the packet assigning means subtracts an estimated transmission time required for one radio device from a transmittable time before one packet is assigned to one radio device. An update process for updating the transmittable time is executed for the m radio devices, and m packets are allocated to the radio device having the positive transmittable time among the m radio devices.

  Preferably, the packet allocating unit stops allocating the packet to the wireless device whose updated transmission available time has become negative.

  Preferably, the packet allocating unit performs a reset process of adding a fixed value to the m transmittable times of the m radios when the transmittable time is negative in all of the m radios, and after the reset process M packets are assigned to m radio devices using m possible transmission times.

  According to the present invention, a wireless network includes the wireless device according to any one of claims 1 to 7.

  In the present invention, packets are distributed and transmitted to m radio units so that the total time required for packet transmission among m radio units capable of transmitting packets among n radio units is minimized. The Distributing the packets to the radios so that the total transmission time required in the m radios that can transmit the packet is minimized is that the transmission capability is relatively high, that is, the transmission time is relative. Therefore, a relatively large number of packets are distributed to short wireless devices, and a relatively low number of packets are distributed to wireless devices having a relatively low transmission capability, that is, a relatively long transmission time. Distributing a relatively large number of packets to a radio device having a relatively high transmission capability and distributing a relatively small number of packets to a radio device having a relatively low transmission capability takes the use efficiency of the radio device into consideration. Thus, the packet is distributed to the wireless devices.

  Therefore, according to the present invention, packets can be distributed in consideration of the utilization efficiency of the wireless device.

  Embodiments of the present invention will be described in detail with reference to the drawings. In the drawings, the same or corresponding parts are denoted by the same reference numerals and description thereof will not be repeated.

  FIG. 1 is a schematic diagram of a wireless network according to an embodiment of the present invention. A wireless network 10 according to an embodiment of the present invention includes a cognitive base station 1 and wireless devices 2 to 5.

  The cognitive base station 1 is connected to the cognitive gateway 30 by a wired cable 31. The cognitive gateway 30 is connected to the network 40 by a wired cable 41.

  The cognitive base station 1 performs wireless communication with the wireless devices 2 to 5 using a plurality of wireless communication methods by a method described later, and also performs communication with the cognitive gateway 30 via the wired cable 31.

  Each of the wireless devices 2 to 5 performs wireless communication with the cognitive base station 1 by a method described later.

  Accordingly, each of the wireless devices 2 to 5 is connected to the cognitive base station 1, the wired cable 31, the cognitive gateway 30, the wired cable 41, and a wireless device that accesses another cognitive gateway (not shown) via the network 40. Wireless communication can be performed.

  FIG. 2 is a schematic block diagram showing the configuration of the cognitive base station 1 shown in FIG. The cognitive base station 1 includes radio devices 11 to 13, a monitor module 14, a distribution module 15, a queue 16, and a communication module 17.

  The radio devices 11 to 13 belong to the data link layer and have queues 11A, 12A, and 13A, respectively. And the radio | wireless machines 11-13 perform radio | wireless communication between the radio | wireless apparatuses 2-5 by a mutually different radio | wireless communication system. More specifically, each of the wireless devices 11 to 13 transmits the packet distributed by the distribution module 15 to the wireless devices 2 to 5 and outputs the packet received from the wireless devices 2 to 5 to the communication module 17. .

  The monitor module 14 belongs to the cognitive layer, and detects the effective throughput in each of the radio devices 11 to 13 at regular intervals by a method described later. Then, the monitor module 14 outputs the detected plurality of effective throughputs to the distribution module 15.

When the distribution module 15 belongs to the cognitive layer and receives a plurality of effective throughputs from the monitor module 14, the distribution module 15 obtains the time required for transmission in each of the radio devices 11 to 13 based on the received plurality of effective throughputs by a method described later. . In addition, the distribution module 15 searches the number P ^q of packets stored in the queue 16 at a constant period. Then, when the number of packets P ^q is positive (P ^q > 0), the distribution module 15 adds the packets read from the queue 16 by the method described later to the total required transmission time in the three radios 11 to 13. Is distributed to the radios 11 to 13 so that the signal becomes minimum. Further, the distribution module 15 uses a time resource (RT: Resource Time) for managing the load on each of the radio devices 11 to 13 when the packet is allocated to the radio devices 11 to 13 based on the time required for transmitting the packet. The load balance in the radio devices 11 to 13 is achieved by a method described later.

  The queue 16 belongs to the cognitive layer and holds the packet PKT received from the communication module 17.

  The communication module 17 includes various modules belonging to the Internet layer, the transport layer, and the application layer. Then, the communication module 17 generates a TCP (Transmission Control Protocol) packet based on the data, and generates an IP packet based on the generated TCP packet. The IP packet is composed of an IP header and an IP data part for storing the TCP packet. Then, when generating the TCP packet, the communication module 17 stores the generated TCP packet in the IP data part to generate an IP packet, and stores the generated IP packet in the queue 16.

  1 has the same configuration as that of the cognitive base station 1 shown in FIG.

  FIG. 3 is a conceptual diagram of the required transmission time. The wireless devices 11 to 13 have queues 11A, 12A, and 13A, respectively. In each of the wireless devices 11 to 13, the packet PKT is stored in the queues 11A, 12A, and 13A at the timing t1, and reaches the head of the queues 11A, 12A, and 13A at the timing t2.

  Thereafter, the packet PKT is extracted from the queues 11A, 12A, and 13A at the timing t3 and transmitted to the transmission destination. Then, the wireless device (any of the wireless devices 11 to 13) that has transmitted the packet PKT receives an acknowledgment (ACK: Acknowledge) from the transmission destination at timing t4.

  In the present invention, the time from timing t1 to timing t4 is measured, and the average value of the measured time for a (a is an integer of 2 or more) packets is required for transmission in each of the radio devices 11-13. Defined as time.

  Thus, the transmission required time is the time from when the packet PKT is stored in the queues 11A, 12A, and 13A to the transmission destination in each of the wireless devices 11 to 13 until the ACK is returned from the transmission destination (timing t1 To the timing t4) is defined as an average value for a packets.

  Therefore, the time required for transmission is a concept including retransmission in a MAC (Media Access Control) layer, and may vary depending on the wireless communication environment of wireless communication performed by each wireless device 11 to 13 according to each wireless communication method. In addition, since the ACK frame varies depending on each wireless communication method, the required transmission time can vary depending on the wireless communication method.

A method for obtaining the transmission time will be described. The monitor module 14 sets a packet transmitted to a certain destination x as p _x = {p ₁ , p ₂ ,..., P _m }, and determines the radio link l _k (which of the radio links by the radios 11 to 13). The effective throughput for the latest a packets for a) is detected by the following equation.

  Note that the monitor module 14 detects the effective throughput by detecting the number of packets successfully transmitted to the transmission destination per unit time and calculating the number of bits included in the detected number of packets.

Monitor module 14, the effective throughput ^R e _(l ¹¹⁾ in the radio 11 to 13 detects the ~R e _{(l 13),} distributes the effective throughput ^R e in which the detected _{^{(l 11) ~R e (l}} 13) Output to module 15.

The distribution module 15 receives the effective throughputs R ^e (l ₁₁ ) to R ^e (l ₁₃ ) from the monitor module 14, and receives the effective throughputs R ^e (l ₁₁ ) to R ^e (l ₁₃ ) and the packet size S ( P _n ) is substituted into the following equation to determine the transmission time T ^e (P _n , l _k ) in each radio link.

The required transmission time T ^e (P _n , l _k ) in each radio link can vary depending on the combination of the packet P _n and the radios 11 to 13. As is clear from the equation (2), the transmission required time T ^e (P _n , l _k ) in each radio link is equivalent to the packet size S (P _n ) divided by the effective throughput R ^e (l _k ). This is because the packet size S (P _n ) and effective throughput R ^e (l _k ) to be substituted into Expression (2) vary depending on the combination of the packet P _n and the radios 11 to 13.

Assume that the packets P _{1 to} P ₃ are distributed to the radios 11 to 13. In this case, the packets P _{1 to} P ₃ have packet sizes S (P ₁ ) to S (P ₃ ), respectively.

When distributing the packets P _{1 to} P ₃ to the radio devices 11 to 13, respectively, the distribution module 15 substitutes the packet size S (P ₁ ) and effective throughput R ^e (l ₁₁ ) into the equation (2) to set the radio device. calculates the transmission time required ^T e _{(P 1, l 11)} at 11, a packet size S _{(P 2)} and the effective throughput ^R e _{(l 12)} sends the required time in the radio 12 into equation (2) T ^e (P ₂ , l ₁₂ ) is calculated, the packet size S (P ₃ ) and the effective throughput R ^e (l ₁₃ ) are substituted into equation (2), and the required transmission time T ^e (P ₃ , L ₁₃ ).

Further, when distributing the packets P ₂ , P ₃ , and P ₁ to the radio devices 11 to 13, the distribution module 15 sets the packet size S (P ₂ ) and the effective throughput R ^e (l ₁₁ ) to Equation (2). Substituting and calculating the required transmission time T ^e (P ₂ , l ₁₁ ) in the radio device 11 and substituting the packet size S (P ₃ ) and effective throughput R ^e (l ₁₂ ) into the equation (2) transmission required time in ^{_{12 T e (P 3, l}} 12) calculates a packet size S _{(P 1)} and the effective throughput ^R e _{(l 13)} sends the required time in the radio 13 into equation (2) Calculate T ^e (P ₁ , l ₁₃ ). As described above, the transmission required time T ^e (P _n , l _k ) in each wireless device varies depending on the combination of the packet P _n and the wireless link l _k by the wireless devices 11 to 13.

The required transmission time T ^e (P _n , l _k ) calculated by the equation (2) is a record of the required transmission time in each of the radio devices 11 to 13 when each of the radio devices 11 to 13 transmits a packets. Yes, it reflects the packet size.

  In the present invention, the required transmission time for a packets that have already been transmitted is set as the estimated required transmission time for the packet to be assigned. That is, when distribution of packets to the radio devices 11 to 13 occurs, each radio device 11 to 13 transmits the time required for transmission of a packets actually transmitted before that time, and each radio device 11 to 13 transmits a packet to be allocated. The time required to do this.

Next, a packet distribution method will be described. When the distribution module 15 calculates the estimated transmission required time T ^e (P _n , l _k ) in each of the radio devices 11 to 13 according to the above-described equation (2), the calculated estimated transmission required time T ^e (P _n , l _k ) is substituted into the following equation to calculate the total T ^e _total of estimated transmission times in the three radios 11 to 13.

In Expression (3), N is the number of radios 11 to 13. Further, a (P _i ) is a set indicating a combination of i packets and three radio devices 11 to 13 when i packets are distributed to the three radio devices 11 to 13.

Distribution module 15, when calculating the total ^T _{e total} estimated transmission duration, total ^T _{e total} estimated transmission time required is minimized by assigning the total ^T _{e total} of the calculated estimated transmission duration to the following equation The optimum combination of the current packet P _n and the radios 11 to 13 is determined.

In Expression (4), / A ^* _K is a vector set indicating the optimal combination of the packet P _n and the radios 11 to 13.

Then, when the distribution module 15 determines the optimum combination of the packet P _n and the wireless devices 11 to 13, the distribution module 15 distributes the packet to the wireless devices 11 to 13 according to the determined optimal combination.

  When distributing the packet to the wireless devices 11 to 13, the distribution module 15 gives time resources RT to the wireless devices 11 to 13 and balances the load in the wireless devices 11 to 13.

  FIG. 4 is a diagram for explaining the concept of the time resource RT. When the distribution of the packet is started, the distribution module 15 gives the initial value RT0 of the time resource RT to each of the radio devices 11 to 13 (see (a) of FIG. 4).

  Then, the distribution module 15 reads the three packets PKT1 to PKT3 from the queue 16, and sets the optimum combination of the read three packets PKT1 to PKT3 and the three radios 11 to 13 to the three radios 11 ˜13 so that the total of the three estimated transmission times is minimized. As a result, the packet PKT1 is distributed to the wireless device 11, the packet PKT3 is distributed to the wireless device 12, and the packet PKT2 is distributed to the wireless device 13.

Then, the distribution module 15 subtracts the estimated required transmission time T ^e (P _n , l ₁₁ ) of the wireless device 11 from the initial value RT0 of the time resource RT, and changes the time resource RT in the wireless device 11 from the initial value RT0 to the value RT1_11. Update to Similarly, the distribution module 15 subtracts the estimated required transmission time T ^e (P _n , l ₁₂ ) of the radio device 12 from the initial value RT0 of the time resource RT to obtain the time resource RT in the radio device 12 from the initial value RT0. The value RT1_12 is updated, and the estimated transmission time T ^e (P _n , l ₁₃ ) of the wireless device 13 is subtracted from the initial value RT0 of the time resource RT to change the time resource RT in the wireless device 13 from the initial value RT0 to the value RT1_13. Update (see FIG. 4B).

Thereafter, the distribution module 15 distributes the packet to each of the radio devices 11 to 13 from the current value of the time resource RT in each of the radio devices 11 to 13 to estimate the required transmission time T ^e (P _n , l _k ) is subtracted to update the time resource RT in each of the radio devices 11-13. Then, the time resource RT is changed from positive to negative by the distribution module 15 repeatedly updating the time resource RT in each of the radio devices 11 to 13. Therefore, the time resource in this invention consists of a concept that changes from plus to minus.

  FIG. 5 is a diagram illustrating a first example of packet distribution using the time resource RT. As a result of the distribution module 15 repeating the operation of distributing packets to the radio devices 11 to 13 by the method described above, the time resource RT11 in the radio device 11 is updated to the value RT2_11, and the time resource RT12 in the radio device 12 is updated to the value RT2_12. And the time resource RT13 in the wireless device 13 is updated to the value RT2_13. In this case, the values RT2_11 and RT2_13 are positive, and the value RT2_12 is negative.

  Then, the distribution module 15 excludes the radio device 12 having the time resource RT12 having the negative value RT2_12 from the object of packet distribution, and the two radio devices having the time resources RT11 and 13 having the positive values RT2_11 and RT2_13. Packets PKT4 and PKT5 are distributed to 11 and 13, respectively. That is, the distribution module 15 regards the radios 11 and 13 having the positive time resources RT11 and 13 as available radios, and the same number of packets PKT4 and PKT4 as the available radios 11 and 13 are available. PKT 5 is distributed to the radios 11 and 13.

  FIG. 6 is a diagram illustrating a second example of packet distribution using the time resource RT. The time resource RT11 of the wireless device 11 is composed of the value RT3_11 (> 0), the time resource RT12 of the wireless device 12 is composed of the value RT3_12 (> 0), and the time resource RT13 of the wireless device 13 is composed of the value RT3_13 (> 0). ). Therefore, the distribution module 15 determines that all of the wireless devices 11 to 13 are usable wireless devices.

  Then, the distribution module 15 detects the number (= 2) of packets PKT6 and PKT7 stored in the queue 16.

  Then, the distribution module 15 compares the number of packets PKT6 and PKT7 (= 2) stored in the queue 16 with the number of available radios 11 to 13 (= 3) and stores the packets in the queue 16. It is determined that the number of packets is smaller than the number of available radios.

  The distribution module 15 reads the packets PKT6 and PKT7 from the queue 16, and sets the optimum transmission required time of the radios that do not distribute packets to “0” as the optimal combination of the read packets PKT6 and PKT7 and the radios 11 and 13. Thus, the packets PKT6 and PKT7 are distributed to the radios 11 and 13, respectively, according to the determined optimum combination.

  As described above, when the number of packets stored in the queue 16 is smaller than the number of available radios, the same number of radios as the number of packets stored in the queue 16 has the minimum sum of estimated transmission times. Select based on the combination. Thereby, the further reduction of the sum total of the time resources RT of the radio devices 11 to 13 is suppressed, and the time resources RT between the plurality of radio devices 11 to 13 can be efficiently used.

  FIG. 7 is a diagram illustrating a third example of packet distribution using the time resource RT. The time resource RT11 of the wireless device 11 is composed of the value RT4_11 (<0), the time resource RT12 of the wireless device 12 is composed of the value RT4_12 (<0), and the time resource RT13 of the wireless device 13 is the value RT4_13 (<0). (See (a) of FIG. 7).

  When all of the time resources RT11 to RT13 in the wireless devices 11 to 13 become negative, the distribution module 15 adds a constant value T to all of the time resources RT11 to RT13 in order to reset the time resources RT11 to RT13. As a result, the time resource RT11 of the wireless device 11 is updated to the value RT5_11 (> 0), the time resource RT12 of the wireless device 12 is updated to the value RT5_12 (> 0), and the time resource RT13 of the wireless device 13 is The value is updated to the value RT5_13 (> 0) (see FIG. 7B).

  Then, the distribution module 15 uses the updated time resources RT11 to RT13 to balance the load among the wireless devices 11 to 13 when distributing packets to the wireless devices 11 to 13 by the method described above.

  As shown in FIG. 7B, the constant value T is added to the values RT4_11 to RT4_13, so that the updated values RT5_11 to RT5_13 reflect the values RT4_11 to RT4_13 before the update. That is, a value obtained by subtracting the minus amount before update from the constant value T is given to the radio devices 11 to 13 as time resources RT11 to RT13.

  As a result, a relatively large number of packets are distributed to a radio device having a relatively long time resource RT, and a relatively small number of packets are distributed to a radio device having a relatively short time resource RT. Therefore, it is possible to balance the load between the wireless devices 11 to 13 when distributing the signals to 13.

  Since the above-described radio with a positive time resource RT is a target of packet distribution, the time resource RT is hereinafter referred to as “transmittable time”.

FIG. 8 is a flowchart for explaining the packet distribution operation. When a series of operations is started, the distribution module 15 determines whether or not the queue 16 is empty (step S1), and counts the number of packets P ^q stored in the queue 16 when the queue 16 is not empty. (Step S2).

  Then, the distribution module 15 determines whether or not the transmittable time of all wireless links (wireless links by all of the wireless devices 11 to 13) is negative (step S3).

  When it is determined in step S3 that the transmittable time of all radio links is negative, the distribution module 15 adds a fixed value T to the transmittable time of all radio links (step S4). Thereby, the transmittable time is updated.

On the other hand, when it is determined in step S3 that the transmittable time of all the radio links is not negative, or after step S4, the distribution module 15 counts radio links by radios having a positive transmittable time. counting the number l ^a of available radio links by (step S5).

The distribution module 15 determines whether the number of packets ^{P q} is smaller than the number ^{l a} radio link (step S6).

In step S6, when the number of packets ^{P q} is determined to be smaller than the number ^{l a} radio link, the distribution module 15 reads out the ^{P q} number of packets from the queue 16 (step S7).

On the other hand, in step S6, when the number of packets ^{P q} is determined to not less than the number ^{l a} radio link, the distribution module 15 reads a ^{l a} number of packets from the queue 16 (step S8).

  Then, after step S7 or step S8, the distribution module 15 performs packet allocation (step S9), and subtracts the estimated transmission time from the transmittable time in each radio device to which the packet is allocated, to obtain the transmittable time. By updating, the transmission possible time of the radio link to which the packet is allocated is updated (step S10).

  Thereafter, the series of operations returns to step S1, and step S1 to step S10 described above are repeatedly executed until it is determined in step S1 that the queue 16 is empty. When it is determined in step S1 that the queue 16 is empty, the series of operations ends.

FIG. 9 is a flowchart for explaining the detailed operation of step S9 shown in FIG. After step S7 or step S8 shown in FIG. 8, the distribution module 15, with l all combinations of ^a number of packets and l ^a number of radio (or P ^q number of packets and P ^q-number of radio All combinations) are set (step S91). For example, when three packets PKT1 to PKT3 are distributed to three wireless devices 11 to 13, all combinations of the three packets PKT1 to PKT3 and the three wireless devices 11 to 13 are [PKT1 → wireless Wireless device 11 / PKT2 → wireless device 12 / PKT3 → wireless device 13], [PKT1 → wireless device 11 / PKT3 → wireless device 12 / PKT2 → wireless device 13], [PKT2 → wireless device 11 / PKT1 → wireless device 12 / PKT3 → wireless device 13], [PKT3 → wireless device 11 / PKT1 → wireless device 12 / PKT2 → wireless device 13], [PKT2 → wireless device 11 / PKT3 → wireless device 12 / PKT1 → wireless device 13], [PKT3 → wireless device] 6 combinations of machine 11 / PKT2 → wireless machine 12 / PKT1 → wireless machine 13] are set.

  Then, the distribution module 15 calculates the total estimated transmission time for each of all the set combinations using the above-described equations (1) to (3) (step S92).

  Thereafter, the distribution module 15 selects the minimum estimated transmission required time among the calculated total estimated transmission required times (step S93), and the selected minimum estimated transmission required time is obtained. A combination of the packet and the wireless device is detected (step S94).

  Then, the distribution module 15 distributes the packet to the radio according to the detected combination (step S95). And after step S95, a series of operation | movement transfers to step S10 shown in FIG.

  As described above, in the present invention, the packets are distributed to the radio devices so that the total estimated transmission time in the available radio devices among the radio devices 11 to 13 is minimized. Then, distributing the packets to the radios so that the total estimated transmission time in the available radios is minimized is a radio with a relatively high transmission capability, that is, with a relatively short estimated transmission time. A relatively large number of packets are distributed to the mobile stations, and the transmission capability is relatively low, that is, relatively few packets are distributed to the wireless devices having a relatively long estimated transmission time. Distributing a relatively large number of packets to a radio device having a relatively high transmission capability and distributing a relatively small number of packets to a radio device having a relatively low transmission capability takes the use efficiency of the radio device into consideration. Thus, the packet is distributed to the wireless devices.

  Therefore, according to the present invention, packets can be distributed in consideration of the utilization efficiency of the wireless device.

  An operation in which the radio apparatus 2 shown in FIG. 1 performs communication with the cognitive base station 1 will be described. When a series of operations is started, the communication module 17 of the wireless device 2 generates a packet and stores it in the queue 16, and the queue 16 holds the packet.

Then, the monitor module 14 of the wireless device 2 detects the effective throughputs R ^e (l ₁₁ ) to R ^e (l ₁₃ ) in the queues 11A, 12A, and 13A of the wireless devices 11 to 13 by the method described above, and detects them. The effective throughputs R ^e (l ₁₁ ) to R ^e (l ₁₃ ) are output to the distribution module 15.

Then, the distribution module 15, in accordance with the flowchart shown in FIGS. 8 and 9, from the queue 16 reads the packet of the ^{l a} number (or ^{P q} pieces), wireless packets ^{l a} number thus read out (or ^{P q} pieces) Distribute to available radios among machines 11-13. Further, the distribution module 15 balances the load of packet distribution using the time resource RT.

  Then, the radio device to which the packet is assigned transmits the packet to the cognitive base station 1.

  The wireless device of the cognitive base station 1 (available wireless device among the wireless devices 11 to 13) receives the packet transmitted from the wireless device 2 and outputs the received packet to the communication module 17. Then, the communication module 17 of the cognitive base station 1 receives the packet transmitted from the wireless device 2.

  The cognitive base station 1 transmits the packet to the wireless device 2 by the same operation as the operation of the wireless device 2 described above.

  Then, the wireless device (available wireless device among the wireless devices 11 to 13) of the wireless device 2 receives the packet transmitted from the cognitive base station 1 and outputs the received packet to the communication module 17. Accordingly, the communication module 17 of the wireless device 2 receives the packet transmitted from the cognitive base station 1.

  Communication between the radio apparatuses 3 to 5 and the cognitive base station 1 is performed in the same manner as communication between the radio apparatus 2 and the cognitive base station 1.

  A simulation result when packet distribution according to the present invention is performed will be described. In the simulation, the wireless link ML1 based on IEEE802.11a and the wireless link ML2 based on IEEE802.11b are used as wireless links, the transmission rate in the wireless link ML1 is set to 12 Mbps, and the transmission rate in the wireless link ML2 is 5 Set to 5 Mbps.

  FIG. 10 is a diagram showing the relationship between the throughput and the average delay when the time resource RT is changed. In FIG. 10, the horizontal axis represents the throughput, and the vertical axis represents the average delay. Further, the time resource RT has been changed to 1 ms, 5 ms, 10 ms, 20 ms, and 50 ms.

  From the results shown in FIG. 10, the highest throughput and the shortest average delay were obtained when the time resource RT was 5 ms.

  Therefore, the following simulation results are simulation results when the time resource RT of 5 ms is used.

  FIG. 11 is a diagram illustrating the relationship between throughput and input traffic volume. In FIG. 11A, the vertical axis represents the average throughput, and the horizontal axis represents the amount of input traffic. A curve k1 shows the relationship between the average throughput and the amount of input traffic when packets are distributed to the radio links ML1 and ML2 by the round robin (RR) method, and the curve k2 shows a weighted round robin (WRR: The relationship between the average throughput and the amount of input traffic when a packet is distributed to the radio links ML1 and ML2 by the Weighted Round Robin) method is shown. A curve k3 is a graph when the packet is transmitted by IEEE802.11a (= radio link ML1). The relationship between the average throughput and the input traffic amount is shown, and a curve k4 shows the relationship between the average throughput and the input traffic amount when packets are distributed to the radio links ML1 and ML2 by the method according to the present invention.

  In FIG. 11B, the vertical axis represents the average effective throughput, and the horizontal axis represents the amount of input traffic. Curve k5 shows the relationship between the average effective throughput and the amount of input traffic when packets are distributed to the radio link ML1 performing radio communication by IEEE802.11a by the round robin method, and the curve k6 is radio by IEEE802.11b. The relationship between the average effective throughput and the amount of input traffic when packets are distributed to the radio link ML2 performing communication by the round robin method is shown. Further, a curve k7 shows the relationship between the average effective throughput and the amount of input traffic when packets are distributed to the wireless link ML1 that performs wireless communication by IEEE802.11a by the weighted round robin method, and the curve k8 shows the IEEE802.11b. Shows the relationship between the average effective throughput and the amount of input traffic when packets are distributed to the wireless link ML2 performing wireless communication by the weighted round robin method. Further, a curve k9 shows the relationship between the average effective throughput and the amount of input traffic when packets are distributed to the radio link ML1 that performs radio communication by IEEE802.11a according to the method of the present invention, and the curve k10 is shown by IEEE802.11b. The relationship between the average effective throughput and the amount of input traffic when packets are distributed to the radio link ML2 that performs radio communication by the method according to the present invention is shown.

  As shown in FIG. 11A, the average throughput when packets are distributed by the round robin method is lower than the average throughput when packets are distributed by the communication method according to IEEE 802.11a (see curves k1 and k3). . In the round robin method, packets are fairly distributed to the radio links ML1 and M12 without considering the state of the radio links ML1 and ML2. As a result, in the round robin method, the capacity of each of the radio links ML1 and ML2 cannot be used effectively, and thus communication bands cannot be effectively aggregated.

  On the other hand, the weighted round robin method and the method according to the present invention achieve higher average throughput than IEEE 802.11a (see curves k1, k2, and k4). This is because the weighted round robin method and the method according to the present invention distribute packets in consideration of the states of the radio links ML1 and ML2.

  As shown in FIG. 11B, the average effective throughput when packets are distributed to the radio link ML2 that performs radio communication by IEEE802.11b by the method according to the present invention is the radio link ML2 that performs radio communication by IEEE802.11b. It is lower than the average effective throughput when packets are distributed by the round robin method or the weighted round robin method (see curves k6, k8, and k10).

  On the other hand, the average effective throughput when packets are distributed to the wireless link ML1 that performs wireless communication by IEEE802.11a by the method according to the present invention is the round robin method or weighted round robin to the wireless link ML1 that performs wireless communication by IEEE802.11a. It is sufficiently higher than the average effective throughput when packets are distributed by the method (see curves k5, k7, and k9).

  When a packet is distributed according to the present invention, a packet having a relatively large size is distributed to the wireless link ML1 that performs wireless communication by IEEE802.11a, and a packet having a relatively small size is wirelessly communicated by IEEE802.11b. Is distributed to the radio link ML2. In the system according to the present invention, the packet is set such that the sum of the estimated transmission time of the radio link ML1 that performs radio communication by IEEE802.11a and the estimated transmission time of the radio link ML2 that performs radio communication by IEEE802.11b is minimized. Is distributed to the two radio links ML1 and ML2.

  As a result, the total average effective throughput when packets are distributed to the radio links ML1 and ML2 according to the present invention is reduced in the radio link ML2, but the capacity of the radio link ML1 is expanded, so that the round robin method or weighted It becomes higher than the total average effective throughput when packets are distributed by the round robin method.

  FIG. 12 is a diagram illustrating the relationship between the average data size of packets held in the queues 11A, 12A, and 13A of the wireless devices 11 to 13 and the amount of input traffic. In FIG. 12, the vertical axis represents the average data size of packets held in the MAC layer queue, and the horizontal axis represents the amount of input traffic.

  Curve k11 shows the relationship between the average data size and the amount of input traffic when packets are distributed to the radio link ML1 that performs radio communication by IEEE802.11a by the round robin method, and the curve k12 is wireless by IEEE802.11b. The relationship between the average data size and the amount of input traffic when packets are distributed to the radio link ML2 that performs communication by the round robin method is shown. Further, a curve k13 shows the relationship between the average data size and the amount of input traffic when packets are distributed to the wireless link ML1 that performs wireless communication by IEEE802.11a by the weighted round robin method, and the curve k14 shows the IEEE802.11b. Shows the relationship between the average data size and the amount of input traffic when packets are distributed to the wireless link ML2 performing wireless communication by the weighted round robin method. Further, a curve k15 shows the relationship between the average data size and the amount of input traffic when packets are distributed to the radio link ML1 that performs radio communication by IEEE802.11a according to the method according to the present invention, and the curve k16 is shown by IEEE802.11b. The relationship between the average data size and the amount of input traffic when packets are distributed to the radio link ML2 that performs radio communication by the method according to the present invention is shown.

  The average data size of the packet held in the MAC layer queue when the packet is distributed by the round robin method to the wireless link ML2 that performs wireless communication by IEEE802.11b is larger than that when the packet is distributed by the method according to the present invention. Increase with lower input traffic (see curves k12, k16).

  On the other hand, the average data size of packets held in the MAC layer queue when packets are distributed to the wireless link ML1 performing wireless communication by IEEE802.11a by the round robin method is such that the amount of input traffic is equal for fair packet distribution. Even if it increases, it hardly increases (see curve k11).

  This is because the number of packets distributed to the wireless link ML2 performing wireless communication by IEEE802.11b is large, and the number of packets distributed to the wireless link ML1 performing wireless communication by IEEE802.11a is small.

  The average data size of the packet held in the queue of the MAC layer when the packet is distributed by the weighted round robin method is almost the same as that when the packet is distributed by the round robin method (see curves k11 to k14).

  On the other hand, the average data size of the packet held in the queue of the MAC layer when the packet is distributed to the wireless link ML1 that performs wireless communication by IEEE802.11a by the method according to the present invention is the wireless link that performs wireless communication by IEEE802.11b. This is similar to the average data size of packets held in the queue of the MAC layer when packets are distributed to ML2 by the method according to the present invention (see curves k15 and k16). This means that the traffic is distributed over two well-balanced radio links ML1, ML2 (radio link for radio communication by IEEE802.11a and radio link for radio communication by IEEE802.11b). .

  FIG. 13 is a diagram showing the relationship between the average delay and the throughput when the input traffic volume increases from 4 Mbps to 14 Mbps. In FIG. 13, the vertical axis represents the average delay, and the horizontal axis represents the throughput.

  Curve k17 shows the relationship between average delay and throughput when packets are distributed by the round robin method, and curve k18 shows the relationship between average delay and throughput when packets are distributed by the weighted round robin method. Curve k19 shows the relationship between the average delay and the throughput when the packet is transmitted by IEEE802.11a, and curve k20 shows the relationship between the average delay and the throughput when the packet is distributed by the method according to the present invention.

  The throughput when packets are distributed and transmitted to the radio links ML1 and ML2 by the method according to the present invention is higher than the throughput when packets are transmitted by round robin, weighted round robin, and a single IEEE 802.11a. The average delay when packets are distributed and transmitted to the radio links ML1 and ML2 by the method according to the invention is lower than the average delay when packets are transmitted by round robin, weighted round robin, and a single IEEE 802.11a ( Curves k17 to k20). Therefore, wireless communication characteristics are improved by using the packet distribution method according to the present invention.

  FIG. 14 is a diagram illustrating the relationship between the throughput and the amount of input traffic when fading occurs. In FIG. 14A, the vertical axis represents the average throughput, and the horizontal axis represents the amount of input traffic. A curve k21 shows the relationship between the average throughput and the input traffic amount when packets are distributed to the radio links ML1 and ML2 by the round robin method, and a curve k22 shows the packets by the weighted round robin method. The curve shows the relationship between the average throughput and the input traffic amount when the packets are transmitted by IEEE802.11a, and the curve k24 shows the relationship between the average throughput and the input traffic amount when the packets are transmitted by IEEE802.11a. The relationship between the average throughput and the amount of input traffic when packets are distributed to the radio links ML1 and ML2 by the method is shown.

  In FIG. 14B, the vertical axis represents the average effective throughput, and the horizontal axis represents the amount of input traffic. Curve k25 shows the relationship between the average effective throughput and the amount of input traffic when packets are distributed to the radio link ML1 performing radio communication by IEEE802.11a by the round robin method, and the curve k26 is wireless by IEEE802.11b. The relationship between the average effective throughput and the amount of input traffic when packets are distributed to the radio link ML2 performing communication by the round robin method is shown. Further, a curve k27 shows the relationship between the average effective throughput and the amount of input traffic when a packet is distributed to the wireless link ML1 that performs wireless communication by IEEE802.11a by the weighted round robin method, and the curve k28 shows the IEEE802.11b. Shows the relationship between the average effective throughput and the amount of input traffic when packets are distributed to the wireless link ML2 performing wireless communication by the weighted round robin method. Further, a curve k29 shows the relationship between the average effective throughput and the amount of input traffic when packets are distributed to the radio link ML1 that performs radio communication by IEEE802.11a by the method according to the present invention, and the curve k30 is shown by IEEE802.11b. The relationship between the average effective throughput and the amount of input traffic when packets are distributed to the radio link ML2 that performs radio communication by the method according to the present invention is shown.

  FIG. 14A shows the average throughput and the amount of input traffic when the Leishan factor K indicating the degree of fading is “0” (a smaller value of the Leishan factor K means stronger fading). FIG. 14B shows the relationship between the average effective throughput and the amount of input traffic when the Leishan factor K is “0”.

  In all cases, the average throughput drops due to packet errors due to fading (see curves k21-k24).

  However, the average throughput when packets are distributed and transmitted by the method according to the present invention is higher than the average throughput when packets are transmitted by IEEE 802.11a even when fading occurs (see curves k23 and k24). .

  The average effective throughput shown in (b) of FIG. 14 is substantially the same as the average effective throughput shown in (b) of FIG.

  FIG. 15 is a diagram illustrating the relationship between the average data size of packets held in the queue of each wireless device and the amount of input traffic when fading occurs.

  In FIG. 15, the vertical axis represents the average data size of packets held in the MAC layer queue, and the horizontal axis represents the amount of input traffic.

  Curve k31 shows the relationship between the average data size and the amount of input traffic when packets are distributed to the radio link ML1 performing radio communication by IEEE802.11a by the round robin method, and the curve k32 is radio by IEEE802.11b. The relationship between the average data size and the amount of input traffic when packets are distributed to the radio link ML2 that performs communication by the round robin method is shown. Furthermore, a curve k33 shows the relationship between the average data size and the amount of input traffic when packets are distributed to the wireless link ML1 that performs wireless communication by IEEE802.11a by the weighted round robin method, and the curve k34 shows the IEEE802.11b. Shows the relationship between the average data size and the amount of input traffic when packets are distributed to the wireless link ML2 performing wireless communication by the weighted round robin method. Further, curve k35 shows the relationship between the average data size and the amount of input traffic when packets are distributed to the radio link ML1 that performs radio communication by IEEE802.11a according to the method of the present invention, and the curve k36 is shown by IEEE802.11b. The relationship between the average data size and the amount of input traffic when packets are distributed to the radio link ML2 that performs radio communication by the method according to the present invention is shown.

  The average data size of packets held in the MAC layer queue when fading occurs is the same as the average data size of packets held in the MAC layer queue when fading does not occur (see FIG. 12).

  FIG. 16 is a diagram illustrating the relationship between the average delay and the throughput when the input traffic volume increases from 4 Mbps to 14 Mbps when fading occurs.

  In FIG. 16, the vertical axis represents the average delay, and the horizontal axis represents the throughput. Curve k37 shows the relationship between average delay and throughput when packets are distributed by the round robin method, and curve k38 shows the relationship between average delay and throughput when packets are distributed by the weighted round robin method. Curve k39 shows the relationship between the average delay and the throughput when the packet is transmitted by IEEE802.11a, and curve k40 shows the relationship between the average delay and the throughput when the packet is distributed by the method according to the present invention.

  The throughput when packets are distributed and transmitted to the radio links ML1 and ML2 by the method according to the present invention is higher than the throughput when packets are transmitted by round robin, weighted round robin, and a single IEEE 802.11a. The average delay when packets are distributed and transmitted to the radio links ML1 and ML2 by the method according to the invention is lower than the average delay when packets are transmitted by round robin, weighted round robin, and a single IEEE 802.11a ( Curves k37 to k40). Therefore, the packet distribution method according to the present invention is an effective packet distribution method for aggregating bands and transmitting packets to a transmission destination with high throughput even in a wireless communication environment in which fading occurs.

  Based on the simulation results described above, the packet distribution method according to the present invention is an effective packet distribution that can transmit packets at a high throughput when transmitting packets to a destination using a plurality of wireless devices that perform wireless communication using different wireless communication methods. It proved to be a method.

In the above description, each of the cognitive base station 1 and the wireless devices 2 to 5 has been described as including the three wireless devices 11 to 13. However, the present invention is not limited thereto, and the cognitive base station 1 and Each of the wireless devices 2 to 5 may include n (n is an integer of 2 or more) wireless devices. Then, each of the cognitive base station 1 and the wireless devices 2 to 5 selects m wireless devices (m is an integer satisfying 1 ≦ m ≦ n) from the n wireless devices, and selects the selected wireless devices. The P ^q (or m = l ^a ) packets are distributed to the m radio devices by the method described above and transmitted to the transmission destination.

  Further, in the present invention, the distribution module 15 that distributes the packets to the wireless devices that can use the packets (= the wireless devices having the positive time resource RT among the wireless devices 11 to 13) according to the flowcharts shown in FIGS. , “Packet distribution means”.

Further, in the present invention, the queue 16 forms a “buffer”, and the distribution module 15 that calculates the estimated transmission required time T ^e (P _n , l _k ) according to the expressions (1) and (2) Means ".

  Furthermore, in the present invention, the distribution module 15 that determines the optimum combination of a packet and a radio so that the total estimated transmission time is minimized constitutes “determination means”.

  The distribution module 15 that distributes the packets to the wireless devices that can use the packet according to the optimal combination (= the wireless device having the positive time resource RT among the wireless devices 11 to 13) constitutes “allocation means”.

  Furthermore, in the present invention, the process of adding the constant value T to the time resources RT11 to RT13 when all of the time resources RT11 to RT13 of the radio devices 11 to 13 are negative constitutes a “reset process”.

  Further, according to the present invention, the estimated transmission time of the radio that has distributed the packet is subtracted from the current value of the time resource RT given to the available radio when the packet is distributed to the available radio. The process of updating the value of the time resource RT given to the available wireless device constitutes an “update process”.

  The embodiment disclosed this time should be considered as illustrative in all points and not restrictive. The scope of the present invention is shown not by the above description of the embodiment but by the scope of claims for patent, and is intended to include all modifications within the meaning and scope equivalent to the scope of claims for patent.

  The present invention is applied to a radio apparatus capable of distributing packets in consideration of utilization efficiency of a radio device. In addition, the present invention is applied to a wireless network including a wireless device capable of distributing packets in consideration of utilization efficiency of wireless devices.

1 is a schematic diagram of a wireless network according to an embodiment of the present invention. It is a schematic block diagram which shows the structure of the cognitive base station shown in FIG. It is a conceptual diagram of transmission required time. It is a figure for demonstrating the concept of a time resource. It is a figure which shows the 1st example of the packet distribution using a time resource. It is a figure which shows the 2nd example of the packet distribution using a time resource. It is a figure which shows the 3rd example of the packet distribution using a time resource. It is a flowchart for demonstrating the distribution operation | movement of a packet. It is a flowchart for demonstrating the detailed operation | movement of step S9 shown in FIG. It is a figure which shows the relationship between a throughput when a time resource is changed, and an average delay. It is a figure which shows the relationship between a throughput and the amount of input traffic. It is a figure which shows the relationship between the average data size of the packet hold | maintained at the queue of each radio | wireless machine, and input traffic amount. It is a figure which shows the relationship between an average delay and throughput when the amount of input traffic increases from 4 Mbps to 14 Mbps. It is a figure which shows the relationship between the throughput and the amount of input traffic when fading arises. It is a figure which shows the relationship between the average data size of the packet hold | maintained in the queue of each radio | wireless machine, and the amount of input traffic when fading has arisen. It is a figure which shows the relationship between average delay and throughput when the amount of input traffic increases from 4 Mbps to 14 Mbps when fading has arisen.

Explanation of symbols

  1 cognitive base station, 2-5 wireless device, 10 wireless network, 11-13 wireless device, 11A, 12A, 13A, 16 queue, 14 monitor module, 15 distribution module, 17 communication module, 30 cognitive gateway, 31, 41 wired Cable, 40 networks.

Claims (8)

N (n is an integer greater than or equal to 2) radios each transmitting assigned packets;
The m radios so that the total required packet transmission time is minimized in m (m is an integer satisfying 1 ≦ m ≦ n) of the n radios. A wireless device comprising: packet allocating means for allocating the packet to a device.   The required transmission time is a time length from a first timing at which one packet is assigned to one radio to a second timing at which an acknowledgment indicating that the one packet has been received is received from the transmission destination. The wireless device according to claim 1, wherein:   The packet allocating means obtains an actual transmission required time, which is an actual transmission required time in one radio, as an estimated transmission required time when the one radio tries to transmit the packet, and the estimated transmission thus obtained The radio apparatus according to claim 2, wherein the packet is allocated to the m radio devices so that a total required time is minimized. A buffer for holding packets;
The packet allocating means reads m packets from the buffer, and allocates the read m packets to the m radio devices so that the total estimated transmission time is minimized. The wireless device described.   The packet allocating unit provides each m radio devices with a transmittable time during which each radio apparatus can execute a packet transmission operation, and the radio transmission time is positive among the m radio apparatuses. The radio apparatus according to claim 4, wherein the m packets are allocated to a machine.   When one packet is assigned to one radio device, the packet assigning means requires the estimated transmission requirement in the one radio device from the transmission possible time before the one packet is assigned in the one radio device. An update process for subtracting time to update the transmittable time is executed for the m radio devices, and the m radio devices having the updated transmittable time are positive among the m radio devices. The wireless device according to claim 5, wherein:   The packet allocating unit performs a reset process of adding a constant value to the m transmittable times of the m radios when the transmittable time is negative in all of the m radios, and performs the reset process. The radio apparatus according to claim 5 or 6, wherein the m packets are allocated to the m radio devices using the m transmittable times after processing.   A wireless network comprising the wireless device according to any one of claims 1 to 7.

Images