JP2015171096A - Gateway device, packet transfer method and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To efficiently perform packet transfer by using a gateway device equipped with a plurality of CPUs.SOLUTION: A gateway device comprises: a packet transfer means which selects a CPU from among a plurality of CPUs in accordance with an IP address included in a packet and has the selected CPU perform a packet transfer process; and an address allocation means which allocates an IP address determined in accordance with operational situations of a plurality of CPUs to a user terminal.

Description

  The present invention relates to a gateway device, a packet transfer method, and a program, and more particularly, to a gateway device provided on a mobile network, a packet transfer method by the gateway device, and a program that operates on the gateway device.

  Patent Document 1 discloses an MME (Mobility Management Entity) that manages mobility of a terminal (UE: User Equipment) and a P-GW (Packet data network Gateway) that serves as a gateway to a PDN (Packet Data Network) that is a carrier network. And a mobile network including an S-GW (Serving Gateway) as a relay device. When the UE accesses the radio access system, authentication and data transfer path establishment are performed by the MME, P-GW, and S-GW.

International Publication No. 2009/025232

  The entire disclosure of the above patent document is incorporated herein by reference. The following analysis was made by the present inventors.

  In the mobile network system described above, in order to improve the transfer capability of the packet transfer unit of the P-GW (Packet Data Network Gateway), a method of configuring the packet transfer unit with a plurality of CPUs (Central Processing Units) can be considered. However, when packet transfer is performed using a plurality of CPUs, division loss may occur depending on the CPU allocation method, and the performance of the entire P-GW may not be sufficiently obtained. The same possibility exists when the S-GW packet transfer unit is configured by a plurality of CPUs.

  Therefore, it is a problem to efficiently perform packet transfer by a gateway device including a plurality of CPUs. An object of the present invention is to provide a gateway device, a packet transfer method, and a program that contribute to solving the above problem.

  The gateway device according to the first aspect of the present invention selects one of a plurality of CPUs (Central Processing Units) according to an IP (Internet Protocol) address included in the packet, and selects the selected CPU. Packet transfer means for performing the packet transfer process, and address assignment means for allocating IP addresses determined according to operating states of the plurality of CPUs to the user terminals.

  According to a second aspect of the present invention, there is provided a packet transfer method in which a gateway device having a plurality of CPUs (Central Processing Units) is selected from the plurality of CPUs according to an IP (Internet Protocol) address included in the packet. A step of selecting the CPU, a step of causing the selected CPU to perform the packet transfer process, and a step of assigning an IP address determined according to the operating status of the plurality of CPUs to the user terminal. .

  According to a third aspect of the present invention, there is provided a program for a computer provided in a gateway device having a plurality of CPUs (Central Processing Units) according to an IP (Internet Protocol) address included in a packet. A process of selecting one of the CPUs, a process of causing the selected CPU to perform the packet transfer process, and assigning an IP address determined according to the operating status of the plurality of CPUs to the user terminal Process. The program can also be provided as a program product recorded on a non-transitory computer-readable storage medium.

  According to the gateway device, the packet transfer method, and the program according to the present invention, packet transfer can be efficiently performed by the gateway device including a plurality of CPUs.

It is a block diagram which illustrates the composition of the gateway device concerning one embodiment. It is a block diagram which illustrates the composition of the mobile network system provided with the gateway device concerning a 1st embodiment. It is a block diagram which illustrates the composition of the packet transfer part of the gateway device concerning a 1st embodiment. It is a figure for demonstrating operation | movement of the call processing control part of the gateway apparatus which concerns on 1st Embodiment. It is a sequence diagram which illustrates the attach operation in the mobile network system provided with the gateway apparatus which concerns on 1st Embodiment. It is a sequence diagram which illustrates the detaching operation | movement in the mobile network system provided with the gateway apparatus which concerns on 1st Embodiment. It is a figure which shows the comparative example with respect to the operation example of the gateway apparatus which concerns on 1st Embodiment. It is a figure which shows the operation example of the gateway apparatus which concerns on 1st Embodiment.

  First, an outline of one embodiment will be described. Note that the reference numerals of the drawings attached to this summary are merely examples for facilitating understanding, and are not intended to limit the present invention to the illustrated embodiment.

  FIG. 1 is a block diagram illustrating the configuration of a gateway device according to an embodiment. Referring to FIG. 1, the gateway device 10 includes a packet transfer unit 12 and an address assignment unit 14. The packet transfer means 12 selects any one of a plurality of CPUs (Central Processing Units) (for example, CPU0 and CPU1) according to an IP (Internet Protocol) address included in the packet, and selects the selected CPU. To transfer the packet. The address assigning unit 14 assigns an IP address determined according to the operating status of a plurality of CPUs (CPU0, CPU1) to the user terminal.

  Here, the address assignment unit 14 determines the IP address so that the load is smoothed among the plurality of CPUs. For example, it is preferable that the address assignment unit 14 determines the IP address so that a CPU having a relatively low load is selected by the packet transfer unit 12.

  As an example, the packet transfer means 12 may select the mod (A, N) -th CPU among N (N is a natural number of 2 or more) CPUs, where A is the IP address of the packet. Then, when there are two CPUs (ie, N = 2), the packet transfer means 12 selects CPU0 if the IP address A is an even number, and selects CPU1 if the IP address A is an odd number. become.

  Further, the address assigning unit 14 mods when the load of the M-th (M = 0, 1,..., N−1) CPU among N (N is a natural number of 2 or more) CPUs is relatively low. An IP address A satisfying (A, N) = M may be assigned to the user terminal. When the number of CPUs is two (ie, N = 2), the address assignment unit 14 assigns an even IP address A to the user terminal if the load on the CPU 0 is relatively low, while the load on the CPU 1 is relatively low. For example, an odd IP address A is assigned to the user terminal.

  According to one embodiment, an IP address that smoothes the processing load while checking the processing load state of each CPU included in the packet transfer means 12 of the gateway device (for example, P-GW: Packet Data Network Gateway). Is assigned to the UE by the address assigning means 14 (for example, the call processing control unit of the P-GW) of the gateway device, the processing load of each CPU is smoothed, and as a result, the performance of the entire device is fully utilized (for example, It can be pulled out (to nearly 100%).

<Embodiment 1>
The gateway device according to the first embodiment will be described with reference to the drawings. Here, a case where the gateway device is a P-GW (Packet Data Network Gateway) provided in the mobile network system will be described.

  FIG. 2 is a block diagram illustrating the configuration of a mobile network system including the gateway device according to this embodiment. Referring to FIG. 2, the mobile network system includes a P-GW 60, an S-GW (Serving Gateway) 50, an eNodeB (evolved NodeB, hereinafter referred to as “eNB”) 30, a UE (User Equipment, terminal) 20, and the Internet. It has.

  The P-GW 60 includes a call processing control unit 64 that controls call processing and a packet transfer unit 62 that performs packet transfer processing. The call processing control unit 64 has a DHCP (Dynamic Host Configuration Protocol) server function, and pays out an IP (Internet Protocol) address to the UE 20 during call processing. The call processing control unit 64 selects the IP address to be paid out from the IP address pool held by itself.

  FIG. 3 is a block diagram illustrating a detailed configuration of the packet transfer unit 62 illustrated in FIG. Referring to FIG. 3, the packet transfer unit 62 includes a plurality of CPUs 0 and 1 and a load balance control unit (hereinafter referred to as “LB unit”) 66. Here, a case will be described in which the packet transfer unit 62 includes two CPUs. However, in the present invention, the number of CPUs is not limited to this. In FIG. 3, the left LB unit 66 distributes a packet from the UE 20 to the Internet to the CPU 0 and the CPU 1. On the other hand, the LB unit 66 on the right side in FIG. 3 distributes the packets from the Internet to the UE 20 to the CPU 0 and the CPU 1.

  By using two CPUs, CPU0 and CPU1, the packet transfer unit 62 preferably exhibits twice the processing capability as compared with the case of one CPU. The LB unit 66 refers to the header information of the input packet, determines whether to distribute to the CPU 0 or the CPU 1, and transmits the packet to the determined CPU. In the present embodiment, the LB unit 66 distributes the packet to the CPU 0 if the IP address of the UE is an even number, and distributes the packet to the CPU 1 if the IP address is an odd number. CPU0 and CPU1 each perform packet transfer processing (for example, packet reception, protocol termination, editing, analysis, transmission, etc.).

  FIG. 4 is a diagram for explaining the details of the operation of the call processing control unit 64 shown in FIG. Referring to FIG. 4, the call processing control unit 64 performs call processing control via the S-GW 50 and the eNB 30 when the UE 20 is activated, and sets a data plane path. At this time, the call processing control unit 64 pays out an IP address to the UE 20. The call processing control unit 64 refers to the load status of the CPU0 and CPU1 of the packet transfer unit 62 when paying out the IP address, and pays out an odd number of IP addresses when the load on the CPU0 is high. Conversely, when the load on the CPU 1 is high, the packet transfer unit 62 pays out even-numbered IP addresses.

  The call processing control unit 64 pays out the IP address to the UE 20 while confirming the load status of the CPU0 and CPU1 included in the packet transfer unit 62, thereby reducing the load state deviation between the CPU0 and CPU1. It is possible to draw out the maximum capabilities of CPU0 and CPU1.

  Next, an operation in which the UE 20 attaches to or detaches from the communication network will be described with reference to the drawings. The presence of UE 20 in the communication network due to physical movement or power-on is called attachment. On the other hand, the end of the communication network location due to physical movement or power off is called detachment.

  At the time of attachment, call processing is performed between the devices, a data plane path is set, and an IP address is assigned to the UE 20. Even at the time of detachment, call processing is performed between the devices, the data plane path is deleted, and the IP address assigned to the UE 20 is also deleted.

  FIG. 5 is a sequence diagram illustrating an operation at the time of attachment in the present embodiment. On the other hand, FIG. 6 is a sequence diagram illustrating an operation at the time of detachment in the present embodiment.

  Referring to FIG. 5, when UE 20 transitions to the attached state (S0), call processing is performed among UE 20, eNB 30, MME 40, S-GW 50, and P-GW 60 (S1 to S4). In the S-GW 50 and the P-GW 60, call information is set from the call processing control units 54 and 64 to the packet transfer units 52 and 62 (S5 and S6), respectively, and a data plane path is set. Further, an IP address is assigned to the UE from the call processing control unit of the P-GW (S7, S8 to S11).

  In this embodiment, when the call processing control unit 64 of the P-GW 60 selects an IP address to be allocated to the UE 20, the load status of the CPU 0 and CPU 1 provided in the packet transfer unit 62 of the P-GW 60 is confirmed, Select an IP address that allows packet forwarding to be performed on the lighter side. Specifically, the call processing control unit 64 assigns an even IP address if the load on the CPU 0 is light, and assigns an odd IP address if the load on the CPU 1 is light.

  When the call processing is completed, the UE 20 can communicate with the Internet via the eNB 30, the S-GW 50, and the P-GW 60. The LB unit 66 provided in the packet transfer unit 62 of the P-GW 60 refers to the IP address of the UE 20, and transfers it to the CPU 0 if it is an even number and transfers it to the CPU 1 if it is an odd number. Note that the LB unit 66 refers to the transmission source IP address for the packet from the UE 20 to the Internet, and refers to the destination IP address for the packet from the Internet to the UE 20.

  Since the IP address assigned to the UE 20 is determined to be even or odd according to the load state of the CPU0 and CPU1 included in the packet transfer unit 62 of the P-GW 60, the processing load of the CPU0 and CPU1 in the packet transfer unit 62 is It will be smoothed.

  Referring to FIG. 6, when the UE 20 transitions to the detached state (S20), call processing is performed between the UE 20, the eNB 30, the MME 40, the S-GW 50, and the P-GW 60 (S21 to S24). In the S-GW 50 and the P-GW 60, the call information is deleted from the call processing control units 54 and 64 to the packet transfer units 52 and 62, respectively (S25 and S26), and the data plane path is disconnected. In addition, the IP address assigned to the UE 20 is deleted by the call processing control unit 64 of the P-GW 60 (S27, S28 to S31).

  Next, the effect brought about by this embodiment will be described. The period from attach to detach differs for each UE. Therefore, when the packet transfer unit 62 is configured using a plurality of CPUs, variations in load occur between the CPUs. According to the present embodiment, the IP address assigned to the UE is determined so that the load between the CPUs is smoothed in consideration of the load on the CPU0 and the CPU1. As a result, it is possible to eliminate the uneven load between the CPUs and to exhibit high performance in the packet transfer apparatus (P-GW 60).

  When call information is set in the packet transfer unit 62 from the call processing control unit 64 of the P-GW 60, the same call information is set in the LB unit 66, and the LB unit 66 determines which CPU to transfer to. A method of making it possible is also conceivable. However, according to such a method, the configuration of the LB unit 66 becomes complicated, and the packet transfer capability of the packet transfer unit 62 may be reduced. On the other hand, according to the present embodiment, the LB unit 66 can have a simple configuration, and the load can be smoothed among a plurality of CPUs.

<Modification 1>
In the first embodiment, the packet transfer unit 62 of the P-GW 60 includes two CPUs, CPU0 and CPU1, and the call processing control unit 64 selects and pays out one of the two even or odd numbers as the IP address. It was supposed to be. On the other hand, the packet transfer unit 62 may include N (N ≧ 3) CPUs.

  Also in this case, the same effect as in the first embodiment can be obtained by selecting the IP address that the call processing control unit 64 pays out to the UE from among N types. As an example, when the load on the Mth (M = 0, 1,..., N−1) CPU among the N CPUs is relatively low, the call processing control unit 64 mod (A, N ) = IP address A satisfying M may be assigned to the UE 20. At this time, if the IP address of the packet is A, the LB unit 66 of the packet transfer unit 62 selects the mod (A, N) th CPU among the N CPUs. However, these operations of the call processing control unit 64 and the packet transfer unit 62 are merely examples, and the present invention is not limited to these operations.

<Modification 2>
In the first embodiment, the configuration having the DHCP server function in the call processing control unit 64 of the P-GW 60 has been described. However, a DHCP server may exist outside the P-GW 60.

  In this case, a new signal is defined between the P-GW 60 and the DHCP server, and the external DHCP server recognizes the load state of each CPU provided in the packet transfer unit 62 of the P-GW 60. Further, the DHCP server determines an IP address to be assigned to the UE 20 so as to smooth the load between the CPUs, for example, according to the notified load state of each CPU. Thereby, the effect similar to 1st Embodiment is acquired.

<Modification 3>
In the first embodiment, the case where the packet transfer unit 62 of the P-GW 60 includes a plurality of CPUs has been described. As a modification, the packet transfer unit 52 of the S-GW 50 may have a plurality of CPUs that process packets, like the packet transfer unit 62 of the P-GW 60.

  In this case, by defining a new signal between the S-GW 50 and the P-GW 60, an IP address that smoothes the load state of each CPU of the packet transfer unit 52 of the S-GW 50 is assigned to the P-GW 60. It is also possible to pay out from the side. According to such a modification, not only the P-GW 60 but also the S-GW 50 can efficiently perform packet transfer using a plurality of CPUs.

<Operation example>
Here, a specific operation example of the gateway device (P-GW 60) according to the first embodiment will be described. Here, eight UEs from UE1 to UE8, ddd. ddd. ddd. A case where eight IP addresses from 001 to 008 are used will be described. Here, as an example, it is assumed that each UE attaches every 2 minutes, UE1, 3, 5, and 7 each detach after 1 minute, and other UEs do not detach.

  The processing load of each CPU provided in the packet transfer unit 62 of the P-GW 60 will be described with reference to FIGS. FIG. 7 shows, as a comparative example, an operation when the call processing control unit 64 assigns IP addresses to UEs in order from the lowest number. In this case, finally, the processing load is biased only to the CPU 1.

  On the other hand, FIG. 8 shows an operation when the call processing control unit 64 of the P-GW 60 determines an IP address to be paid out according to the load state of the CPU 0 and the CPU 1. In this case, finally, two UEs are linked to each of CPU0 and CPU1, and the load is distributed. Thus, according to the first embodiment, load smoothing can be realized between CPUs.

In the present invention, the following modes are possible.
[Form 1]
As in the gateway device according to the first aspect.
[Form 2]
The address assignment means determines an IP address so that a load is smoothed among the plurality of CPUs;
The gateway device according to aspect 1.
[Form 3]
The address assigning means determines an IP address so that a CPU having a relatively low load is selected by the packet transfer means;
The gateway device according to mode 2.
[Form 4]
The packet transfer means selects a mod (A, N) -th CPU among N (N is a natural number of 2 or more) CPUs, where A is the IP address of the packet.
4. The gateway device according to mode 2 or 3.
[Form 5]
When the load on the M-th (M = 0, 1,..., N−1) CPU among N (N is a natural number greater than or equal to 2) CPUs is relatively low, the address allocating unit mod (A , N) = assign IP address A satisfying M to the user terminal,
5. The gateway device according to mode 4.
[Form 6]
The address allocating means notifies the operation status of the plurality of CPUs to an externally allocated address allocation device instead of allocating the IP address to the user terminal, and allocates the address according to the operation status of the plurality of CPUs. Assigning the IP address determined by the device to the user terminal;
The gateway apparatus as described in any one of form 1 thru | or 5.
[Form 7]
The gateway device is a PDN-GW (Packet Data Network Gateway) on a mobile network.
The gateway apparatus as described in any one of form 1 thru | or 6.
[Form 8]
The gateway device according to aspect 7,
S-GW (Serving Gateway) on the mobile network,
The S-GW selects a second CPU of the plurality of second CPUs according to the IP address of the packet, and causes the selected second CPU to perform the transfer process of the packet. Packet transfer means,
Notification means for notifying the gateway device of the operating status of the plurality of second CPUs,
The address assigning means assigns an IP address determined according to an operating status of the plurality of second CPUs to a user terminal;
A communication system characterized by the above.
[Form 9]
The packet transfer method according to the second aspect is as described above.
[Mode 10]
The gateway device determines an IP address so that a load is smoothed among the plurality of CPUs.
The packet transfer method according to mode 9.
[Form 11]
The gateway device determines an IP address so that a CPU with a relatively low load is selected.
The packet transfer method according to mode 10.
[Form 12]
The gateway device selects a mod (A, N) -th CPU among N (N is a natural number of 2 or more) CPUs, where A is an IP address of a packet.
The packet transfer method according to Form 10 or 11.
[Form 13]
When the load of the M-th (M = 0, 1,..., N−1) CPU among N (N is a natural number of 2 or more) CPUs is relatively low, the gateway device mod (A, N) = assign an IP address A satisfying M to the user terminal,
The packet transfer method according to mode 12.
[Form 14]
Instead of allocating the IP address to the user terminal, the gateway device notifies the operation status of the plurality of CPUs to an external address allocation device, and the address allocation device according to the operation status of the plurality of CPUs. Assigning to the user terminal an IP address determined by
The packet transfer method according to any one of forms 9 to 13.
[Form 15]
The program according to the third aspect is as described above.
[Form 16]
Causing the computer to execute a process of determining an IP address so that the load is smoothed among the plurality of CPUs.
The program according to form 15.
[Form 17]
Causing the computer to execute a process of determining an IP address so that a CPU with a relatively low load is selected.
The program according to mode 16.
[Form 18]
When the IP address of the packet is A, the computer is caused to execute a process of selecting a mod (A, N) -th CPU among N (N is a natural number of 2 or more) CPUs.
The program according to Form 16 or 17.
[Form 19]
When the load of the M-th (M = 0, 1,..., N−1) CPU among N (N is a natural number of 2 or more) CPUs is relatively low, the gateway device mod (A, N) causing the computer to execute a process of assigning an IP address A satisfying M to a user terminal.
The program according to Form 18.
[Form 20]
Instead of assigning an IP address to a user terminal itself, the operating status of the plurality of CPUs is notified to an address allocation device provided outside, and the IP determined by the address allocation device according to the operating status of the plurality of CPUs Causing the computer to execute a process of assigning an address to the user terminal;
The program according to any one of forms 15 to 19.

  It should be noted that the entire disclosure of the above patent document is incorporated herein by reference. Within the scope of the entire disclosure (including claims) of the present invention, the embodiment can be changed and adjusted based on the basic technical concept. Further, various combinations or selections of various disclosed elements (including each element of each claim, each element of each embodiment, each element of each drawing, etc.) are possible within the framework of the entire disclosure of the present invention. is there. That is, the present invention of course includes various variations and modifications that could be made by those skilled in the art according to the entire disclosure including the claims and the technical idea. In particular, with respect to the numerical ranges described in this document, any numerical value or small range included in the range should be construed as being specifically described even if there is no specific description.

DESCRIPTION OF SYMBOLS 10 Gateway apparatus 12 Packet transfer means 14 Address allocation means 20 UE
30 eNB
40 MME
50 S-GW
52 packet transfer unit 54 call processing control unit 60 P-GW
62 packet transfer unit 64 call processing control unit 66 LB unit

Claims (10)

A packet transfer unit that selects any one of a plurality of CPUs (Central Processing Units) according to an IP (Internet Protocol) address included in the packet, and causes the selected CPU to perform the packet transfer process; ,
Address allocation means for allocating IP addresses determined according to operating states of the plurality of CPUs to user terminals,
A gateway device characterized by that. The address assignment means determines an IP address so that a load is smoothed among the plurality of CPUs;
The gateway device according to claim 1. The address assigning means determines an IP address so that a CPU having a relatively low load is selected by the packet transfer means;
The gateway device according to claim 2. The packet transfer means selects a mod (A, N) -th CPU among N (N is a natural number of 2 or more) CPUs, where A is the IP address of the packet.
The gateway device according to claim 2 or 3. When the load on the M-th (M = 0, 1,..., N−1) CPU among N (N is a natural number greater than or equal to 2) CPUs is relatively low, the address allocating unit mod (A , N) = assign IP address A satisfying M to the user terminal,
The gateway device according to claim 4. The address allocating means notifies the operation status of the plurality of CPUs to an externally allocated address allocation device instead of allocating the IP address to the user terminal, and allocates the address according to the operation status of the plurality of CPUs. Assigning the IP address determined by the device to the user terminal;
The gateway apparatus of any one of Claim 1 thru | or 5. The gateway device is a PDN-GW (Packet Data Network Gateway) on a mobile network.
The gateway apparatus of any one of Claim 1 thru | or 6. A gateway device according to claim 7;
S-GW (Serving Gateway) on the mobile network,
The S-GW selects a second CPU of the plurality of second CPUs according to the IP address of the packet, and causes the selected second CPU to perform the transfer process of the packet. Packet transfer means,
Notification means for notifying the gateway device of the operating status of the plurality of second CPUs,
The address assigning means assigns an IP address determined according to an operating status of the plurality of second CPUs to a user terminal;
A communication system characterized by the above. A gateway device including a plurality of CPUs (Central Processing Units), selecting one of the plurality of CPUs according to an IP (Internet Protocol) address included in the packet;
A step of causing the selected CPU to transfer the packet;
Assigning an IP address determined according to the operating status of the plurality of CPUs to a user terminal,
And a packet transfer method. For a computer provided in a gateway device having a plurality of CPUs (Central Processing Units)
A process of selecting one of the plurality of CPUs according to an IP (Internet Protocol) address included in the packet;
A process for causing the selected CPU to transfer the packet;
A process of assigning an IP address determined according to the operating status of the plurality of CPUs to a user terminal,
A program characterized by that.

Images