JP6012080B2 - Communication system and handover method thereof - Google Patents

Description

  The present invention relates to control between devices via a network and a communication system and communication method for transferring a communication flow according to the control.

  The conventional network (NW: Network) has integrated routing and packet transfer. The switch determines the transfer destination by looking at the header of the input packet and transfers it to the next switch. . Such a transfer method has the feature that each network device can communicate autonomously, and the distributed control is the basis of the TCP / IP network that realizes a highly fault-tolerant network. is there. However, flexible route control overlooking the entire network cannot be performed, and there is a trouble that settings must be made for each network device in order to perform route control.

  In conventional network devices, transfer control is performed in a layer structure. In the L2 switch, each device performs transfer control based on the MAC address, router IP address, and firewall TCP / UDP port number.

  On the other hand, in the flow control used in OpenFlow [Non-Patent Document 1], the traffic is identified by looking at all the addresses and identifiers of each layer, and an action is performed. Actions include unicast, multicast, bandwidth control, discard, load balancing, failure recovery, and virtual port forwarding control. OpenFlow has a centralized architecture in which a dedicated controller controls the route and centrally controls packet forwarding on the switch side. The OpenFlow controller that performs route control optimizes the route for each flow so that the load is not concentrated, or dynamically changes the route so that communication can be recovered quickly in the event of a failure. Therefore, a “flow table” based on various control information such as a physical port and a layer 4 protocol in addition to the existing MAC / IP address is created and propagated to the switch.

  In the mobile core network system, a method for always managing the connection position of a mobile terminal that is constantly moving is known [Non-Patent Document 2].

  FIG. 14 shows a configuration diagram of a mobility management method using a communication system according to the prior art. A plurality of radio base stations (eNodeB) 702 are connected to a single mobility management control device (MME) 701, and a plurality of radio terminals are connected to each radio base station (eNodeB) 702. The mobility management control device (MME) 701 is connected to the subscriber information database device (HSS) 703. In addition, a radio base station (eNodeB) 702 and an S-GW (Serving Gateway) 704 are connected to perform actual U plane (User plane) processing.

  For the C plane (Control plane) processing of mobile networks, functions are realized by using inexpensive general-purpose server equipment, and control processing by scale-out technology, IF integration between functions in the virtualization server, function arrangement, and integration. Efficiency is being considered. In other words, C-plane processing is not only scaled out by operating on general-purpose hardware, but also the simple integration of existing functions, as well as integration between functions to eliminate the need for IF between devices Yes.

  Also, for the U-plane of mobile networks, GTP-U (GPRS (General Packet Radio Service) Tunneling Protocol for User Plane) processing is currently implemented with dedicated equipment / software, and is linked to GTP-C and GTP-U. Since a dedicated in-device IF is required for operation, it cannot be as simple as a C-plane. In other words, what is realized with the IF for each device / inside the device is being studied in the direction of linking with the virtualization server. However, if the vendor or the function to be controlled changes, the IF changes and control is impossible.

OpenFlow Configuration and Management Protocol OF-CONFIG 1.0 (3GPP TS 23.401 V12.0.0 (2013-03))

  The conventional network (NW: Network) is integrated with route control (C plane) and packet forwarding (U plane), and the switch determines the forwarding destination by looking at the header of the incoming packet. Thus, since the data is transferred to the next switch, flexible route control over the entire network cannot be performed. For this reason, there has been a problem that settings must be made for each network device in order to perform path control.

  Furthermore, in the IP network, a routing table has been formed by a protocol such as RIP / OSPF / IS-IS so far, and the connection path at the end-end by IP in each network is controlled, and a protocol such as EGP / BGP is used between IP networks. By controlling it, communication between end-to-end nodes is realized by IP address. However, performing software processing on these C planes on the transfer device leads to an increase in device functions and a decrease in performance.

  Also, NGW (Network) control devices such as SGW and PGW in mobile networks, edge routers in fixed networks, etc. are also integrated with C plane and U plane processing, and either performance becomes a bottleneck, This is a performance limitation.

  That is, as shown in FIG. 15, a conventional NW (Network) device 810 has a control function and U-plane packet processing functions 811, 812, and 813 integrated on dedicated hardware, and the transfer NW device 820 can also calculate a route. The route calculation function 821 is distributed and deployed so as to be distributed.

  As a result, C plane processing (path calculation, control function) and U plane (switching, packet transfer) are integratedly implemented in devices such as switches and routers that make up the transfer system NW, and servers that make up the control system NW. Therefore, the performance degradation of the entire system due to the bottleneck of either resource is a problem. In addition, since each device has different functions and there are an enormous number of dedicated devices, in addition to the cost of the device itself, an enormous cost is required to operate them. Furthermore, the C-plane processing resources (CPU, memory) of these devices can only operate protocols and functions unique to the device such as RIP / OSPF if it is a router and mobility management / GTP if it is a control server. Therefore, a decrease in the utilization efficiency of C plane resources in the entire NW becomes a problem.

  In addition, a fixed and mobile specific control (C plane) function, and a device that integrates dedicated C / U plane processing to perform U plane processing will be used, not only between individual networks but also between multiple networks. Improving the resource utilization efficiency of C plane and U plane is an issue.

  In view of the above problems, the object of the present invention is to completely separate the C plane and U plane processing in the network control function and transfer function, and from the control function realized on a general-purpose C plane processing resource. It is to provide a communication system and a communication method that can solve the conventional problems by determining necessary processes and target devices according to received control contents, updating a flow table, and setting a path of a U plane. .

In order to achieve the above object, the present invention implements each function related to the C plane in the fixed network and the mobile network as a virtualization server on the general-purpose server, and a plurality of functions related to the U plane in the fixed network and the mobile network. A fixed network composed of a transfer flow switch network having a general-purpose switch and a mobile network, and a controller that performs a linkage process between the function related to the C plane and the transfer network related to the U plane In the communication system, the function relating to the C plane includes S-GW / P-GW and MME in a mobile network, and the controller is configured between a base station and S-GW / P-GW in the mobile network. The GTP tunnel ID (TEID) in the GTP tunnel for identifying and controlling the communication path is converted into an L2 tag, and the end point address of the GTP tunnel ID Means for converting the MAC address of the request source base station and generating conversion data in which QCI is converted into frame priority, and the general-purpose switch of the transfer network includes an L2 tag, a destination / transmission source based on the conversion data Means for setting a flow control rule for an action for a forwarded packet using a MAC address as a matching rule, and the terminal moves from the first base station to the second base station in the mobile network and from the terminal to the second base When the handover execution command is sent to the station, the EPS bearer context is taken over from the first base station to the second base station, and the connection destination of the GTP tunnel is transferred from the first base station to the second base station. The second base station notifies the MME that the switch has been made, and the MME notifies the S-GW / P-GW of the notification and notifies the EPS. The A-context is updated, and the S-GW / P-GW that has received the notification transmits a GTP flow update request including EPS bearer context information to the controller, and the controller that has received the GTP flow update request Extracts the tag information assigned from the tunnel ID and IP address information, updates the IP address for the general-purpose switch of the forwarding network, sets the MAC address corresponding to the IP address, and deletes the old flow control information And requesting an update of the new matching rule and action, notifying the S-GW / P-GW that the update of the GTP flow is completed after the request is completed, and receiving the notification of the S-GW The GW / P-GW confirms that the path switching accompanying the handover is completed via the MME to the second A communication system characterized by being configured to notify a station is proposed.

In order to achieve the above object, the present invention implements each function relating to the C plane in the fixed network and the mobile network as a virtualization server on the general-purpose server, and also provides a function relating to the U plane in the fixed network and the mobile network. A handover method in a communication system comprising a common transfer network composed of a transfer flow switch network including a plurality of general-purpose switches, and including a controller that performs cooperation processing between the function related to the C plane and the transfer network related to the U plane The function relating to the C plane includes S-GW / P-GW and MME in the mobile network, and the controller identifies between the base station and S-GW / P-GW in the mobile network. The GTP tunnel ID (TEID) in the GTP tunnel for controlling the communication path is converted into an L2 tag, and the end point of the GTP tunnel ID Means for converting the address into the MAC address of the requesting base station and generating conversion data in which the QCI is converted into frame priority, and the general-purpose switch of the transfer network uses an L2 tag, a destination / Means for setting a flow control rule for an action on a forwarded packet using a source MAC address as a matching rule, and a terminal moves from a first base station to a second base station in the mobile network and from the terminal to the second When a handover execution command is sent to the first base station, the EPS bearer context is taken over from the first base station to the second base station, and a GTP tunnel connection is established from the first base station to the second base station. The second base station notifies the MME that the destination has been switched, and the MME notifies the S-GW / P-GW of the notification. The EPS bearer context is updated, the S-GW / P-GW that has received the notification transmits a GTP flow update request including EPS bearer context information to the controller, and the controller that has received the GTP flow update request , The tag information assigned from the information of the GTP tunnel ID and the IP address is extracted, the IP address is updated to the general-purpose switch of the forwarding network, and the MAC address corresponding to the IP address is set, and the old flow control information , Request the update of the new matching rule and action, notify the S-GW / P-GW that the update of the GTP flow is completed after the request is completed, and receive the notification The S-GW / P-GW confirms that the path switching accompanying the handover has been completed via the MME. A handover method characterized by notifying a second base station is proposed.

  According to the communication system and the communication system of the present invention, the C plane processing function is realized on the virtualization server with all the fixed network and mobile network control functions and the route calculation. By implementing with a general-purpose switch, the C-plane processing and U-plane processing are completely separated, and the general-purpose switch is controlled by the control function realized on the C-plane processing resources. Therefore, it is not necessary to make settings for each network device, and processing in the transfer network can be simplified to improve performance. Furthermore, since the virtualization server that implements the C-plane processing function and the transfer network that implements the U-plane processing function are completely separated, either resource does not become a bottleneck, and the overall system performance Improvement and resource utilization efficiency can be improved.

  Furthermore, according to the present invention, by providing a common transfer network and virtualization server for each of the fixed network and the mobile network, it is possible to improve the resource utilization efficiency of the C plane and the U plane among a plurality of networks.

The figure explaining the communication system in one Embodiment of this invention The figure which shows the communication system of 1st Example of this invention. The figure explaining the detail of the architecture which shared the control and transfer plane technique in the fixed network and mobile network of 2nd Example of this invention The figure explaining the prior art example corresponding to 2nd Example The figure explaining the detailed connection structure of the WiFi offload which shared the control and transfer plane technology in the fixed network and mobile network of 3rd Example of this invention The figure explaining the prior art example corresponding to Example 3 The figure explaining the network configuration method in 4th Example of this invention The figure explaining the network configuration method of the prior art example corresponding to 4th Example The figure explaining the mobile tunnel control of 5th Example of this invention The figure explaining the prior art example corresponding to 5th Example The figure explaining the flow of a process when a terminal implements a hand-over in the communication system of one Embodiment of this invention. The figure explaining the flow of a process when a terminal implements a hand-over in the communication system of a prior art example The figure explaining the detail of the command generation number between the controller and general purpose switch in one Embodiment of this invention, and a prior art example Configuration diagram for explaining a mobility management method in a communication system of a conventional example The figure explaining the subject in the communication system of a prior art example

  The present invention is an invention of a communication system and a communication system having an architecture that completely separates a C plane (route control) process and a U plane (packet transfer) process in a network control function and a transfer function.

  In other words, in the present invention, the C plane processing is realized on general-purpose server software by virtualizing server hardware. The U plane is realized by a low-function transfer flow switch network equivalent to SDN (Software-Defined Networking). In addition, the controller performs the linkage function between the C plane and the U plane, determines the processing required for the control content received from the control function and the target device, and updates the flow table, so that the NW (Network) While maintaining the same function as, it was possible to respond to new service additions.

  That is, in the communication system 1 of the present invention, as shown in FIG. 1, the C plane processing 2 and the U plane processing 3 are completely separated, and a plurality of flow switches 11 provided in the general purpose transfer hardware 10 are controlled via the controller 21. The architecture is adopted, and the C plane processing function 2 is realized on a virtual server using general-purpose hardware, and the control function and route calculation of the fixed network and the mobile network are all performed.

  As is well known, server virtualization is a technology that divides one physical server computer into multiple virtual computers, each running a separate OS and application software. is there. As a result, the processor, memory, and disk can be virtually divided into multiple areas, each acting as if it were one computer, and different OSs and applications can be executed simultaneously. Further, compared to a case where a plurality of computers are physically prepared, there is an advantage that it is possible to save time and effort for managing physical resources and to flexibly allocate resources according to demand.

  Accordingly, the service control function 22 and the network control function 23 operate on the common general-purpose server hardware 24 and 25, exchange necessary control information between the virtualization servers, and instruct the controller 21 to perform necessary U-plane processing. To realize services and functions.

  As a result, different functions between the fixed network and the mobile network can be realized independently on the virtualization server, and physical server resources and transfer resources can be shared, and necessary resources can be allocated and used as appropriate. Become. Therefore, it is possible to improve the resource utilization efficiency of the C plane / U plane between a plurality of networks.

  With the above architecture, C plane processing and U plane processing can be separated and resource allocation can be performed independently, so that either one will not become a bottleneck and the C / U plane can be scaled out independently. it can.

  In addition, the fixed network and the mobile network do not use dedicated devices, but use the common hardware to realize individual services only by the difference in control software. Resource efficiency can be realized.

  Next, a communication system according to a first embodiment of the present invention will be described with reference to FIG.

  As shown in FIG. 2, the C plane processing is realized by a general-purpose server, and the U plane is realized by a general-purpose flow switch. The server resource 4 in the figure is a general-purpose server group, and the transfer resource 5 is a general-purpose flow switch group that is distributed and arranged on the network. Further, the service control system resource 6 has a common fixed / moving transfer system.

  Further, in this embodiment, necessary resources are allocated from server resources common to the respective control functions without using dedicated hardware on the server resources (general-purpose server group).

  Rather than using dedicated hardware specialized for individual functions and services for both server resource 4 and transfer resource 5, software required for fixed / moving specific functions and services is operated on a general-purpose server. The transfer resource 5 also implements the routing / switching processing settings necessary for the fixed / moving service to the common general-purpose hardware via the controller 21, thereby enabling the fixed / mobile service. This can be realized while sharing the server resource 4 and the transfer resource 5.

  In other words, the C plane and server processing are converted to general-purpose servers and assigned to service / NW control resources. In addition, NW control functions such as mobility management, tunnel control, AAA, subscriber DB, etc. are centrally processed on the virtualization server, and functions common to fixed and mobile use the same control server, QoS, DHCP, NAT, Service control functions such as proxy and SIP (session control, etc.) can be shared by fixed / moving to improve server resource utilization efficiency.

  In addition, the transfer plane dynamically allocates transfer resources necessary for fixed / mobile services using a common technology, and transfers resources by sharing resources using a common transfer technology (common transfer network 70) for fixed / mobile services. Can be used effectively.

  As described above, in this embodiment, based on the above-described characteristics, the fixed service requires IMS (IP Multimedia Subsystem), tunnel control, AAA (Authentication, Authorization, Accounting)). The subscriber management is allocated to the fixed system service from the general-purpose server resource 4 as much as necessary, and the above functions are realized on a virtual server basis. Further, flow switching information for the control signal is set to the necessary transfer resource 5 via the controller 21 so that the control signal is appropriately routed to the virtualization server.

  When the user connects to the network and completes authentication with the virtual IMS or AAA server, the control server such as IMS or AAA requests flow control for appropriately routing the user packet to the controller 21 based on this result. To do.

  In response to this, by adding the flow control information for routing the PPP and VoIP packets necessary for the fixed service to the appropriate SIP server or other network via the controller 21 to the necessary transfer resource 5, Realize Internet access and voice services.

  Similarly, for the mobile service, a control server such as mobility management, tunnel control, AAA, subscriber DB, etc. necessary for the mobile service is secured from the server resource 4 and started as a virtual server on the general-purpose server. . Further, flow switching information for the control signal is set to the necessary transfer resource 5 via the controller 21 so that the control signal is appropriately routed to the virtualization server.

  In addition, when a mobile terminal connects, moves, and communicates with a mobile network, a control signal is transmitted to the server. Based on the result, user data is appropriately sent to the general-purpose switch via the controller 21 to the Internet or a partner terminal. It controls via the controller 21 so that it may reach.

  In response to this, the transfer resource 5 related to the controller 21 is secured, and necessary flow control information is input, whereby mobility management, handover, and bearer control can be realized.

  With the above configuration, independent scale-out is possible with a C / U plane separation type configuration via a common control IF for fixed and moving.

  In the conventional 3GPP architecture, there is a problem that the balance between the xGW C plane and U plane resources cannot be changed flexibly by integrating them.

  However, as described above, according to the present invention, since the C plane processing 2 is realized by a general-purpose server and the U plane 3 is realized by a general-purpose flow switch, both can be scaled out independently. In addition, the server resource 4 and the transfer resource 5 are virtualized and shared by the fixed / moving function, so that each fixed / moving NW (Network) subscriber needs a minimum amount according to demand in a single network. Limited resources can be flexibly allocated. Thereby, since the service can be provided, the resource utilization efficiency can be improved as compared with the conventional case.

  Next, a second embodiment of the present invention will be described with reference to FIG.

  In the second embodiment, details of an architecture in which a control / transfer plane technology in a fixed network and a mobile network is shared will be described.

  As shown in FIG. 3, in the second embodiment, since the fixed network 40 and the mobile network 50 have a common control PF (platform) 110 and transfer plane 120, the interwork function of the control signal protocol is realized by software. It can be realized on a common platform on hardware. That is, “edge control”, “line”, “conversion GW”, “HSS”, “MME”, “EPC control”, etc. are performed by the common control PF 110, and transfer of “edge”, “EPC” is performed by the common transfer plane 120. To do. In this way, by sharing the technology, each network's unique function / IF becomes unnecessary, and interworking can be realized by superposition / conversion between parameters, so that the linkage function can be suppressed to a small scale. This eliminates the need for a dedicated device and leads to cost reduction.

  Further, since the common transfer plane 120 covers the entire common transfer network 70, the U-plane protocol processing of the fixed network 40 and the mobile network 50 is not performed, and the minimum necessary processing that can be realized by the function of the general-purpose flow switch is started from the control PF 110. Instructed to transfer plane 120.

  By applying the present invention as in the above-described embodiment, the cooperation function of the fixed network 40 and the mobile network 50 can be closed to the software part of the C plane. Thus, since a general-purpose server and a general-purpose switch can be used, the cost can be reduced. For this reason, it is possible to realize an inexpensive cooperative service between the mobile network 40 and the fixed network 50. Furthermore, by sharing the technology, the unique function / IF of each network is not necessary, and the interwork can be realized by superimposing / converting parameters, so that the cooperation function can be reduced to a small scale.

  On the other hand, in the conventional WiFi offload architecture as shown in FIG. 4, protocol conversion of authentication / control signals and U-plane packets is required between the fixed network 40 and the mobile network 50. A conversion GW (gateway) corresponding to both NW (Network) protocols of the mobile network 50 is required. That is, it is necessary to have a GW that supports all the functions of the fixed network and the mobile network, and interworkes (protocol conversion and session state management cooperation). For this reason, the conversion GW apparatus becomes complicated, and many functions are required on high-performance hardware, and dedicated hardware tends to be expensive.

  As described above, the fixed network 40 and the mobile network 50 are conventionally configured using a technology that satisfies the requirements of each network. Therefore, in the case of cooperation between fixed and mobile, such as WiFi offload, both the mobile and fixed networks are used. There is a problem that it is necessary to deploy a GW corresponding to the functional requirements of the above, and the intermediate GW device tends to be highly functional.

  However, as described above, according to the present embodiment, by using the technology common to both networks and constructing the functions necessary for each network therein, the cooperation function can be realized more easily. The function can realize the WiFi offload function.

  Next, a third embodiment of the present invention will be described.

  In the third embodiment, a detailed connection configuration of WiFi offload in which control and transfer plane technologies in a fixed network and a mobile network are shared will be described.

  As shown in FIG. 5, in the third embodiment, the EAP-SIM / AKA authentication process and the mobile authentication server connection required by SaMoG are all realized on the common control platform on the general-purpose server 61.

  Each function is implemented on a software basis, and by deploying a function for performing cooperative control as SaMoG, authentication and address assignment of the terminal 82 connected to WiFi and a communication path to the service are set. All this processing is performed by software processing on the general-purpose server 61, and only the switching part to the PDN (public data network) 62 after connection is completed is connected to the general-purpose flow switch group (general-purpose switch network (common By instructing the transfer network 70), connection to the mobile network service via WiFi is realized.

  As described above, when realized by the NW (Network) configuration method of the present embodiment, a function equivalent to PPP / PPPoE / GTP is achieved by processing of the L2 layer in which a plurality of L2 switches (L2SW) 41 configuring the fixed network 40 are provided. While maintaining, the U-plane function can be realized at low cost by using a general-purpose switch. Furthermore, high function processing (C plane) can eliminate state management in the common transfer network 70 (fixed network 40, mobile network 50) by eliminating inter-device IF and performing centralized processing in the virtualization server. it can.

  On the other hand, in the conventional SaMoG architecture, as shown in FIG. 6, authentication and tunnel control functions are distributed among a plurality of devices, and state management control is performed in cooperation between the devices.

  That is, in the conventional mobile fixed cooperation architecture, in order to connect to the mobile network 50 from the fixed network 40 or the like by WiFi, the EAP-SIM / AKA protocol termination and the authentication server connection of the mobile network 50, address assignment by DHCP, movement by GTP Connection to the network and cooperative operation of the above three functions are required in the apparatus, and it is necessary to implement many functions and processes with the conversion GW function.

  Next, a fourth embodiment of the present invention will be described.

  In the fourth example, an NW (Network) configuration method in the above embodiment to which the present invention is applied will be described.

  As shown in FIG. 7, in the fourth embodiment, the mobility control replaces the GTP function with an equivalent function of the transfer layer to realize a handover or the like. For example, tag-based switching. Fixed control configures a VLAN for each control surface and ISP (Internet Service Provider).

  That is, in this embodiment, functions equivalent to PPPoE and GTP tunnels are realized by flow control (VLAN tag control, etc.) on a general-purpose flow switch network 70 (common transfer network) in which a plurality of general-purpose switches 71 (for example, VLAN Switch) are arranged. . This simplifies PPPoE and GTP protocol header processing and reduces the U plane processing load of the GW apparatus (BRAS, S / P-GW).

  In this embodiment, the fixed terminal 81 and the mobile terminal 82 first communicate with an authentication server (LAC-RADIUS, MME / HSS) on the general-purpose server 61 to perform authentication processing. At this time, flow control information (such as VLAN tag control information) is input in advance so that the authentication signal is routed to an appropriate server in the general-purpose switch network (common transfer network 70).

  Next, the distribution control server in the general-purpose server 61, the GTP-C processing of the MME / S / P-GW determines the flow control necessary for connecting the terminal and an appropriate service network (ISP etc.), and the controller (Not shown). For this purpose, a VLAN tag or the like can be assigned and used to control a path for connecting to a service for each terminal.

  In this way, the C plane and U plane are separated, the C plane is centrally controlled, and the U plane processing is processed in common, so that handover is realized by NW (Network) control. In addition, there is a possibility of improving the throughput.

  In contrast, in the conventional fixed network and mobile network Internet connection and mobile network mobility control processing, the mobile terminal control creates a GTP tunnel between S / P-GW-eNodeB and adjusts to the movement of the terminal. The EPS bearer context is updated to perform handover, and the fixed terminal control is combined with PPPoE, L2TP, and PPP to achieve NTE allocation.

  That is, as shown in FIG. 8, in the fixed network 40, the PPPoE tunnel connection 45 is made from each home to the BRAS device 42 for the Internet connection of the fixed terminal 81, and the BRAS device 42 operates in cooperation with the LAC-RADIUS 43 to the contractor. The contract ISP 63 is determined. An Internet access is realized by connecting to the NTE 44 to which the ISP 63 is connected by using an L2TP 46 to construct an L2 route, and making a PPP connection directly to the ISP 63 to which each home has contracted.

  In order to connect the mobile terminal 82 to the Internet, a service network (ISP) 62 called a PDN is connected to the P-GW 51 in the mobile network 50. The mobile terminal 82 and the PGW 51 are connected by the GTP tunnel 52, and handover is realized by appropriately switching the tunnel communication path by transmitting and receiving signals between the MME and the S / P-GW according to the movement of the terminal. At this time, information for recognizing information such as GTP tunnels used by each mobile terminal 82 on the C plane is called an EPS bearer context 53, which is managed by the MME and S / P-GW and connected to the tunnel. Is managing.

  Next, a fifth embodiment of the present invention will be described.

  In the fifth example, mobile tunnel control (handover, etc.) in the above embodiment to which the present invention is applied will be described.

  Specifically, with reference to FIG. 9, an application method when the present invention is applied and mobility control (handover or the like) is realized on a general-purpose server or a general-purpose switch network will be described.

  A controller that controls the general-purpose switch 71 of the general-purpose flow switch network (common transfer network) 70 by deploying the GTP control portions of the MME 64 and S / P-GW 65 and 66 as software on the general-purpose server 61, exchanging control information with them. 21 is also deployed.

  In order to connect the mobile terminal 82 to a service network such as an ISP network, an EPS bearer context 67 is generated on the MME, S / P-GW (C) server, and the EPS bearer context 67 is converted into general-purpose bearer information 68. A communication path is set on the general-purpose flow switch network 70 by instructing the controller 21 based on the information. The general-purpose bearer information 68 at this time includes “bearer ID”, “L2 tag”, “radio bearer ID”, “terminal IP address”, “S1 endpoint MAC”, “frame priority”, and the like.

  The bearer controls the L2 tag information as a key on the general-purpose switch, and transfers the packet toward each mobile terminal 82 based on the L2 tag. That is, the control instruction to the controller 21 is performed mainly using “L2 tag”, “S1 end point MAC”, and “frame priority” of the general-purpose bearer information 68.

  When the mobile terminal 82 moves to another eNodeB, the destination eNodeB recognizes the movement of the terminal, updates the IP address of the S1 endpoint, that is, the IP address of the destination eNodeB, between the MME and S / P-GW, The controller 21 is requested to update the flow control. In response to this, session information conversion processing is performed, the MAC address corresponding to the IP address is specified, and the controller 21 updates the flow table for the switch group related to before and after the movement via a general-purpose IF such as Openflow carry out.

  In the fifth embodiment, the EPS bearer context 67 is converted into general-purpose bearer information 68 on the controller 21 and instructed to the general-purpose flow switch network (common transfer network) 70, so that the function of the existing technology GTP is the basic of the general-purpose switch. The signal transmitted from the controller 21 can be suppressed by substituting it with a typical function.

  On the other hand, as an example of the prior art, as shown in FIG. 10, a method (existing research (Ericsson paper)) in which a GTP-U function (GTP extension 209) is mounted on a general-purpose switch 207 and bearer control is known. It has been.

  In the technology described in the above Ericsson paper, the GTP-U packet is controlled to a group of switches (207, 222) extended via the controller 206 based on the MME 202 and S / P-GW (C) 203, 204 on the general-purpose server 201. Instructed to do so through a dedicated control IF. Therefore, the EPS bearer context 210 on the control server side can use information common to the current EPC, and can identify and control the communication path of the mobile terminal 82 based on the GTP tunnel ID (TEID). Information of the EPS bearer context 205 includes “bearer ID”, “GTP tunnel ID”, “radio bearer ID”, “terminal IP address”, “IP address of S1 endpoint”, “QCI”, and the like.

  However, the above method cannot be realized with only the basic functions of the general-purpose switch 207, but requires an extended function 222 and an extended switch for a mobile unit. The problem is that the cost increases.

  Next, the flow of processing when a terminal performs handover in the communication system of this embodiment will be described with reference to FIG. Here, the flow of processing when the terminal performs handover is shown. Moreover, eNodeB (evolved NodeB: base station) is referred to as eNB.

  A UE (User Equipment: terminal) 301 is connected to a Source eNB (302) whose MAC address is MAC1, and its GTP tunnel ID is TEID1 and the end point address of TEID1 is IP1. In the controller (307), TEID1 is converted to L2 tag 1, IP1 is converted to the MAC address (MAC1) of Source eNB (302), and QCI is converted to frame priority.

  Based on this conversion data, tag 1 is assigned to the general-purpose switch group (general-purpose switch 305 in the figure) through which the flow passes, and the action (addition / deletion of tag, It is assumed that a flow control rule such as (to which physical port) is set.

  When UE (301) moves and connects to Target eNB (303), a handover execution command is sent from UE (301) to Target eNB (303) (S401), and Source eNB (302) to Target eNB (303) The EPS bearer context is taken over (S402).

  Thereafter, uplink data traffic from the UE (301) passes through the general-purpose switch network (305) via the Target eNB (303) and is sent to the Internet or the like (S403, S404).

  On the other hand, downlink traffic to the UE (301) requires bearer switching. Therefore, the Target eNB (303) notifies the MME (Mobility Management Entity) (304) that the TEID1 GTP tunnel has been switched to the IP2 eNB (S405), and the MME (304) sends it to the S / P-GW ( C) Notify (306) and update the EPS bearer context of each control server (S406).

  When this is completed, a GTP flow update request including the relevant EPS bearer context information is transmitted to the controller (307) (S407). The controller (307) extracts the assigned tag information from the TEID and IP address information, updates the IP address for the general-purpose switch group (305), and sets the MAC address corresponding to the IP address. The flow control information is deleted, and a new matching rule and action update is requested (S408).

  When this is completed, the controller (307) notifies the P-GW (C) (306) that the update of the GTP flow has been completed (S409), and that the path switching accompanying the handover has been completed via the MME (304). The target eNB (303) is notified (S410).

  When the above is completed, the downlink traffic route is switched and the handover procedure is completed.

  Next, the flow of processing when a terminal performs handover in a conventional communication system will be described in detail with reference to FIG.

  The UE (311) is connected to the Source eNB (312) whose address is IP1, its GTP tunnel ID is TEID1, and the end point address of TEID1 is IP1. The controller (317) sends GTP header information corresponding to TEID1 to the general-purpose switch (315), and adds a GTP header using a virtual port. It is assumed that a physical port for transferring a packet to the destination eNB is set.

  When UE (311) moves and connects to Target eNB (313), a handover execution command is sent from UE (311) to Target eNB (313) (S501), and Source eNB (312) to Target eNB (313) The EPS bearer context is taken over (S502). Thereafter, uplink data traffic from the UE (311) passes through the general-purpose switch network (315) via the Target eNB (313) and is sent to the Internet or the like (S503, S504).

  On the other hand, downlink traffic to the UE (311) requires bearer switching. Therefore, the Target eNB (313) notifies the MME (314) that the GTP tunnel of TEID1 has been switched to the eNB of IP2 (S505), and the MME (314) notifies it to S / P-GW (C) (316) To update the EPS bearer context of each control server (S506).

  When this is completed, a GTP flow update request including the relevant EPS bearer context information is transmitted to the controller (317) (S507). From the specified EPS bearer information, the controller (317) updates / adds / deletes the flow table and GTP encapsulation table in the related general purpose switch (315) (S508, S509).

  When this is completed, the P-GW (C) (316) is notified of the completion of GTP flow control update (S510), the path switching completion is sent to the Target eNB (313) via the MME (314) (S511), The switch is complete.

  Thus, the handover process is completed.

  Next, with reference to FIG. 13, the number of commands generated between the controller 307 and the general-purpose switch 305 in the communication system according to the present embodiment described with reference to FIG. 11, and the description with reference to FIG. Details of the number of commands generated between the controller 317 and the general-purpose switch 315 in the conventional communication system will be described.

  In FIG. 13, the vertical axis represents the number of commands generated per hour. Each of (a) to (l) represents as follows. (a): Attach, (b): Detach, (c): Location registration (X-GW not changed), (d): Location registration (X-GW changed), (e): Handover (X-GW changed) None), (f): Handover (with X-GW change), (g): PDN connection (added), (h): PDN connection (deleted), (i): QoS setting for PCRF activation, (j): Radio resource release by network activation, (k): Radio resource acquisition by terminal activation, (l): Radio resource acquisition by network activation.

  In the communication system of the conventional example, the number of commands per hour of (a) to (l) is (a): 0.2, (b): 0.2, (c): 8, (d ): 0.02, (e): 20, (f): 0.2, (g): 0.2, (h): 0.2, (i): 0.2, (j): 75, (k): 20, (l): 10.

  On the other hand, in the communication system of this embodiment, the number of commands per hour of (a) to (l) is (a): 0.1, (b): 0.1, (c ): 4, (d): 0.01, (e): 10, (f): 0.1, (g): 0.1, (h): 0.1, (i): 0.1, (j): 60, (k): 10, and (l): 5.

  As described above, according to the present embodiment, the number of commands generated per hour between the controller 307 and the general-purpose switch 305 is reduced by about 22% compared to the conventional example.

  As described above, according to the communication system of the present embodiment, the C plane and U plane processing in the network control function and the transfer function is completely separated, and the control content received from the control function of the C plane processing is used. Thus, since necessary processes and target devices are determined, the flow table is updated, and the path of the U plane is set, the conventional problem can be solved.

  The present invention relates to a communication system and a communication method for transferring a communication flow in accordance with control between devices via a network (transfer network), and a C-plane process and a U-plane process. Therefore, it is not necessary to make settings for each network device to perform path control, and the processing in the transfer network can be simplified to improve performance. Furthermore, since the virtualization server that implements the C-plane processing function and the transfer network that implements the U-plane processing function are completely separated, either resource does not become a bottleneck, and the overall system performance Improvement and resource utilization efficiency can be improved.

  DESCRIPTION OF SYMBOLS 1 ... Communication system, 2 ... C plane process, 3 ... U plane process, 4 ... Server resource, 5 ... Transfer resource, 6 ... Service control system resource, 10 ... General purpose transfer hardware, 11 ... Flow switch, 21 ... Controller, 22 ... Service control function, 23 ... Network control function, 24,25 ... General-purpose server hardware, 110 ... Control PF, 120 ... Transfer plane, 40 ... Fixed network, 41 ... L2 switch, 42 ... BRAS device, 43 ... LAC-RADIUS 44 ... NTE, 46 ... L2TP, 50 ... mobile network, 52 ... GTP tunnel, 53 ... EPS bearer context, 61 ... general server, 62 ... PDN, 63 ... ISP, 64 ... MME, 65 ... S-GE, 66 ... P-GW, 67 ... EPS bearer context, 68 ... General purpose bearer information, 70 ... Common transfer network, 71 ... General purpose switch, 81 ... Fixed terminal, 82 ... Mobile terminal, 301 ... UE (terminal), 302 ... Source eNB, 303 ... Target eNB, 304 ... MME, 305 ... General-purpose switch, 306 ... S / P-GW (C), 307 ... Controller.

Claims (7)

Each function related to the C plane in the fixed network and the mobile network is implemented as a virtualization server on the general-purpose server, and the function related to the U-plane of the fixed network and the mobile network is composed of a transfer flow switch network including a plurality of general-purpose switches. A communication system including a controller configured by a common transfer network in a fixed network and a mobile network, and performing a linkage process between the function related to the C plane and the transfer network related to the U plane,
The functions related to the C plane include S-GW / P-GW and MME in the mobile network,
The controller converts a GTP tunnel ID (TEID) in a GTP tunnel for identifying and controlling a communication path between a base station and an S-GW / P-GW in a mobile network into an L2 tag, and the GTP tunnel Means for converting the address of the end point of the ID into the MAC address of the requesting base station, and generating conversion data in which the QCI is converted into frame priority;
The general-purpose switch of the transfer network includes means for setting a flow control rule for an action on a transfer packet based on the conversion data, using an L2 tag and a destination / source MAC address as a matching rule.
When the terminal moves from the first base station to the second base station in the mobile network and a handover execution command is sent from the terminal to the second base station,
Take over the EPS bearer context from the first base station to the second base station,
The second base station notifies the MME that the connection destination of the GTP tunnel has been switched from the first base station to the second base station;
The MME notifies the S-GW / P-GW of the notification to update the EPS bearer context,
The S-GW / P-GW that has received the notification sends a GTP flow update request including EPS bearer context information to the controller,
The controller that has received the GTP flow update request extracts the tag information assigned from the information of the GTP tunnel ID and the IP address, updates the IP address to the general-purpose switch of the transfer network, and corresponds to the IP address. Set MAC address, delete old flow control information, request new matching rule and action update, update GTP flow to S-GW / P-GW after the request is completed Notice that
The S-GW / P-GW that has received the notification is configured to notify the second base station via the MME that the path switching accompanying the handover has been completed. . Each function related to the C plane was built on a common control platform.
The communication system according to claim 1. EAP-SIM / AKA authentication processing and mobile authentication server connection required by SaMoG were built on the common control platform.
The communication system according to claim 2. In the transport network, the function equivalent to PPPoE and GTP tunnel is configured by flow control that controls the path for connecting the terminal to the service network.
The communication system according to any one of claims 1 to 3. While using a VLAN switch as a general-purpose switch of the transfer network,
The controller controls the VLAN switch by VLAN tag so that it is routed properly
The communication system according to any one of claims 1 to 4, characterized in that: The action for the transfer packet includes tag addition / deletion and an instruction of a packet transfer destination physical port.
The communication system according to any one of claims 1 to 5, wherein: Each function related to the C plane in the fixed network and the mobile network is implemented as a virtualization server on the general-purpose server, and the function related to the U-plane of the fixed network and the mobile network is composed of a transfer flow switch network including a plurality of general-purpose switches. A handover method in a communication system including a controller configured by a common transfer network and performing a linkage process between the function related to the C plane and the transfer network related to the U plane,
The functions related to the C plane include S-GW / P-GW and MME in the mobile network,
The controller converts a GTP tunnel ID (TEID) in a GTP tunnel for identifying and controlling a communication path between a base station and an S-GW / P-GW in a mobile network into an L2 tag, and the GTP tunnel Means for converting the address of the end point of the ID into the MAC address of the requesting base station, and generating conversion data in which the QCI is converted into frame priority;
The general-purpose switch of the transfer network includes means for setting a flow control rule for an action on a transfer packet based on the conversion data, using an L2 tag and a destination / source MAC address as a matching rule.
When the terminal moves from the first base station to the second base station in the mobile network and a handover execution command is sent from the terminal to the second base station,
Take over the EPS bearer context from the first base station to the second base station,
The second base station notifies the MME that the connection destination of the GTP tunnel has been switched from the first base station to the second base station;
The MME notifies the S-GW / P-GW of the notification to update the EPS bearer context,
The S-GW / P-GW that has received the notification sends a GTP flow update request including EPS bearer context information to the controller,
The controller that has received the GTP flow update request extracts the tag information assigned from the information of the GTP tunnel ID and the IP address, updates the IP address to the general-purpose switch of the transfer network, and corresponds to the IP address. Set MAC address, delete old flow control information, request new matching rule and action update, update GTP flow to S-GW / P-GW after the request is completed Notice that
The S-GW / P-GW that has received the notification notifies the second base station via the MME that the path switching accompanying the handover has been completed.
A handover method in a communication system.

Images