CN1582051A - Method for realizing network layer switching in mobile IP - Google Patents

Abstract

The invention discloses a method for realizing network layer switching in mobile IP. The key point is to establish a dual link for communication between a mobile node and its current base station. The method includes the following steps: a. After handover, disconnect one of the established dual links, and establish a new link connection between the current mobile node and the target base station to be handed over to; b. The mobile node registers with the home agent through the newly created link , and perform a network layer handover; c. After the registration is completed, disconnect another link connection between the current mobile node and the original base station, and establish a link connection between the current mobile node and the target base station to be handed over to. Adopting the method can ensure that the communication is not interrupted during the switching process of the mobile node, thereby increasing the switching speed and reducing the packet loss rate during the switching process.

Description

The Method of Realizing Network Layer Handover in Mobile IP

technical field

The invention relates to mobile IP switching technology, in particular to a method for realizing network layer switching in mobile IP.

Background technique

With the rapid popularization of the Internet, the extensive use of portable devices, the rapid development of wireless communication devices, and the increase in people's dependence on the network, mobile users hope to access the corporate network anytime, anywhere in a more flexible way. Resources and Internet resources, that is to say, the user has a demand for the mobility of the host, hoping that the host does not need to interrupt the existing communication connection when changing its location. For this reason, the Internet Engineering Task Force (IETF) has defined Mobile IP (Mobile IP). Mobile IP is a network layer solution that supports host mobility. On the basis of hardware, it provides a location tracking and IP data message forwarding mechanism. When a host moves, the IP data message sent to the host can be safely forwarded to the current location of the host.

Mobile IP is a scheme that provides mobile functions on the global Internet, so that mobile nodes (MN) can still maintain ongoing communication when switching links. The IP routing mechanism provided by mobile IP enables mobile nodes to connect to any link with a permanent IP address.

The introduction of mobile IP technology enables mobile users to continue to use all resources in the original network without modifying the original IP address of the computer during cross-network movement and roaming. It shields users from changes in IP access networks during movement the impact. Simply put, Mobile IP is to realize the smooth movement and roaming functions of mobile users in the network.

Fig. 1 is the basic network topological structure figure that each entity is formed in the mobile IP, and the mobile node 10 shown in the figure refers to a mobile computer or router, also can be referred to as mobile host (MH), this mobile node 10 still remains after moving. The original IP address can be used for communication; the correspondent (CN) 11 shown in the figure is a computer communicating with the mobile node 10, and can also be called a correspondent host (CH). 1, for the mobile node 10, the home network of the mobile node 10 is called a home network (HN), and the network where the mobile node 10 is in a roaming state is called a foreign network (FN). foreign network. The permanent IP address assigned to mobile node 10 is called home address (HomeAddress), and this address does not change with the change of node position; In the home network, the link of mobile node 10 home network is called home link (HL); The router on the home network of the mobile node 10 is called a home agent (HA, HomeAgent) 12, which is responsible for saving the current location information of the mobile node, and is also responsible for intercepting the message sent to the mobile node, and forwarding the message to the The current location of the mobile node. In the foreign network, the link of the foreign network where the mobile node 10 is located is called a foreign link (FL); the router running on the foreign network is called a foreign agent (FA), which is the agent for the mobile node to send messages in the foreign network. The default router also provides tunnel decapsulation service sometimes, and Fig. 1 includes foreign agent 13 and foreign agent 14 belonging to foreign network 1 and foreign network 2 respectively.

Still taking Figure 1 as an example, when the mobile node 10 roams from the home network to the foreign network 1 or from the foreign network 1 to the foreign network 2, the mobile node 10 will undergo network handover, and the usual handover process is: the mobile node 10 according to After the currently detected information finds that it has entered a new network, the mobile node 10 will obtain a care-of address by receiving the agent broadcast message sent by the foreign agent, or by means of Dynamic Host Configuration Protocol (DHCP), and identify the mobile node with this care-of address 10 Current location. The care-of address may be a foreign agent care-of address, that is, the IP address of the foreign agent; it may also be a configured care-of address, that is, an IP address allocated on the foreign network. After obtaining the care-of address, the mobile node 10 registers with its home agent through the routing and forwarding function provided by the foreign agent, and notifies the home agent of its current care-of address. That is to say, whenever the location of the mobile node changes, it must re-register with the home agent and update its location information, so that the home agent can know the current location information of the mobile node at any time.

Since Mobile IP is a solution at the network layer, the above-mentioned handover process is all carried out at the network layer, but in fact, the handover between different networks during the roaming process of the mobile node mainly occurs at two levels: one is the lower layer The handover on the physical network includes the processing on the physical layer and the link layer; the other is a series of operations such as registration and reconfiguration of routes on the network layer according to the Mobile IP protocol. In the standard mobile IP protocol, only when the mobile node has completed the handover at the lower layer, that is, after disconnecting from the original network, does the handover at the network layer begin, which will interrupt the communication and cause packet loss. Moreover, if the communication interruption time is too long, it will affect applications with high real-time requirements, making roaming opaque to upper-layer applications and affecting the use effect of mobile IP.

In order to solve the impact of handover delay on applications, it is necessary to realize seamless handover. The so-called seamless switching includes two aspects: fast switching and smooth switching. Among them, the purpose of fast handover is to reduce or eliminate the delay in establishing a new communication path for the mobile node; the purpose of smooth handover is to reduce the loss rate of data packets.

At present, the industry has proposed many different solutions for fast switching and smooth switching:

In solving the fast switching problem, the main idea is to distribute the functions of the home agent, implement local registration, and reduce the registration time. It mainly includes two types of methods: hierarchical mobility management and mobility management based on specific host routing. Among them, the hierarchical mobility management methods include: MIP-RR, IDMP and other solutions; the mobility management methods based on specific host routing include: cellular IP, HAWAII and other solutions. In the above methods, the wireless access network is divided into individual wireless domains, and localization of registration is realized in the wireless domains. The difference between the two methods lies in: the protocol for packet transmission in the wireless domain, the former method uses the conventional mobile IP protocol, and the latter method uses the protocol based on specific host routing.

In terms of smooth switching, a packet buffer forwarding mechanism is required, which is divided into unicast forwarding and multicast forwarding. Unicast forwarding means: the current base station agent, that is, the foreign agent, caches packet data packets during the handover process and forwards them to the new base station agent after handover, but the packet forwarding between the old and new base station agents may take a long time; Multicast forwarding uses the multicast protocol to forward the cached packet data packets. For example, each mobile node has a multicast address. When the mobile node registers with the new base station agent, the new base station agent requires the adjacent base station agent to join the multicast group. , to pre-cache packets, but this solution has problems of multicast address conflicts and neighbors needing to be manually configured. If a domain foreign agent (DFA) is used to assign multicast addresses to mobile nodes to solve the problem of multicast address conflicts, Lack of intelligent neighbor discovery mechanism and overhead of multicast state.

In addition, in the above schemes, whether it is to achieve fast switching or smooth switching, it is a solution limited to the network layer. Although it is trying to shorten the delay time of network layer switching, in fact, the time of network layer switching is only milliseconds. , and the communication interruption process is at the second level, if simply shortening the switching time of the network layer, the impact on improving the overall switching speed is not obvious. Therefore, some research institutes propose to combine the link layer and the network layer, and use the link layer to trigger the network layer to perform pre-registration, so that the layer-3 handover is also completed when the layer-2 handover is completed, thereby realizing fast handover. This kind of scheme mainly includes: An Internet draft proposed by Ericsson - "Fast Switching in Mobile IPv4". The main idea of this scheme is: when the mobile node is in the overlapping area, register in advance, and use the original link to register with the new foreign agent. However, since the implementation of this solution is still based on a single link, the time coordination between the Layer 3 handover and the Layer 2 handover is very important. If the original link is disconnected before the network layer handover is completed, the communication will still be interrupted. causing packet loss.

Contents of the invention

In view of this, the main purpose of the present invention is to provide a method for implementing network layer handover in Mobile IP, which can ensure that the communication is not interrupted during the handover process of the mobile node, thereby improving the handover speed and reducing the packet loss rate in the handover process .

In order to achieve the above object, technical solution of the present invention is achieved in that way:

A method for implementing network layer handover in a mobile IP, establishing a dual link for communication between a mobile node and its current base station, the method comprising the following steps:

a. After the network layer determines that handover is required, disconnect one of the established dual links, and establish a new link connection between the current mobile node and the target base station to be handed over to;

b. The mobile node registers with the home agent through the newly created link in step a, and performs network layer handover;

c. After the registration is completed, disconnect another link connection between the current mobile node and the original base station, and establish a link connection between the current mobile node and the target base station to be handed over to.

In the above solution, when the mobile node switches from the home network to the foreign network, step b specifically includes:

b11. The mobile node obtains a new care-of address through the link connection established with the target base station;

b12. The mobile node establishes a tunnel, and sends a registration request to the home agent through the newly established link;

b13. After receiving the registration request, the home agent establishes a tunnel leading to the care-of address, sends a registration response, and sends an agent address resolution protocol ARP message;

Then the completion of the registration in step c is: the mobile node receives the registration response sent by the home agent.

When the mobile node switches from one foreign network to another foreign network, step b specifically includes:

b21. The mobile node obtains a new care-of address through the link connection established with the target base station;

b22. The mobile node establishes a tunnel after obtaining the care-of address, and sends a registration request to the home agent through the newly established link, and the registration request carries an instruction to retain the original care-of address;

b23. After receiving the new registration request, the home agent retains the original care-of address, stores the new care-of address carried in the registration request, and sends a registration response to the mobile node;

b24. After the mobile node receives the registration response, it sends a deregistration request from the link reserved between the mobile node and the original base station;

b25. After receiving the deregistration request, the home agent deletes the original care-of address and sends a registration response to the mobile node;

Then the completion of the registration in step c is: the mobile node receives the registration response sent by the home agent after deleting the original care-of address.

When the mobile node switches back to the home network from the foreign network, step b specifically includes:

b31. The mobile node sends a gratuitous ARP message in the home network, and sends a deregistration request to the home agent;

b32. After receiving the deregistration request, the home agent deletes the care-of address and tunnel of the mobile node, and returns a registration response to the mobile node;

Then the completion of the registration in step c is as follows: the mobile node deletes the tunnel established with the foreign agent after receiving the registration response sent by the home agent.

For the above solution, during the handover process, after receiving duplicate data sent from different links, the mobile node filters a piece of duplicate data according to the existing protocol.

In the above solution, before the handover of the mobile node or after the handover is completed, the two links established by the mobile node and the current base station are in the same network and share a care-of address.

For the above solution, the method also includes: pre-selecting the trigger parameters to be detected, and setting corresponding trigger conditions; before performing step a, the method further includes: the current mobile node detects the selected trigger parameters of all base stations currently discovered Parameter information; judge whether the detected trigger parameter information meets the set trigger condition, if so, then judge whether the base station that currently satisfies the trigger condition belongs to its current subnet, if not, trigger network layer handover ; If the trigger condition is not met or it belongs to the current subnet, continue detection.

Then, the method further includes: the current mobile node obtains and stores all the base station information of the subnet where it is currently located, and obtains the base station information of all the currently discovered base stations in real time; The base station information of the base station that satisfies the trigger condition is matched with the base station information of all base stations stored by itself. If none of them match, the base station that meets the trigger condition does not belong to the current subnet of the current mobile node, otherwise it belongs to the current subnet of the current mobile node .

In the above solution, the selected trigger parameter is received signal strength or signal-to-noise ratio. The base station information is the base station identification ID, or the service set identification SSID of the subnet to which the base station belongs.

The method for realizing network layer switching in a kind of mobile IP provided by the present invention has following advantages and characteristics:

1) The present invention introduces the concept of dual links in mobile IP handover, that is: the mobile node maintains the communication of two wireless links simultaneously, and in the handover process, one link is used to establish a new connection, while the other link uses In order to maintain the original connection and maintain data communication, it can ensure that the communication is not interrupted during the switching process, thereby reducing the packet loss rate.

2) In the present invention, when the mobile node is in the non-handover stage, two links are used in parallel, so that hardware resources can be fully utilized, bandwidth can be increased, and the purpose of load balancing can be achieved.

3) In the present invention, when the mobile node enters the overlapping network coverage area that may be handed over, it is judged whether it is necessary to perform network layer handover when the original network is still available, and if necessary, a link is used to establish a link with the new network Connection, so as to realize the simultaneous access of the original network and the new network before the original network is unavailable, fundamentally improve the switching speed, and effectively reduce packet loss, which is an effective way to achieve seamless switching in mobile IP .

4) The present invention is an extension to the standard mobile IP protocol, only slightly changes the existing processing, and is fully compatible with the standard mobile IP system, therefore, it is simple and convenient to realize, and is easy to carry out function expansion on the existing mobile IP system.

Description of drawings

Fig. 1 is the basic network topology structure chart that each entity in the mobile IP is formed;

Fig. 2 is the processing flowchart that the present invention realizes;

Fig. 3 is a schematic diagram of the handover process of the mobile node moving from the home network to the foreign network in the present invention;

FIG. 4 is a schematic diagram of a handover process in which a mobile node moves between two foreign networks in the present invention;

Fig. 5 is a schematic diagram of the handover process of the mobile node returning to the home network from the foreign network in the present invention;

FIG. 6 is a schematic diagram of the relationship between dual links and upper-layer applications in the present invention;

FIG. 7 is a network topology diagram of a specific embodiment of the present invention.

Detailed ways

The core idea of the present invention is: establish dual links for each mobile node, when the mobile node is in the non-handover state, the mobile node uses the established dual links to communicate in parallel; when the mobile node is in the handover state, only disconnect One of the dual links, that is, use one link to maintain the original connection, and use the other link to establish a new connection. After the network layer switching of the new link is completed, the link that maintains the original connection is disconnected. Create a new connection.

Each mobile node includes two states: one is the state that the mobile node is in the process of handover; the other is the state that the mobile node is in normal communication before or after the handover.

The premise of the present invention is that a dual link for communication is established between the mobile node and the base station to which it currently belongs.

Based on this premise, in the present invention, the process in which the mobile node utilizes dual links to realize network switching is shown in Figure 2, including the following steps:

Step 201: When the network layer of the mobile node judges that handover is required, the system releases the current connection of the first link, that is, disconnects the original connection of any link in the dual link, and the second link assumes the current network All communication of the , the first link is used to establish a connection with the base station in the new wireless subnet.

Step 202: After the bottom layer, that is, the connection between the physical layer and the link layer is established, the mobile node performs network layer handover and registers with the home agent through the newly established link. That is to register between the new foreign agent and the home agent, including: first, the mobile node needs to obtain a new care-of address from the new foreign agent, then register with the home agent, and send the new care-of address to the home agent; After receiving the registration of the new care-of address, the original care-of address is not deleted immediately, but the two registrations are kept bound, and each data packet sent to the mobile node is sent to each care-of address of the mobile node through the tunnel . In this case, the mobile node will receive duplicate data through the dual link, then the mobile node will filter a piece of duplicate data according to the processing method for duplicate data in the existing protocol.

Step 203: After the new link is stably formed, the system releases the original connection of the second link, and establishes a connection between the second link and the base station in the new wireless subnet and the current foreign agent of the mobile node, At this point, the normal communication process is restored.

The handover of the mobile node can be divided into three situations: the mobile node moves from the home network to the foreign network, the mobile node moves between different foreign networks, and the mobile node returns to the home network from the foreign network.

When the mobile node leaves the home network for handover, it uses one link to register and another link to communicate during the handover process. As shown in FIG. 3 , link 30 is a link for maintaining the original connection to continue communication, and link 31 is a link for disconnecting the original connection and establishing a new connection with the foreign agent for registration. The handover process when the mobile node moves from the home network to the foreign network includes the following steps:

a1. When the mobile node moves out of the home network and enters a foreign network, the mobile node connects any one of the two links established by itself to the foreign network. When a new connection is established, the mobile node passes this connection Obtain the care-of address in the new network, for example: use the link 31 to obtain the care-of address.

b1. A tunnel is established between the mobile node and the foreign agent, and a registration request is sent to the home agent through the newly established link, that is, the mobile node registers with the home agent via links 31, paths 32, 33, and 34.

c1. After receiving the registration request, the home agent establishes a tunnel leading to the care-of address, sends a registration response, and sends an Agent Address Resolution Protocol (ARP) message. Here, the purpose of sending the ARP message is to indicate that the mobile node has left the home network, and all data packets sent to the mobile node will be received and forwarded by the home agent.

d1. After the mobile node receives the registration response from the home agent, indicating that the new connection has been formed stably, it disconnects the other link that maintains the original connection in the home network, reconnects to the foreign network, and completes the handover.

When the mobile node leaves a foreign network and enters another foreign network, the home agent adopts multiple binding methods during the handover process, that is, it stores two care-of addresses for the mobile node and binds two routes at the same time. Referring to Fig. 4, the handover process when the mobile node moves between different foreign networks includes the following steps:

a2. When the mobile node roams from foreign network 1 to foreign network 2, it disconnects one of the two links established with foreign network 1 and connects to foreign network 2. After a new connection is established, the mobile node The node obtains the care-of address in the new network through this connection.

b2. The mobile node establishes a tunnel after obtaining the care-of address, and sends a registration request to the home agent through the newly established link. It should be noted that: at this time, the registration request needs to set the "S" bit. The format of the registration request message stipulated in the standard mobile IP protocol is shown in Table 1, wherein, the S bit represents the binding information at the same time, if S is set, it means that the mobile node requires the home agent to retain its previous binding information. 0 1 2 3 type S B D. m G R T x Lifetime Home Address Home Agent Care-of Address Description Extensions...

Table I

c2. Since the mobile node has already registered a care-of address on the home agent, the home agent does not delete the original care-of address after receiving a new registration request, but just adds another one, and sends a registration response to the mobile node. Thereafter, the home agent sends each data packet received and sent to the mobile node to each care-of address of the mobile node through the tunnel.

d2. After receiving the registration response, the mobile node sends a deregistration request message from the original link.

e2. The home agent receives the deregistration request, deletes the original configured care-of address, and sends a registration response. Thereafter, the home agent forwards data only to the new care-of address.

f2. After receiving the registration response, the mobile node also disconnects the other link that maintains the original connection with the foreign network 1, reconnects to the foreign network, and completes the handover.

When the mobile node returns to the home network from the foreign network, the handover process adopts a link registration, and the other link continues to communicate through the foreign network. As shown in Figure 5, the handover process when the mobile node returns to the home network from the foreign network includes the following steps:

a3. When the mobile node returns to the home network from the foreign network, it disconnects one of the two links established with the foreign network and connects to the home network. When the new connection is established, the mobile node is in the home network Send a gratuitous ARP message and send a deregistration request to the home agent.

Here, the purpose of sending the ARP message is to inform the mobile node that it has returned to the home network.

b3. After receiving the deregistration request, the home agent deletes the care-of address and tunnel of the mobile node, and returns a registration response to the mobile node.

c3. After the mobile node receives the registration response sent by the home agent, it deletes the tunnel established with the foreign agent, and disconnects the other link that maintains the original connection in the foreign network, and reconnects to the home network. Complete the switch.

In FIG. 3 , FIG. 4 , and FIG. 5 , dotted lines represent one-way tunnels, solid lines represent registration messages, dotted lines represent normal communication data, and rectangles on solid lines represent tunnels.

Before the handover of the mobile node or after the handover is completed, if the mobile node is in a foreign network, the home agent only binds the care-of address of the network where the mobile node is located, so all data packets sent to the mobile node are forwarded to the on the care-of address. At this time, the two links established by the mobile node are in the same subnet and share a care-of address. Although there are two links, a virtual link layer can be used to make them use one IP address for communication, so that the upper layer application cannot realize the existence of dual links. As shown in Figure 6, in practical applications, the two links of the mobile node can be implemented through two network cards respectively. From the perspective of the upper layer application of the mobile node, the mobile node has only one IP address, but the datagram it sends is through transmitted over two links. When using which link to send, you can choose according to the busy or idle status of the link. For example, when you need to choose, first judge the busy or idle status of the two links, and use the link with the light load to send, so that both It can improve bandwidth and realize load balancing.

In the above dual-link switching process, the establishment of the new link needs to be carried out while the original link is still available. Therefore, a method for triggering network layer switching proposed in another patent application of the applicant can be combined to trigger Establishment of a new link in a dual link. The main idea of this method is: the current mobile node detects the information of the selected trigger parameters of all the currently discovered base stations, and obtains the base station information of all the currently discovered base stations in real time, and the current mobile node also obtains and stores the current subnet information of all base stations; then, the current mobile node judges whether the detected trigger parameter information satisfies the set trigger condition, and if so, judges according to the base station information whether the base station currently satisfying the trigger condition belongs to its current subnet, if If it does not belong, trigger network layer switching; if it does not meet the trigger condition or belongs to the current subnet, continue detection. That is: match the base station information of the base station that satisfies the trigger condition with the base station information of all base stations stored by itself, if none of them match, the base station that meets the trigger condition does not belong to the subnet where the current mobile node is currently located, and triggers a network layer handover; otherwise , belongs to the subnet where the current mobile node is currently located, and continues to detect.

Wherein, the physical layer channel information includes the ID of the currently connected base station, the list of available basic service sets, the received signal strength, the signal-to-noise ratio, etc., and one of them can be selected as a trigger parameter; the trigger condition refers to The threshold corresponding to the selected trigger parameter is set in advance; base station information refers to information that can uniquely identify each base station, such as: base station ID, SSID of the subnet to which the base station belongs, and so on. The mobile node can use the existing information in the existing message, such as the SSID of the subnet to which the base station belongs to obtain all the base station information of the current network; or use the extensible field in the existing message to carry the required base station information, such as using the registration message The extension field carries all base station IDs.

Referring to FIG. 7 and in conjunction with a specific embodiment, the handover process of establishing dual link and lower layer active triggering of network layer handover will be described in detail. FIG. 7 is a network topology diagram of the mobile IP system where the mobile node 70 is located. As shown in Figure 7, the mobile IP system is a wireless local area network using the 802.11 protocol, and the wireless local area network includes three subnets: home network, foreign network 1 and foreign network 2. The three subnets are connected by a router, and each subnet includes an agent, namely: home agent, foreign agent 1 and foreign agent 2. Each subnet includes two base stations, ie, access points AP1-AP6. The six dashed circles in FIG. 7 represent the coverage areas of AP1-AP6 respectively. Wherein, the access points AP1 and AP2 belong to the home network, the access points AP3 and AP4 belong to the foreign network 1 , and the access points AP5 and AP6 belong to the foreign network 2 . The dotted line in Fig. 7 represents the moving track of the mobile node 70. It can be seen that the mobile node 70 starts from the home network, first roams to the foreign network 1, then roams from the foreign network 1 to the foreign network 2, and finally roams from the foreign network 2 back home. The moving process includes three handovers: handover from home network to foreign network, handover between two foreign networks, and handover from foreign network back to home network.

In this embodiment, the mobile node 70 selects two 802.11 wireless network cards to simulate dual links, and the network layer switching is actively triggered by the lower layer according to the detected physical channel information.

The mobile node 70 is initially located in the home network, and when the mobile node 70 moves to the overlapping coverage area 71 of AP2 and AP3, the second-layer trigger starts to notify the network layer that the mobile node 70 will move from the home network to the foreign network 1, triggering the network layer to start handover . At this time, the mobile node 70 disconnects one of the two links established with AP2, and establishes a connection with AP3; while the other link remains connected with AP2. After the link between the mobile node 70 and AP3 is established, the mobile node 70 obtains the care-of address 1 in the foreign network 1 through this connection; and establishes a tunnel according to the care-of address 1, and sends a registration request to the home agent through the link with AP3.

After the home agent receives the registration request, it sets up a tunnel leading to the care-of address 1, and sends a registration response to the mobile node 70, and sends a proxy ARP message at the same time; It is also disconnected, and the link is connected to AP3 as the second link between the mobile node 70 and AP3. In this way, the entire handover process from the home network to the foreign network 1 is completed.

Taking the signal-to-noise ratio as the trigger parameter and the base station information as the MAC address of the AP as an example, the specific process of the two-layer trigger start is: when the mobile node 70 enters the junction of the home network and the foreign network 1, that is, when moving from AP2 to AP3, As the distance from AP3 gets closer, the mobile node 70 will find that the SNR of AP3 has met the set trigger condition, that is, the preset SNR threshold; at this time, the mobile node 70 searches and obtains the AP3 MAC address, compare the MAC address of AP3 with the MAC addresses of all APs in the current subnet currently saved by itself, and find that the MAC address of AP3 does not match all the APs currently saved by itself, so it can be determined that AP3 belongs to the new subnet , then the mobile node 70 actively triggers the handover of the network layer.

When the mobile node 70 moves to the overlapping coverage area 72 of AP4 and AP6 along the moving track, the layer 2 trigger starts, notifies the network layer that the mobile node 70 is about to enter a new foreign network, and triggers the network layer to start switching. Mobile node 70 disconnects a link connection in the two links established between AP4, and establishes connection with AP6, but still maintains a link and connects with AP4; Care-of address 2 in network 2. The mobile node 70 establishes a tunnel, and sends a registration request to the home agent through the link with AP6. At this time, the registration request needs to set the "S" bit.

Although the mobile node 70 has registered a care-of address 1 on the home agent, the home agent does not delete the care-of address 1 after receiving a new registration request with the S bit set, but adds the care-of address 2 to the binding table, and send a Registration Reply to the mobile node 70. Thereafter, the home agent sends each data packet sent to the mobile node 70 to the two care-of addresses 1 and 2 through the tunnel. After receiving the registration response, the mobile node 70 sends a deregistration request to the home agent through the link with AP4. After receiving the deregistration request, the home agent deletes the care-of address 1, and returns a registration response to the mobile node 70, indicating that the care-of address 1 has been deleted. After that, the home agent only forwards data to care-of address 2. After receiving the registration response, the mobile node 70 disconnects the link with AP4, and inserts the link into AP6 as the second link between the mobile node 70 and AP6. The entire switching process from network 1 to foreign network 2.

When the mobile node 70 moves to the overlapping coverage area 73 of AP5 and AP2 along the moving track, the layer 2 trigger starts, notifies the network layer that the mobile node 70 will return to the home network from the foreign network 2, and triggers the network layer to start switching. At this time, the mobile node 70 disconnects one of the two links established with AP5, establishes a connection with AP2, and maintains the other link connection with AP6. After the link with AP2 is established, the mobile node 70 sends a free ARP message in the home network through the connection established with AP2, indicating that it has returned to its hometown, and sends a deregistration request. After receiving the deregistration request, the home agent deletes the care-of address 2 and the tunnel of the mobile node 70, and sends a registration response to the mobile node 70. After the mobile node 70 receives the registration response, it deletes the tunnel, disconnects the link connection with AP5, and connects to AP2 on the home network as the second link between the mobile node 70 and AP2. The entire switching process from the foreign network 2 back to the home network.

In this embodiment, the scheme that the network layer switching is performed ahead of the low layer switching is combined with the scheme of establishing a dual link, which not only greatly shortens the network layer switching time, but fundamentally improves the switching speed; moreover, for various switching situations, all It can ensure uninterrupted communication during the whole switching process and avoid packet loss.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention.

Claims (10)

1. A method for realizing network layer handover in a mobile IP, characterized in that, a dual link for communication is set up between a mobile node and its current affiliated base station, the method may further comprise the steps: a. After the network layer determines that handover is required, disconnect one of the established dual links, and establish a new link connection between the current mobile node and the target base station to be handed over to; b. The mobile node registers with the home agent through the newly created link in step a, and performs network layer handover; c. After the registration is completed, disconnect another link connection between the current mobile node and the original base station, and establish a link connection between the current mobile node and the target base station to be handed over to. 2. The method according to claim 1, wherein when the mobile node switches from the home network to a foreign network, step b specifically includes: b11. The mobile node obtains a new care-of address through the link connection established with the target base station; b12. The mobile node establishes a tunnel, and sends a registration request to the home agent through the newly established link; b13. After receiving the registration request, the home agent establishes a tunnel leading to the care-of address, sends a registration response, and sends an agent address resolution protocol ARP message; Then the completion of the registration in step c is: the mobile node receives the registration response sent by the home agent. 3. The method according to claim 1, wherein when the mobile node switches from one foreign network to another foreign network, step b specifically includes: b21. The mobile node obtains a new care-of address through the link connection established with the target base station; b22. The mobile node establishes a tunnel after obtaining the care-of address, and sends a registration request to the home agent through the newly established link, and the registration request carries an instruction to retain the original care-of address; b23. After receiving the new registration request, the home agent retains the original care-of address, stores the new care-of address carried in the registration request, and sends a registration response to the mobile node; b24. After the mobile node receives the registration response, it sends a deregistration request from the link reserved between the mobile node and the original base station; b25. After receiving the deregistration request, the home agent deletes the original care-of address and sends a registration response to the mobile node; Then the completion of the registration in step c is: the mobile node receives the registration response sent by the home agent after deleting the original care-of address. 4. The method according to claim 1, wherein when the mobile node switches from a foreign network to a home network, step b specifically includes: b31. The mobile node sends a gratuitous ARP message in the home network, and sends a deregistration request to the home agent; b32. After receiving the deregistration request, the home agent deletes the care-of address and tunnel of the mobile node, and returns a registration response to the mobile node; Then the completion of the registration in step c is as follows: the mobile node deletes the tunnel established with the foreign agent after receiving the registration response sent by the home agent. 5. The method according to any one of claims 1 to 4, characterized in that, during the handover process, after receiving duplicate data sent from different links, the mobile node filters a copy of duplicate data according to an existing protocol. 6. The method according to claim 1, wherein before or after the handover of the mobile node, the two links established by the mobile node and the current base station are in the same network and share a care-of address. 7. The method according to any one of claims 1 to 4, characterized in that the method further comprises: pre-selecting the trigger parameters to be detected, and setting corresponding trigger conditions; before performing step a, the method also Further includes: The current mobile node detects the selected trigger parameter information of all base stations currently discovered; judges whether the detected trigger parameter information satisfies the set trigger condition, and if so, then judges whether the base station that currently satisfies the trigger condition belongs to its current If it does not belong to the subnet it belongs to, it will trigger network layer switching; if it does not meet the trigger conditions or it belongs to the current subnet, it will continue to detect. 8. The method according to claim 7, further comprising: the current mobile node obtains and stores all base station information of the subnet where it is currently located, and obtains the base station information of all currently discovered base stations in real time; Then the judgment whether it belongs to the current subnet is as follows: the current mobile node matches the base station information of the base station satisfying the trigger condition with the base station information of all base stations stored in itself, if none of them match, the base station satisfying the trigger condition does not belong to the current mobile node. The subnet where the node is currently located, otherwise it belongs to the subnet where the current mobile node is currently located. 9. The method according to claim 7, wherein the selected trigger parameter is received signal strength or signal-to-noise ratio. 10. The method according to claim 8, wherein the base station information is a base station identification ID or an SSID of a subnet to which the base station belongs.

Images