CN1829177A - Method for Realizing Load Balancing among Access Devices in Wireless Local Area Network - Google Patents

Abstract

The present invention reefs to load equalization method between access equipment wireless local area networks. It contains configuring accessing priority and priority modifying strategy to each covering same area AP, each AP according to own priority modifying strategy and own accessing priority, maintenance adjacent AP neighbor information in each AP, each AP through detecting AP link or receiving neighbor AP broadcast information through wireless channel, updating oneself maintenance neighbor information, AP through comparing own and adjacent AP priority determining own accessing site state. The present invention overcame load unbalance in access equipment in wireless local area networks of current technology, convenient realizing access equipment load equalization.

Description

Method for Realizing Load Balancing among Access Devices in Wireless Local Area Network

technical field

The invention relates to the field of communication technology, in particular to a method for realizing load balancing between access devices in a wireless local area network.

Background technique

WLAN (Wireless Local Area Network) uses wireless technology to transmit data, voice and video signals in the air. As an alternative or extension of the traditional wiring network, it allows users to expand and extend the wired network arbitrarily. From personal and desktop wireless access to cross-regional wireless data communication transmission, users can obtain information anytime and anywhere, improving the efficiency of network use.

As shown in Figure 1, it is a schematic diagram of the WLAN network structure. In a WLAN network, a wireless AP (wireless access point) constitutes a BSS (Basic Service Set), and the BSSID (Basic Service Set Identification Code, generally the MAC address of the AP) To identify. The services that the AP can provide are identified by the SSID (service set identification code, that is, a character string that identifies the services that the AP can provide), and multiple APs can use the same SSID. The STA (station, such as a notebook with wireless function) can access the BSS and network resources other than the BSS through the DS (distribution system) service provided by the AP.

In a wireless local area network (WLAN) system, a STA has two ways to select an AP to access: one is to select access in a passive manner, and the other is to select access in an active manner. The passive access mode is that the STA scans various channels, discovers APs in the WLAN by receiving Beacon frames sent regularly by the AP, and then selects an AP to access from the scanning results. The active access method is that the STA sends a probe request (Probe Request) frame according to the SSID to inquire about the APs in the network that support the SSID, and the STA selects an AP to access according to the inquiry result.

Since WLAN is a shared medium technology, when the number of access users increases, the competition among STAs will increase, and the bandwidth allocated to each user will decrease. Therefore, in practical applications, in order to increase the user capacity of the WLAN, multiple APs are configured to cover the same area on different channels for networking. At this time, bandwidth resources can be fully utilized only if the load among APs can be balanced.

Figure 2 shows the channel settings of APs covering the same area. When APs work on the same channel, the signals between APs will interfere with each other. Therefore, APs covering the same area need to work on independent channels. Taking 802.11b as an example, three independent channels can be used: 1 channel, 6 channels and 11 channels.

For STAs, regardless of whether they access APs passively or actively, STAs do not consider the load of APs when they choose to access APs. enter. For APs, since there is no protocol for APs to notify each other of their loads, the current common practice is that APs do not communicate with each other, and each AP configures its own maximum user independently to decide whether to access STAs. When the number of AP access users reaches the maximum, it will refuse to access new users. The disadvantage of this method is that multiple APs provide access services at the same time, and the STA may always select the one with the strongest signal for access. In this way, the load among APs cannot be balanced, and some APs access too many STAs, resulting in performance degradation, while some APs are idle.

At present, there is no established standard for load balancing. Although many APs also support load balancing, they all use self-defined private protocols, and there is no way for APs from different manufacturers to play the role of load balancing.

In some existing technologies, APs are wired to each other through Ethernet to exchange their own load information, so as to perform load balancing. The AP notifies its own load information in real time, including the number of users connected to the AP, existing traffic information, and so on. For an AP that performs load balancing based on the number of users, it needs to notify other APs in the network of its user information in real time every time a user goes online or offline; Advertises its traffic information to other APs in the network.

The disadvantage of the prior art is that: since there are many APs connected to the same LAN through Ethernet, the STA can search for multiple APs when the coverage areas of the APs overlap each other. The load between the APs cannot be balanced, and there are too many STAs connected to some APs, resulting in performance degradation, while some APs are idle. If the load information is transmitted between APs in a wired manner, the problem is that the APs connected by wires do not necessarily cover the same area, and the AP cannot determine which APs cover the same area as itself, and the neighbor information thus obtained has a lot of redundancy. . Moreover, since the information such as the number of users and traffic of an AP changes in real time, the AP needs to constantly notify the adjacent APs of their own load information, and there is a large amount of information transmission. At the same time, using Ethernet for AP interconnection will increase the workload of networking equipment and wiring, which is not suitable for flexible networking. Wired transmission of information limits the installation and use of APs.

Contents of the invention

The technical problem to be solved by the present invention is: to overcome the unbalanced load among the access devices in the prior art wireless local area network, or to transmit load information through wired interconnection, which will lead to a large amount of redundant information transmission and inflexible shortcomings, and to provide a wireless local area network A method for implementing load balancing among access devices, so as to realize load balancing of access devices conveniently and flexibly.

The technical scheme that the present invention adopts for solving the problems of the technologies described above is:

This method for implementing load balancing between access devices in a wireless local area network includes the following steps: configuring access priority and priority change policies for each wireless access point covering the same area, each wireless access point according to its own The priority change policy changes its own access priority; the neighbor information of adjacent wireless access points is maintained on each wireless access point, and each wireless access point regularly updates the neighbor information maintained by itself; the wireless access point The access point determines the status of its own access site by comparing the priority of itself and the adjacent wireless access point. Only the wireless access point in the allowed access state can broadcast beacon information and respond to the inquiry request of the site.

The wireless access point updates the neighbor information maintained by itself by detecting the link between the wireless access points or receiving the information broadcast by the neighbor wireless access point through the wireless channel regularly; when the wireless access point cannot receive the information within a period of time Information about a certain neighboring wireless access point, or after detecting that the link of a certain neighboring wireless access point is interrupted, the neighboring wireless access point may be deleted from the maintained neighbor information. If the current priority of the wireless access point is the highest and there is no wireless access point with the same priority in the neighbors, the access status of the wireless access point is allowed, and the rest are not allowed; when there are multiple wireless access points When the current priority of the access point is the highest, each wireless access point adopts the same strategy to determine its own access status, ensuring that only one wireless access point currently has an access status of allowed access.

The wireless access point can use the number of access users as the strategy for changing the priority, and configure the user access threshold. When the number of access users of the wireless access point reaches the access threshold, the priority will drop by one level. After the access state of the wireless access point in the allowed access state becomes not allowed, it continues to broadcast beacon frames, but no longer responds to new inquiry requests until all access stations of the wireless access point go offline.

When a station is only within the coverage of one wireless access point and is not allowed to access, if the wireless access point receives the inquiry request from the same station several times in succession, the wireless access point can respond to ensure Stations always have access to a wireless access point.

The wireless access points can communicate in a broadcast or multicast manner, encrypt the exchange information between the wireless access points, and before exchanging priority information, the wireless access points authenticate each other and establish Communication link, only when the state between the wireless access points is connected and authenticated, the priority information can be transmitted between the wireless access points.

When a wireless access point joins the network or its status changes, it can use the status notification message to actively notify the adjacent wireless access points through the wireless channel, so that the adjacent wireless access points update their neighbor information. The status notification The message at least includes the media access control address of the wireless access point, the service set identification code and the access priority. When a wireless access point exits the network, it can use the exit message to actively notify the adjacent wireless access point through the wireless channel, so that the adjacent wireless access point updates its neighbor information. The exit message includes at least wireless The media access control address and service set identifier of the access point.

When the wireless access point supports multiple SSIDs, neighbor information, access priority and access state can be maintained for each SSID.

The beneficial effects of the present invention are: the present invention provides a method for APs to inform each other of wireless access information through a wireless channel, and on this basis to realize load balancing among APs. AP has the ability of neighbor discovery and self-organizing network, can automatically adjust its own access status according to network conditions, and dynamically realize load balancing between APs; at the same time, it can support load balancing between APs in multiple SSID environments, and can Ensure the security of information exchange between APs and prevent the existence of illegal APs.

Each AP can independently determine its own access priority according to its own configuration, and each AP can flexibly set its own access threshold, such as accessing according to the maximum number of users, accessing according to the maximum flow, etc., can support multiple A load balancing strategy. Only when the load of the AP causes the access priority of the AP to change, the AP needs to notify other APs, instead of notifying other APs every time a user goes online or goes online or the traffic changes, so that the interactive information between APs is reduced, and the access is convenient and flexible. Device load balancing.

Description of drawings

FIG. 1 is a schematic diagram of a WLAN network structure;

Figure 2 is a schematic diagram of AP channel settings covering the same area;

FIG. 3 is a structural schematic diagram of multiple APs simultaneously providing access services in the same area of the present invention;

Fig. 4 is a protocol stack diagram of communication between APs of the present invention;

FIG. 5 is a schematic diagram of a connection state machine between APs in the present invention.

Detailed ways

Below according to accompanying drawing and embodiment the present invention will be described in further detail:

In order to overcome the deficiencies of the prior art, the present invention defines a protocol for communicating wireless access information between APs through a wireless channel, and realizes load balancing between APs on this basis. The wireless channel for communication between APs and the channel for user access can use different frequencies or channels, or they can be the same, but it should be ensured that the communication between APs uses the same channel, otherwise it is difficult to ensure that other AP news.

As shown in Figure 3, it is a schematic diagram of the access service structure provided by multiple APs in the same area at the same time. The present invention configures access priority and priority change policies for each wireless access point covering the same area. Each wireless access point Change its own access priority according to its own priority change strategy; maintain the neighbor information of adjacent wireless access points on each wireless access point, and each wireless access point detects the link between wireless access points Or receive the information regularly broadcast by neighboring wireless access points through the wireless channel, and update the neighbor information maintained by itself. The wireless access point determines the status of its own access site by comparing the priorities of itself and neighboring wireless access points. Only wireless access points in the access-allowed state can broadcast beacon information and respond to station inquiry requests.

The realization process of the present invention is described in detail below:

1. Configure access priority and priority change policy

Configure the access priority and priority change policy for each AP. For example, you can configure the user access threshold. When the number of AP access users reaches the access threshold, the priority will be lowered by one level. Each AP can independently configure its own priority change strategy. For example, the number of access users is used as the priority change strategy. The AP priority change rules are as follows:

a) After the STA is successfully connected, the current number of users connected to the AP + 1; if the number of users currently connected to the AP >= the threshold, the priority is -1; in other cases, the priority remains unchanged;

b) After the STA is disconnected, the current number of AP access users -1; if the current number of AP users = threshold -1, the priority is +1; in other cases, the priority remains unchanged.

2. Maintain neighbor information

APs that can support load balancing need to maintain information about neighboring APs. Each AP with load balancing capabilities should actively notify neighboring APs when it joins the network, changes status, or exits the network, so that neighboring APs can update their neighbor information. At the same time, each AP needs to maintain its own neighbor table, including detecting the link status between APs, or the AP periodically broadcasts its own information, so that neighboring APs can periodically age the neighbor information maintained by themselves. When the information of a certain AP cannot be received for a period of time, or the link of a certain AP is detected to be interrupted, the AP should be deleted from the neighbor table.

The load sharing between APs takes the same SSID as the unit, and load sharing is performed between APs configured with the same SSID. In the case of supporting multiple SSIDs, the AP needs to maintain a separate neighbor information table, access priority, and access status for each SSID.

3. Priority comparison and access state determination

APs decide whether to access STAs by comparing their own priorities with those of neighboring APs. If the APs have the same priority, the APs can determine that an AP has the right to effectively access the STA based on the same rules. For example, the AP with the smaller MAC address can be selected to have the right to access the STA preferentially.

When the access priority of an AP changes, all neighbors need to be notified. And change the access state according to the neighbor information received by itself. Each AP determines its own access status according to the priorities of itself and neighboring APs. The following rules ensure that currently only one AP sets its access status to enable and can access users.

a) If the current priority of the AP is the highest and there is no AP with the same priority in the neighbors, the access status of the AP is enabled;

b) When there are multiple wireless access points with the highest current priority, the wireless access points use the same strategy to determine their own access status, for example, select the one with the smallest Media Access Control (MAC) address, and its wireless access Click the access status to enable;

c) Based on the strategy selected in b), if the current priority of the AP is the highest, but there is an AP with a smaller MAC address among the APs with the same priority, the access status of the AP is disabled;

d) In other cases, the access status of the AP is disabled;

When the priority of any one of the neighboring APs changes, or the topology changes, all neighboring APs need to adjust their access status according to the above rules. Topology changes include:

a) A new AP is added to the network;

b) The AP actively withdraws from the network;

c) The AP cannot actively withdraw from the network due to sudden power failure, etc. When other APs detect that an AP has no signal for a long time, it will be aged out from the neighbor list.

To sum up, when the priority of itself changes, the priority of neighbor APs changes, or the topology of neighbors changes (new APs join or APs exit or age), APs need to adjust their own access status.

Only the AP that is currently in the enabled state can broadcast beacon (beacon frame) information and respond to the inquiry request of the STA. Since the STA generally maintains the connection with the AP by receiving the beacon information of the AP, in order to ensure that the connected STA will not go offline after the access status of the enabled AP changes to disabled, it should continue to broadcast beacon frames, but no longer respond. A new inquiry request until all STAs go offline. At this time, although the AP can broadcast the beacon, it cannot respond to the STA's inquiry request and cannot access new STAs.

When the STA is connected, it can search for beacon information. Since only one AP broadcasts beacon information within a coverage area, it is guaranteed that the STA can only access one AP. Or, in order to find a suitable AP, the STA first initiates a Probe Request (probing request). When a STA is within the overlapping coverage of multiple APs, multiple APs receive the Probe Request at the same time, and each AP decides whether to access the corresponding STA according to its own access status. The AP whose current access status is enabled sends a Probe Response (probe response) to the STA.

When the STA is only within the coverage of one AP, if the access status of this AP happens to be enable, the AP can access users. If the access state of this AP is just in the disabled state, in order to ensure that the STA can always access an AP, if the AP in the disabled state receives the probe request from the same STA n consecutive times, the AP in the disabled state can respond .

If the message transmission between APs is transmitted in the form of plain text, the problem faced is that the illegal AP can set its own priority to the highest, and the user will always connect to the illegal AP, while the legal AP is in an idle state. Therefore, message transmission between APs requires the guarantee of a security mechanism. Since neither the 802.11 control frame nor the management frame can be encrypted, the information exchanged between APs is carried in the 802.11 data frame, which can be encrypted using WEP/WPA or other encryption methods. The 802.11 protocol layer provides functions such as data encryption, integrity verification, and replay protection by supporting protocols such as WEP/802.11i. As shown in Figure 4, it is a protocol stack diagram of communication between APs of the present invention, the MAC (Media Access Control) layer adopts 802.11 MAC, on which functions such as security are provided by enhanced protocols such as 802.11i, and the interactive protocol between APs is carried as data in In the data of the 802.11MAC frame.

Before exchanging priority information between APs, APs need to authenticate each other and establish a communication link. In the method of establishing a communication link, authentication between APs can use mechanisms such as PSK and 802.1x defined by 802.11i. During the authentication process, APs negotiate information such as keys for data encryption between APs.

Figure 5 is a schematic diagram of the connection state machine between APs. The connection states between APs are divided into:

a) Not connected; no communication link established between APs;

b) Connected but unauthenticated; a communication link has been established between APs, but mutual identity authentication has not been performed between APs; at this stage, the link between APs can be removed, that is, the connection between APs can be cancelled, and the connection between APs can be returned to the unauthenticated network. Connection Status.

c) Connected and authenticated; when the mutual identity authentication is performed between the APs, the AP enters the authenticated state. At this stage, the link between APs can be removed, that is, the connection between APs can be canceled to return to the unconnected state.

Only when the status of the APs is connected and authenticated, can the priority information be transmitted between the APs; thereby ensuring the security of the data between the APs. The AP establishes a neighbor relationship according to the received priority information, which ensures that illegal APs cannot spread false message information.

In the present invention, APs communicate by broadcasting or multicasting (including layer-2 multicasting and layer-3 multicasting). Two kinds of messages can be defined - status notification message and exit message. When an AP joins the network and its status changes, it uses a status notification message to notify the neighboring APs. After receiving the status notification message, the neighboring APs update the neighbor information and their own access status. To maintain AP neighbor information, the AP periodically sends status notification packets in broadcast or multicast mode. If it does not receive a status notification message from an AP within a certain period of time, it will delete the AP from its neighbors and update its access status. The detection of neighbors by the AP can also be completed by detecting the link status between APs, and at this time, there is no need to periodically send status notification messages. When an AP exits the network, the exit message is used to notify neighboring APs.

The message format of the communication protocol between APs is defined as follows:

0 1 2 3 Code Identifier Length Attributes…

Wherein: each message between APs needs to include message type code, message identifier Identifier and message length information. Other information that needs to be transmitted between APs is carried in the message as Attributes.

code: 1 byte, used to identify the message type:

1 is status-notify (status notification)

2 is leave (leave)

Identifier: 1 byte, packet identifier.

Length: 2 bytes. All packet lengths including Code, Identifier, Length, and Attribute. The minimum length of the message is 20, and the maximum length is 4096.

Attributes include the MAC address of the AP, the SSID of the AP, and the access priority of the AP.

(1) Message type

The value of the message Code field determines the message type. For status notification packets, the code type is status-notify. . The AP status notification message must include the MAC address of the AP, the SSID of the AP, and the access priority of the AP. Wherein, the MAC address of the AP uses the AP_id attribute, the SSID of the AP uses the AP_ssid attribute, and the access priority of the AP uses the AP_priority attribute.

For the leave message, the code type is leave; the AP leave message must include the MAC address of the AP and the SSID of the AP. Wherein, the MAC address of the AP uses the AP_id attribute, and the SSID of the AP uses the AP_ssid attribute.

(2) Attributes

The design of the message content adopts the TLV format. Its format is:

0 1 2 type Length value

Type: 1 byte, attribute type;

Length: 2 bytes, attribute length, indicating the length of this attribute.

Value: The value of the attribute. The value types are divided into text, string, address, integer, and time.

The following describes the attributes defined in the status notification message and the exit message:

1) Ap_id attribute

The MAC address used to identify the AP; the type is string and the length is 48bit

0 1 2 3 Ap_id Ap_id

2) Ap_ssid attribute

The SSID used to identify the AP; the type is string, and the length is specified by the attribute length.

3) AP_Priority attribute

It is used to identify the access priority of the AP; the type is integer, and the length is 4 bytes.

0 1 2 3 Ap_priority

Value: The priority decreases as the number increases, and 0 is the maximum priority.

The invention provides a method for informing each other of wireless access information between APs through a wireless channel, and on the basis of this, a method for realizing load balancing between APs. AP has the ability of neighbor discovery and self-organizing network, can automatically adjust its own access status according to network conditions, and dynamically realize load balancing between APs; at the same time, it can support load balancing between APs in multiple SSID environments, and can Ensure the security of information exchange between APs and prevent the existence of illegal APs.

Each AP can independently determine its own access priority according to its own configuration, and each AP can flexibly set its own access threshold, such as accessing according to the maximum number of users, accessing according to the maximum flow, etc., can support multiple A load balancing strategy. Only when the load of the AP causes the AP access priority to change, the AP needs to notify other APs, instead of notifying other APs every time the user goes online or goes online or the traffic changes, so that the interactive information between APs is reduced.

The present invention is applicable to APs covering the same area, these APs are configured with different channels, but the SSID, the authentication mode for STA, and the encryption mode are the same, and the STA can choose any AP to access. Those skilled in the art do not depart from the essence and spirit of the present invention, there can be many variants to realize the present invention, the above description is only a preferred and feasible embodiment of the present invention, and it does not limit the scope of rights of the present invention. The equivalent structural changes made in the description of the invention and the accompanying drawings are all included in the scope of rights of the present invention.

Claims (10)

1. A method for implementing load balancing between access devices in a wireless local area network, comprising the following steps: Configure access priority and priority change policy for each wireless access point covering the same area, and each wireless access point changes its own access priority according to its own priority change policy; Maintain the neighbor information of adjacent wireless access points on each wireless access point, and each wireless access point regularly updates the neighbor information maintained by itself; The wireless access point determines the status of its own access site by comparing the priority of itself and the adjacent wireless access point. Only the wireless access point in the access-allowed state can broadcast beacon information and respond to the inquiry request of the site. 2. The method for realizing load balancing between access devices in a wireless local area network according to claim 1, wherein the wireless access point detects the link between the wireless access points or receives a regular load from the neighboring wireless access points. Update the neighbor information maintained by itself through the information broadcast through the wireless channel; when the wireless access point cannot receive the information of a certain adjacent wireless access point within a period of time, or check the information of a certain adjacent wireless access point After the link is interrupted, the adjacent wireless access point is deleted from the maintained neighbor information. 3. The method for realizing load balancing between access devices in a wireless local area network according to claim 1 or 2, characterized in that: if the current priority of the wireless access point is the highest and there is no wireless access with the same priority among the neighbors point, the access status of the wireless access point is allowed access, and the rest are not allowed to access; when there are multiple wireless access points with the highest current priority, each wireless access point uses the same strategy to determine its The access status of the wireless access point is guaranteed to ensure that there is only one access point whose access status is allowed. 4. The method for implementing load balancing between access devices in a wireless local area network according to claim 3, wherein the wireless access point uses the number of access users as a strategy for changing priority, and configures a user access threshold, when When the number of users accessing the wireless access point reaches the access threshold, the priority is lowered by one level. 5. The method for realizing load balancing between access devices in a wireless local area network according to claim 3, wherein the wireless access point in the access state continues to broadcast after the access state becomes disallowed. Beacon frames, but no longer respond to new probe requests until all access stations of the wireless access point go offline. 6. The method for realizing load balancing between access devices in a wireless local area network according to claim 3, wherein: when the station is only within the coverage of one wireless access point, and access is not allowed, if the When the wireless access point receives the inquiry request from the same station several times in a row, the wireless access point responds to ensure that the station can always access a wireless access point. 7. The method for realizing load balancing between access devices in a wireless local area network according to claim 3, characterized in that: the wireless access points communicate in a broadcast or multicast manner, and the wireless access points Encrypt the interaction information between them, and before exchanging priority information, the wireless access points authenticate each other and establish a communication link. Only when the state between the wireless access points is connected and authenticated, the wireless access point Priority information can only be passed between them. 8. The method for realizing load balancing between access devices in a wireless local area network according to claim 7, characterized in that: when a wireless access point joins the network or its status changes, the status notification message is used to actively notify the corresponding device through the wireless channel Neighboring wireless access points, so that adjacent wireless access points update their neighbor information, and the status notification message includes at least the media access control address, service set identification code and access priority of the wireless access point. 9. The method for realizing load balancing between access devices in a wireless local area network according to claim 7, characterized in that: when a wireless access point withdraws from the network, it actively notifies the adjacent wireless access points through the wireless channel through the exit message The ingress point enables the adjacent wireless access point to update its own neighbor information, and the exit message includes at least the media access control address and service set identification code of the wireless access point. 10. The method for realizing load balancing between access devices in a wireless local area network according to claim 3, characterized in that: when the wireless access point supports multiple service set identification codes, maintain neighbors for each service set identification code information, access priority and access status.

Images