Methods for implementing a dynamic frequency selection (dfs) and a temporary channel selection

Dynamic frequency selection and temporary channel selection

Related Application This application is a continuation (CIP) of US Patent Application No. 10/406049, entitled "Method for Implementing Dynamic Frequency Selection (DFS) Function for WLAN Equipment", filed by Atheror Communications on April 2, 2003. And merged with this by reference.

BACKGROUND OF THE INVENTION Field of the Invention The present invention relates to wireless local area network (WLAN) devices, and more particularly to a technique for enabling radar scanning in the local area network, identifying alternate channels for channel switching, and effectively changing the channel during radar detection events in the operating channel.

Related Technology Description Wireless Local Area Network (WLAN) devices operating in the 5GHz spectrum coexist with radar systems. Various adjustment standards (for example, drafted European Radiocommunication Standards Institute (ETSI) EN 301,893, version V1.2.1, published in July 200) require: 5GHz WLAN equipment with dynamic frequency selection (DFS) function. When there is a co-channel radar, the DFS function switches the operating channel, and can evenly expand the operation within a certain wide frequency band.

The adjustment criteria generally provide simple guidelines for radar detection and/or avoidance of companion channel radars. For example, the current European adjustment standards require: complete a 60-second scan of each allowed channel before the start of operation. Moreover, once a radar signal is detected on the current operating channel, the WLAN device (ie, access point or base station) must suspend communication within a short time frame (e.g., 0.2-1.0 seconds). Note that the pending revision of the European adjustment standard requires that the comprehensive transmission time of an access point or base station is limited to a total of 260 milliseconds from the moment the radar is detected. Similar rules have been proposed in other adjustment areas. For example, the drafting of the IEEE 802.11h specification (which adapts the 802.11a standard to European regulations) suggests that the base station responds to channel switching broadcast frames from the access point by stopping data transmission, thereby avoiding harmful interference to the companion channel radar system.

However, none of these adjustment standards provide specific implementation details. Therefore, various dedicated solutions have been proposed. For example, US Patent Application 10/XXX,XXX (McFarland hereinafter) filed by Atheror on December 6, 2001, titled "Radar Detection and Dynamic Frequency Selection for Wireless Local Area Networks", the application date is December 2001 Written by Atheros Communications Corporation on the 6th and incorporated by reference, this patent teaches how to effectively detect companion channel radars. In McFarland, signal pulses can be received as detection events. It can eliminate any detection events corresponding to network traffic. At this point, any events that cannot be eliminated can be checked to determine whether they correspond to radar signals. This check includes identifying the pulse repetition frequency, pulse period, or number of pulses in a predetermined period of time.

McFarland also pointed out a technique for stopping transmission between an access point and its associated base station based on radar detection. In this technology, the Node Coordination Function (PCF) can be used to control which base stations transmit and when they transmit. The node coordination function is provided in the IEEE 802.11 standard. In particular, a PCF beacon (beacon) sent by the access point announces the start of the polling period, at this time all base stations must wait until polled by that access point before sending. During the gap between the PCF beacon and the polling of the next base station, the access point may perform a radar detection loop. At the end of the radar detection cycle, the access point can perform normal PCF polling.

Other implementation technologies involving DFS are currently being developed. These techniques should include: defining special algorithms for performing start-up radar scans, determining acceptable alternate channels for possible channel switching, and effectively switching channels in the case of radar detection of operating channels. Preferably, these technologies should further include a mechanism so that legacy base stations, that is, those base stations that do not implement DFS functionality, can be easily controlled by the access point during channel switching operations and can often operate on new channels.

The summary of the invention The main parts of the various amendments and standards publish standards that define how radio equipment operates in certain frequency bands. The 5GHz frequency band is of particular importance for the revised authority, because the radar system operates in a part of this frequency band. The radar system can be used by military, aviation, meteorological, and other government agencies. Therefore, because of the importance of these radar systems, radio equipment operating in the 5GHz band must be able to detect radar and avoid any frequencies used by the radar system. Radar detection and avoidance are the main features of periodic dynamic frequency selection (DFS) capabilities. According to one aspect of the present invention, the following techniques are provided for: performing a radar start scan, determining an acceptable alternate channel for possible channel switching, and being able to switch channels effectively within a radar detection event of an operating channel. These technologies are conducive to meeting the current regulatory standards governing DFS, while minimizing network startup delays and interruptions to users during radar detection events.

Generally speaking, the time of the start-up procedure of the radio equipment is very critical. In particular, the base station (and the human user of the base station) should consider that the lengthy startup procedure is a product defect. The access point can be restarted to initiate a scan in response to an invalid channel availability, thereby significantly reducing the availability and utilization of the wireless network. Some adjustment areas require a relatively long period of time to test a channel that is not used by the radar. For example, European regulations require that each channel be scanned for 60 seconds to determine whether it is an idle channel for radar. Therefore, this demand must be met while also minimizing the time taken to start the program.

According to one aspect of the present invention, the access point assembles a short list (for example, one or two) of spare channels during initial startup, and stores this short list during normal operation. In the case that radar is detected on the current channel, the alternate channel facilitates fast and uninterrupted channel switching. Using this short channel table can eliminate another lengthy start-up procedure (for example, reaching 19 channels×60 seconds = 19 minutes), because the short channel table is pre-scanned by the radar for the required period (for example, 60 seconds).

During startup, the access point can quickly scan multiple channels of the radar. In one embodiment, normal operation can be started on the first channel where no radar is found. During normal operation, the access point or one or more base stations requested by the access point can perform a short background scan to identify other radar idle channels. The scanning process can continue until one or two spare radar channels are found. At this time, the radar free channel can be stored in the access point for use during future channel switching events.

Fortunately, the timing of these scans can be easily changed, thereby avoiding the short period of channel scans performed by the access point and the base station, and the radar on the standby channel sends radar bursts with timing that is exactly the same as the short period. Possibility. What is important is that multiple short background scans can be performed during normal operation on the selected backup channel in order to achieve an effective full range scan of the radar (for example, 60 seconds), as required by some adjustment standards.

Note that the radar can transmit effective energy in multiple channels adjacent to the current channel of the WLAN device (ie, on both sides of the lower and higher sides). Therefore, according to a feature of the present invention, a backup channel can be selected to reduce the possibility that the new channel is adjacent to the current channel in frequency. This selection process is also beneficial to meet the European demand for uniform expansion of operating channels. Moreover, because many legacy base stations still cannot operate in certain parts of the 5GHz spectrum, for example, the 5470MHz-5725MHz sub-band, the selection process should be weighted so that most or all of the current base stations related to the access point Can run on the new channel.

For example, in one embodiment of the present invention, the access point establishes a spare channel list by selecting a channel from each sub-band of 5GHz (for example, 5150-5250MHz, 5250-5350MHz, and 5470-5725MHz). It is important to note that the random selection of channels distributed in each sub-band (as opposed to the random selection of channels in the entire range of the channel) can generate a pseudo-random table of alternate channels, and it will be used to reduce the frequency of the spectrum. The information of the end part (ie, 5150-5350MHz) is weighted. Because many legacy base stations cannot operate in the 5470-5725MHz sub-band, minimizing the channel selection in this larger sub-band can increase the possibility of compatibility with such legacy devices.

Advantageously, in the 5150-5250MHz sub-band, it is rarely found that radars can operate in certain target areas (for example, Europe). Therefore, the revised technology in the United States and Europe requires these devices to detect radars in this sub-band. Moreover, it is also hoped that a relatively small number of radars will be found in the 5250-5350MHz sub-band. On the contrary, in many countries, it is hoped that the wider 5470-5725MHz sub-band can be used by commonly deployed weather radars. Therefore, this weighting also reduces the possibility of encountering radar on the candidate alternate channel.

According to another aspect of the present invention, the access point also takes into consideration the existence of other WLAN devices already operating in certain frequency bands. In one embodiment, the BSS idle channel table is combined from the allowed channel table in the current area. The access point can preferably select its spare channel from the BSS idle channel list. However, if the number of spare channels is too small, the access point can also select its spare channel from the list of allowed channels to maintain compliance with the channel extension rule.

When a radar is detected on the current channel of operation, the access point can select its alternate channel from the list of channels not used by the radar, that is, on the general frequency (that is, higher or lower) from the current channel where the radar is detected The farthest channel. In one embodiment, the channel switching process favors the selection of alternate channels in the 5150-5350 MHz sub-band. As noted above, the channel is less likely to contain radar energy or more likely to allow the operation of legacy base stations. For example, if the current channel is in the sub-band of 5250-5350MHz, and at least one spare channel is in the sub-channel of 5150-5250MH, then the access point can set the new channel to the lowest channel in the radar free channel list. Conversely, if the current channel is in the 5250-5350MHz sub-channel and there is no spare channel in the 5150-5250MHz sub-band, then the access point can set the new channel to the highest channel in the radar free channel list. Finally, if the current channel is in the 5470-5725MHz sub-band, the access point can set the new channel to the lowest channel in the radar free channel list. This channel switching operation combined with the alternate channel selection (ie, combined radar idle channel table) advantageously minimizes the network interruption to legacy and non-legacy base station equipment, and compared with completely random channel switching, it can be significantly Improve the operation of the wireless network.

Once the radar is detected on the current channel of operation, the access point broadcasts the channel switching frame to all base stations, thus ensuring that all base stations that can interpret this information will immediately stop data transmission. At this time, the access point has an additional control frame, that is, the revocation of the authentication frame will broadcast the frame to all base stations so that all base stations can understand that they immediately suspend sending data to the access point and return to the activated state.

Importantly, the de-authentication frame is sent after the designated channel switching time. Without this timing, most, if not all base stations suitable for 802.11h, can respond to the de-authentication frame by immediately stopping the communication with the access point. Therefore, it loses the need for no interruption to execute the channel switching process. The ability to switch channels. By sending a de-authentication frame at a specified time after the channel switching frame, before the access point sends the de-authentication frame, any receiving, understanding, and 802.11h suitable base station that accepts the channel switching frame can be from the current operating channel Remove. Moreover, any non-803.11h suitable (ie, legacy) base stations will quickly stop their operation, thereby reducing harmful interference to the companion channel radar, and reducing the time required to detect that the access point has left the channel and initiate the recovery process .

It also provides a way to perform startup operations for the access points in the adjustment area. Note that the adjustment area has a frequency spread index, and the access point can only use the channels allowed in the frequency spectrum to communicate. In this method, a list of allowed channels can be combined with the allowed channels. The BSS idle channel table can be combined with any allowed channel that contains the acceptable level of current WLAN transmission. Then, the idle channel table of the radar can be combined with the BSS idle channel table and the allowed channel table.

The first channel can be randomly selected from the free channel list of the radar. If the first channel is in the first sub-band and the adjustment area is exempt from scanning in the first sub-band, then the first channel can be designated as the current channel, and the standby channel can be selected from the radar idle channel in the first sub-band Select from the table, and the operation of the access point can start with the first channel.

On the other hand, if the first channel is not in the first sub-band, and the area is adjusted to avoid scanning in the first sub-band, then, except for any channel in the first sub-band, the radar scan can be in the radar free channel list. On each channel within. Similarly, if the first channel is in the first sub-band, and the adjustment area needs to be scanned in the first sub-band, then the radar scan can be performed on each channel in the radar free channel list. In both cases, any channel containing radar can be detected in the radar free channel table. The above steps can be repeated until a predetermined number of channels are included in the radar idle channel list. At this time, the first channel can be selected in the radar idle channel list, and the first channel can be used to start the normal operation of the access point.

In some adjusted areas, when the radar is detected, the access point can be helpful to select a new channel from the radar free channel list for operation without additional radar scanning. In other adjustment areas, such as the United States adjustment area, an access point should perform a 60-second radar scan immediately before operations on non-radar idle channels (ie, channels that require radar detection). Therefore, the radar scan performed during startup and refresh does not meet this criterion. The continuous 60-second radar scan from the radar detection of the current channel is not expected to cause the access point to interrupt the operation of its current associated base station. In other words, fast channel change (without losing the association with the base station) will be impossible, because the access point and the base station need to vacate the channel within 10 seconds, and therefore, if at this time, if you are still trying on the candidate channel For any type of 60-second scan, the access point cannot broadcast the fast channel change information to the base station.

Therefore, according to one aspect of the present invention, once the channel change operation is initiated by the radar detected on the current channel, if the selected new channel is in the radar exempt sub-band, then the access point can immediately restart operation on the new channel . In other words, regardless of the interpretation of the adjustment area for channel availability detection, a 60-second scan is not required to start operation in the radar exempt subband. Note that the new channel can be selected using the pseudo-random channel algorithm. And therefore it also satisfies the uniform expansion rule.

On the other hand, if the selected new channel is in the non-radar exempt sub-band, then the access point can select a temporary new channel in the radar exempt sub-band. In one embodiment, the temporary new channel may be randomly selected. In another embodiment, at startup, the access point can store information about channel traffic, and the access point can select the radar idle channel with the smallest amount of standby. In this temporary new channel, it can continue to operate normally until the access point has completed the comprehensive background radar scan of the previously selected new channel. If the radar is detected in the previous new channel, the access point can select another channel from the radar free channel list and perform another integrated background scan. If no radar is detected during the integrated background radar scan, the access point can quickly switch from this temporary new channel to the previous new channel. With this channel switching technology, it is beneficial for network users to not endure service interruption, because only fast channel changes and background scanning can meet the applicable revision indicators.

Brief Description of the Drawings Figure 1 depicts an exemplary process suitable for channel selection during the start/start of an access point; Figures 2A, 2B, and 2C describe a combination of the allowed channel table and the BSS idle channel table. An exemplary processing procedure of the radar idle channel list used; Figure 3A depicts an exemplary radar scanning processing procedure performed by the access point during normal operation; Figure 3B depicts the selection applicable to radars detected in the current channel One embodiment of the new channel; Figures 4A and 4B describe an exemplary channel selection technique that can be performed by the access point. In this channel switching technology, if the newly selected channel is not in the radar exempt sub-band, then the access point can temporarily perform the fast channel change operation of the access point and the relevant base station on the channel in the radar exempt sub-band, and Immediately start a comprehensive background radar scan on the selected new channel. After confirming that the newly selected channel is a channel not used by the radar, the access point performs the second fast channel change operation to switch to the newly selected channel.

Fig. 5 describes an embodiment of the channel change process in the base station. This channel change process allows the base station to maintain compatibility with the IEEE 802.11h standard, maintain compatibility with non-802.11h access points, and comply with mandatory European standards.

Figures 6A and 6B illustrate that the access point can request the relevant base station to assist in another exemplary start-up operation of the radar detection function.

Figure 7 depicts another exemplary process suitable for channel selection during the start/start of the access point.

Detailed Description of the Drawings The detailed description of the various aspects of the present invention is now applicable to perform start-up radar scanning; determine an acceptable alternate channel for channel switching; and facilitate effective channel change in the case of radar detection in the operating channel to satisfy The current management standard adjusted by DFS also minimizes network startup delays and interruptions to users during radar incidents.

Channel selection during the start/restart of the access point FIG. 1 describes an exemplary process of the alternate channel selection during the start/restart of the access point.

In step 101, the access point can search for the channels on which the access point will operate as permitted by the adjusted area. In one embodiment, the access point has access to a look-up table (LUT), which includes a known adjustment area and related allowed operating channels. If the adjustment area is Europe, then the allowed channels are within the following frequency ranges (also referred to as sub-bands in this article): 5150-5250MHz (4 channels), 5250-5350MHz (4 channels), and 5470-5725MHz (11 Channels). These allowed channels can be added to the allowed channel list. This article refers to this process as "combining" a list.

In step 102, the access point performs a startup scan for the basic service setup (BBS) on the allowed channel list. BBS is a group of IEEE 802.11h compatible devices that can operate as a fully connected wireless network. In one embodiment, it takes 200ms per channel to initiate the scan. Therefore, assuming that all sub-bands are allowed, this start-up scan will last no more than 200ms×19 channels=4 seconds. At this time, the access point can combine the BSS idle channel list of the allowed channels. These allowed channels are channels that are not currently non-transmitted or have an acceptable level of current WLAN transmission.

In step 103, the access point can use the processing procedure described with reference to FIG. 2 to assemble the radar idle channel table. In the embodiment described in FIGS. 2A and 2B, there are two or three candidate channels in the radar idle channel table. Note that at this time, these candidate channels have not been tested for radar.

In step 104, the access point can perform a radar scan on each candidate channel in the radar free channel list. In one embodiment, the radar scan complies with the current European standard of 60 seconds per channel. After the radar scan operation, the access point can delete any channel where the radar signal is detected from the radar free channel list. Typically, at this time, the radar idle channel table contains 0, 1, 2, or 3 channels.

In step 105, if the radar free channel list contains two or more channels, then in step 106, the access point can randomly select a channel from the radar free channel list and set that channel as the currently operating channel. At this time, the access point can delete the current channel from the radar idle channel table, thereby leaving one or two spare channels in this table. The access point can keep the radar idle channel list in case the radar is detected in the current channel as a backup. In step 107, the access point can start pointing and perform normal operations on the current channel. In one embodiment, the BSS idle channel table can be discarded after step 106 to save the available storage area.

On the other hand, referring back to step 105, if the radar free channel table contains less than two channels (ie, 0 or 1), then the access point may send a control message indicating that a sufficient number of channels without radar have been found. Then, the process returns to step 101 to repeat the start operation.

For the adjustment area that forces the revision of the technical requirements, the WLAN device must no longer occupy the channel where the radar is detected (for example, usually at least 30 minutes after the radar is detected), so as to achieve an additional protection. In this case, the WLAN device can store the number of channels on which the radar is detected along with the time stamp of the radar event related to that channel. In one embodiment, this information can be stored in non-volatile memory. During normal operation, this time mark can be periodically refreshed to count down the required period (for example, 30 minutes), during which the non-reoccupied mark can be removed. Once the WLAN device is restarted, the candidate radar free channel table (Figure 1, step 103) can be compared with the number of channels stored in the non-volatile memory, and the appropriate channel can be removed. Once the WLAN device is restarted for the second time thereafter, a separate lifetime flag (the lifetime flag can also be stored in non-volatile memory and decremented) can be used to refresh this non-reoccupied channel table. The number of restarts since storing the non-reoccupied channel data is counted.

An alternative implementation of the non-reoccupied table can be implemented in the following way, without the need for a real-time clock or time stamps for each channel in the list. The list of non-occupied channels can be compiled as described above, and the new channel will be stored in the non-volatile memory immediately after the radar is detected. The entire list of non-reoccupied channels can only be reset after 30 minutes from the time when any channel is added to the list. If the device is reset or restarted due to detection of radar or other conditions, then the existing non-reoccupied channel list is checked at the start and a new 30-minute timer is started. Only after a continuous 30-minute period has elapsed (during this time, no new channels have been added to the table), the list stored in the non-volatile memory will be cleared. In this way, the time stamp or real-time clock of each channel cannot be used. Note that a period longer than 30 minutes may elapse before some suitable channels are removed from this list. Because of the effective number of total channels and sub-bands that are not subject to radar detection technology requirements (and therefore are generally available for operation, regardless of radar activity in the area), this removal technology may affect the operation and/or availability of wireless networks The impact is minimal. In order to fully realize the benefits of the short standby channel list (for example, quick start and non-disruptive refresh of the standby channel), it is considered that the radar free channel list should be combined in step 103. In particular, the choice of alternate channels should reduce the risk of channel switching operations causing the access point to transfer to a channel where the radar is still operating. The revised index has not yet been indicated: 5GHz WLAN equipment can take special measures to avoid moving to a new channel still occupied by the radar after the current channel detects the radar, in addition to the technical requirement: during the initial start-up scan, the radar that previously detected the new channel.

The important thing is that radar can work in a wider transmission bandwidth than WLAN equipment. Moreover, some radars, called frequency agile radars, can switch at will on multiple adjacent frequencies within the spectrum wave. In addition, other radars, such as portable, airborne, or even fixed radars may operate intermittently or with lengthy and random directional scan sequences (e.g., weather radars). In addition, because a large number of out-of-band parasitic emissions near the radar enter the WLAN receiver with limited out-of-band suppression, the WLAN device located near the radar will receive the effective energy in the channel adjacent to the current operating channel. In this way, unfortunately, it is entirely possible in the real world to transfer to a channel where the radar energy still exists.

Therefore, the traditional spare channel table, even if it needs to be refreshed every time (described in Figure 3A), cannot protect the access point from performing channel switching to the spare channel. This spare information does not contain radar, nor is it close to radar in frequency. The unsuitable channel. Such channel switching should also require other channel switching with interruption capabilities related to its network operation. The revised technical requirement to extend the operation of the WLAN device over the entire range of the available spectrum makes this problem more complicated.

According to a feature of the present invention, the selection of the spare channel can be weighted to provide a wide range of channel frequencies in the 5GHz spectrum while maintaining the randomness required to implement the extended operation of the WLAN device. In addition, this weighting increases the possibility that the legacy base station can still operate when the access point performs channel switching.

Figures 2A, 2B, and 2C describe an exemplary weighting process for the combined radar idle channel table. In one embodiment, the access point can combine the radar idle channel table with the allowed channel table and the BSS idle channel table. Preferably, two or three channels weighted toward the lower end of the spectrum can be selected, thereby significantly improving the support of legacy devices.

In step 201, if the allowed channel table contains channels with frequencies between 5150-5250MHz, 5250-5350MHz and 5470-5725MHz, then in step 204, the access point determines whether the BSS idle channel table contains less than two channels. In other words, if two or more channels currently do not contain transmissions (or acceptable levels) from other basic service settings, then these channels are preferably available if radar is detected on the current channel As a spare channel. If the BSS idle channel table contains no less than two channels, that is, two or more channels are found in the BSS idle channel table, then in step 206, the process can determine whether the BSS idle channel table is in each of 5150-5250MHz, At least one channel is contained in the 5250-5350MHz and 5470-5725MHz sub-bands. If the BSS idle channel table contains at least one channel in each of the 5150-5250MHz, 5250-5350MHz and 5470-5725MHz sub-bands, then in step 207, the access point pseudo-randomly selects three channels from the BSS idle channel table. In one embodiment, a channel can be selected from each of 5150-5250MHz, 5250-5350MHz, and 5470-5725MHz (hence, the selection is considered "pseudo-random").

On the other hand, in each of the 5150-5250MHz, 5250-5250MHz, and 5470-5725MHz sub-bands, if the BSS idle channel table contains less than two channels or if the BSS idle channel table does not contain at least one channel, then in step 207. The access point is pseudo-randomly selecting three channels from the allowed channel list. In one embodiment, one channel can be selected from various sub-bands, namely, 5150-5250 MHz, 5250-5350 MHz, and 5470-5725 MHz.

In step 202, if the allowed channel list contains channels whose frequencies are only between 5150-5250MHz and 5250-5350MHz, then in step 208, the access point determines that the BSS idle channel list contains less than two channels. If not, then in step 210, the access point determines whether the BSS idle channel table contains at least two channels in the 5150-5250 MHz sub-band and at least one channel in the 5250-5350 MHz sub-band. If the BSS idle channel table contains at least two channels in the 5150-5250MHz sub-band and at least one channel in the 5250-5350MHz sub-band, then in step 211, the access point pseudo-randomly selects three channels from the BSS idle channel table . In one embodiment, two channels can be selected from the 5150-5250 MHz sub-band, and one channel can be selected from the 5250-5350 MHz sub-band.

On the other hand, if the BSS idle channel table contains less than two channels or the BSS idle channel table does not contain at least two channels in the 5150-5250MHz sub-band and one channel in the 5250-5350MHz sub-band, then in step 209, access Click to select three channels pseudo-randomly from the list of allowed channels. In one embodiment, two channels can be selected from the frequency range 5150-5250, and one channel can be selected from the frequency range 5250-5350MHz.

In step 203, if the allowed channel list contains channels whose frequencies are only 5150-5250 MHz, then in step 212, the access point determines whether the BSS idle channel list contains less than two channels. If not, then in step 214, the access point randomly selects two channels from the BSS idle channel list. On the other hand, if the BSS idle channel list contains less than two channels, then in step 213, the access point randomly selects two channels from the allowed channel list. In the case that the allowed channel list contains frequencies only within 5150-5250MHz, since there is less possibility of radar signals in this sub-band, the choice of only two channels (step 211/212) is appropriate and ideal of. Moreover, reducing the number of channels in the radar idle channel table from three to two is beneficial to reduce the start-up time and maintain the expansion of the required channels.

In one embodiment, in the case where a channel must be selected from the allowed channel list due to a shortage of channels in the BSS idle channel list, the above steps can be modified to take into account the relative level of BSS activity on the occupied channel. In this embodiment, the access point can build a history of received signal strength (RSSI) levels and activity levels from BSS signals received in each allowed channel. Those channels with the combined RSSI and active traffic level that constitute the lowest value can be combined in the BSS idle channel table. In this way, the BSS idle channel table can also be used to combine the radar idle channel table, even when there is some BSS activity on most or all of the allowed channels. This process can advantageously reduce the possibility of the access point operating on a channel with severe co-channel BSS traffic. In another embodiment, a single absolute RSSI threshold can be used instead of the numerical history. In this way, it is found that the allowed channels containing the minimum and/or acceptable level of BSS activity can be combined with the BSS idle channel table. In other words, BSS traffic below the minimum RSSI value will have no actual impact on the operation of the access point, and thus can be safely ignored.

Radar Scan Performed by an Access Point in Normal Operation FIGS. 3A and 3B describe an embodiment of a radar scan that can be performed by an access point during normal operation. According to one aspect of the present invention, an ongoing radar scan of the current channel is required. In each step 301, this scan is performed on the current channel during normal operation to ensure compliance with the revised technical requirements. If a radar is detected on the current channel, as determined in step 302, then the processing is input to step 311, which can be explained with reference to FIG. 3B.

If no radar is detected in the current channel, then in step 303, the access point determines whether the refresh time has been reached. This refresh time is called a certain time period, during which the radar in the spare channel in the radar idle channel list should be re-detected. In particular, some revised technical specifications require the WLAN device to rescan any channel in the radar idle channel list, otherwise the access point cannot change to the alternate channel during the channel switching operation. The refresh cycle is usually a 24-hour cycle. Various European technologies require that during the refresh cycle, the access point must perform another 60-second scan of the radar on each spare channel. If the refresh time is not reached in step 303, the access point returns to step 301 and continues to perform periodic radar scanning on the current channel. If the refresh time is reached, the access point suspends normal operation, and in step 304, rescans the spare channels in the radar idle channel table.

Preferably, the refresh scan can be performed at a specified time, instead of trying to refresh at a predetermined time that is difficult for channel switching due to radar detection on the current channel. In particular, the operations of sending channel switching broadcast and channel suspension can be performed in a much shorter time frame of 60 seconds required to perform each spare channel scan. Therefore, in order to save time during channel switching and reduce network interruption, refresh scanning should be performed during normal operation. However, if the wireless network operation is interrupted for 60 seconds, even refreshing one of the spare channels will cause all base stations to lose communication with higher-layer protocols and applications, and may lose data or other unacceptable interruptions.

Therefore, an exemplary refresh operation includes: the access point uses short (for example, 2-5 seconds) periodic background scanning of each backup channel to meet the 60-second comprehensive scanning time, thereby meeting the revised technical requirements, while making The impact on users and the network is reduced to a minimum. In another embodiment, one or more related base stations may be used to perform short periodic background scans of various backup channels (see, for example, Fig. 6B). According to a feature of the present invention, identifying a short spare list with 1 or 2 channels can advantageously ensure that the refresh of the spare channel is performed significantly faster than refreshing a larger list. In addition, the use of short-period refresh can dynamically reduce or even eliminate the interruption of network operation.

If during the refresh scan of each step 305, radar is detected on one or more spare channels, then in step 306, delete the channel with radar from the radar free channel list, and once the radar is detected, give up immediately Complete all comprehensive scans. In another embodiment, the access point attempts to revalidate the deleted channel within a certain predetermined interval (e.g., 30 minutes). If the radar is not found during the revalidation period, the channel can be added to the radar free channel list. If at least one spare channel remains in the radar free channel list (as determined in step 307), or if no radar is detected in step 305, then in step 308, the access point can resume its normal operation on the current channel and Step 309, reset the refresh time. At this time, the access point returns to step 301 to scan the radar on the current channel.

If no channel remains in the radar free channel list (step 307), then in step 310, the access point is separated from all base stations. At this time, for a start (or restart) scan, the access point returns to step 101. In another embodiment, in step 320, the access point performs background scanning of any allowed channels (for example, periodic scanning in 200ms increments) until a radar idle channel is identified and added to the radar idle channel list. Then, the process can go to step 308. In one embodiment, if no radar idle channel is found, then the process goes to step 310, that is, separation of all base stations.

Figure 3B depicts an exemplary selection process for a new channel after the radar is detected on the current channel. In step 311 of this process, the access point stops the current transmission on the channel and rejects the new parallel request. At this time, the access point can respond based on the sub-band (for example, 5150-5250MHz, 5250-5350MHz, and 5470-5725MHz) without characterizing the current channel characteristics.

In particular, as previously noted, some adjustment areas may not require radar scanning in a few predetermined sub-bands even if there is radar. For example, the adjustment area in the United States and Europe generally does not require radar scanning in the 5150-5250 MHz sub-band. The United States does not require radar scanning in the 5725-5850MHz sub-band. In this paper, the sub-bands that do not require radar scanning are called radar exempt sub-bands. Therefore, in step 302, if a radar is detected on the channel, then the current transmission can be performed in the 525-5350 MHz sub-band or in the 5470-5725 MHz sub-band.

Because the radar in the close range enables the WLAN to receive effective energy on multiple channels adjacent to the current channel, setting the new channel at the frequency as far away as possible from the current channel can help reduce the possibility that the new channel contains radar . For example, if the current channel is in the sub-band of 5250-5350 MHz, and at least one channel in the radar free channel table is in the sub-band of 5150-5250 MHz, as determined in step 314, then in step 315, the access point can advantageously Set its new channel to the lowest channel in the radar free channel list. On the contrary, if the current channel is in the 5250-5350MHz sub-band, and no channel in the radar free channel list is in the 5150-5250MHz sub-band, as determined in step 316, then in step 317, the access point can transfer its new channel Set to the highest channel in the radar channel table. Finally, if the current channel is in the 5470-5725MHz sub-band, as determined in step 318, then in step 319, the access point can set its new channel to the lowest channel in the radar free channel list. After setting up a new channel, the access point can switch to the channel switching technique described with reference to Figures 4A and 4B.

Temporary channel selection ensures the continuity of transmission. In certain adjusted areas, when a radar is detected, the access point can advantageously select a channel from the radar free channel list for operation without additional radar scanning. US Patent Application No. 10/406049 discusses this type of channel switching technique in more detail. In other adjusted areas, for example, the United States adjusted area, the access point should perform a 60-second radar scan, and the moment before the operation on the non-radar idle channel (ie, the channel that needs radar detection). Therefore, the radar scan performed during startup (step 104) and refresh (step 304) may not meet this technical requirement. As previously noted, it is not expected that continuous radar scans for 60 seconds will cause the access point to interrupt the operation of its current associated base station.

Therefore, referring to FIG. 4A, if the selected new channel is in the radar exempt sub-band, for example, 5150-5250MHz or 5725-5850MHz, as determined in step 401, the access point immediately restarts operation on this new channel. In other words, regardless of the interpretation of the channel availability detection in the adjustment area, the 60-second scan does not need to start running in the radar exempt sub-band. Note that the new channel is selected using the pseudo-random channel algorithm and therefore also satisfies the uniform expansion rule.

If the selected new channel is in a non-radar exempt sub-band, step 402 may select a temporary new channel in the radar exempt sub-band (for example, 5150-5250 MHz or 5725-5850 MHz). In one embodiment, the temporary new channel may be selected randomly. In another embodiment, at startup, the access point can store information about channel traffic, and step 402 can include selecting a radar idle channel with the smallest traffic. Normal operation can continue on this temporary new channel until the access point has completed the radar scan of the previously selected new channel (ie, the new channel selected by one of steps 315, 316, and 317), as This is described below with reference to FIG. 4B.

Note that in one embodiment, if the adjustment area does not require frequency extension, the temporary new channel can be used as a new channel, and the switching step described in FIG. 4B does not need to be performed.

Operation of the access point during the channel change After detecting the radar, the access point can send control and/or management frames and reach the maximum movement time (MaxMoveTime) as provided by the current revision (for example, 10.24 seconds). However, in order to reduce the impact on potentially sensitive applications, the access point should perform channel changes as fast as possible (ie, on the order of less than 3 seconds).

In one embodiment, the access point can be quickly separated from non-802.11h-compatible base stations because they cannot act on the channel change frames broadcast by the access point. The access point can advantageously maintain the parallel connection with most 802.11h compatible base stations on the new channel by referring to the channel switching technology described in steps 403-410. In particular, in step 403, the access point can broadcast a channel change frame to its associated base station. What is important is that 802.11h compatible base stations, once they receive a channel change frame, the channel switching time specified in the channel change frame should be able to respond by stopping normal data transmission and changing their transmission channel to a new channel. However, some 802.11h compatible base stations may be dormant or have lost channel change frames (for example, due to interference from ongoing co-channel radar pulses or system failure).

In step 404, the access point determines whether the channel change time has expired. If the channel change time has not expired, then in step 405, the access point can repeatedly broadcast the channel change frame to reach MaxMoveTime (maximum movement time), but usually a duration of less than 3 seconds. If the channel change time zone expires, then in step 406, the access point broadcasts a de-authentication frame to any other base stations on the old channel. The broadcast de-authentication frame is sent only after the channel change time expires. The de-authentication frame prohibits all base stations and causes separation from the access point. This timing prevents the 802.11h-compliant base stations from receiving the de-authentication frame (otherwise, this will cause these base stations to immediately disconnect from the access point instead of performing a smaller split channel change procedure). Note that any resulting separation will not conflict with the IEEE 802.11h standard or revised technical requirements. Advantageously, once contact with the access point is lost, a base station can be programmed to stop data transmission within 1 second, thereby limiting any impact on any radar system, if any. Note that these separated base stations can quickly re-associate with the access point on the new channel (discussed with reference to step 409) or with the new access point on another channel.

In step 407, the access point transfers to the new channel and starts to send beacons on this new channel. In step 408, the access point can resume communication with the 802.11h compatible base station that successfully transferred to the new channel. In step 409, the access point can be associated with any base station that became separated during the channel switch and/or any new base station associated with the request. In step 410, the access point continues its normal operation.

Step 420 determines whether the new channel being operated is temporary. If it is not, then the process returns to step 301 to perform a radar scan on the channel (if necessary). If the new channel is temporary, then step 421 can perform background channel detection on the previous channel selected by the pseudo-random, weighted algorithm (e.g., step 317 or 319). Obviously, this background channel detection can use the comprehensive total of the scanning time to meet the 60-second scanning technical requirement. This background channel detection can be performed by the access point alone or with help from any available base station, which affirmatively responds to a request to perform part of the 60 second scan.

If the 60-second background scan shows that there is no radar, as determined in step 422, then step 425 can remove the temporary new channel designation and return to step 403 to start the second fast channel with the previous new channel change. In this way, after performing steps 402-408, the operation will restart on the latest scan channel. Advantageously, because only the use of fast channel changes and background scanning can meet the applicable revised technical requirements, network users will experience uninterrupted service.

Because this channel has undergone continuous 60-second radar scans during the start of the access point, and perhaps even a combined 60-second background scan during the refresh cycle (see step 304), there is a high possibility: in step 422 A background scan performed will find that there is no radar. However, if a radar is detected, then step 423 marks the non-occupied channel of the channel and selects another channel from the radar free channel list. At this time, step 424 performs a comprehensive 60-second radar scan on this new channel. Then, the process returns to step 422. Note that until the channel passes this background scan, the access point can conveniently continue to operate normally on the temporary channel.

DFS implementation for WLAN base stations Figure 5 describes an exemplary channel change process that can be performed in a WLAN base station. This channel change process advantageously allows the base station to maintain compatibility with the IEEE 802.11h standard, maintain compatibility with non-802.11h access points, and also comply with mandatory European technical requirements.

In step 501, the base station determines whether it has received a channel change or revocation authentication frame from an access point. If the channel change frame has been received, then in step 502, the base station stops all transmissions and responds with an acknowledgment. If the base station supports the new channel, as determined in step 503, then in step 504, the base station sets up its operation for the new channel. In one embodiment, the re-authentication of the base station performed in step 504 is not required.

At this time, the base station is ready to listen to the signal (i.e., beacon) from the access point on the new channel. In one embodiment, the suitable time for waiting to listen to the access point in step 505 includes: the time required for the access point to change the channel plus the time required for the access point to send a predetermined number of beacons (for example, 7 beacons) . Note that during this channel change with this timing or different timings, various timers in the base station can be frozen.

If in step 505, the base station hears the access point on the new channel, then in step 507, the base station can resume operation on the new channel. On the other hand, if the base station does not hear the access point on the new channel, then in step 506, the base station starts a normal passive scanning sequence on all allowed channels. If the base station receives a revocation authentication frame from an access point, then in step 509, the base station immediately stops all transmissions. This interruption generally does not precede the base station's confirmation of the access point. Note that non-802.11h access points (ie, legacy access points) respond to radar detection or when the base station loses the channel change frame from 802.11h access, it can send a de-authentication frame (recall that the access point sent before leaving the channel) Revocation of authentication frame). Once the de-authentication frame is received, in step 506, the base station starts the scanning sequence on the allowed channel.

The base station can also be configured with a relatively short lost beacon timer (ie, 1 second beacon interval × 2 lost beacons) to further protect the base station from receiving neither the de-authentication frame nor the channel change Radar in the case of a frame. In this case, the base station will stop all transmissions within 2 seconds after the access point vacates the current channel due to a radar event. A shorter value for lost beacons can result in a degradation of the base station performance of the edge link to the access point.

Note that if the base station fails to detect the radar, the base station can be configured to prohibit the use of any 5GHz channel for special communication, and in step 506, force passive scanning. This configuration ensures that the base station will not transmit on any 5GHz channel until after the access point has checked the radar and started to send beacons. In particular, the revised technical requirements allow the base station not to have the function of detecting radar, so as to remain under the control of the'master station' equipment, that is, the access point responsible for checking the radar before and during the use of any channel. The embodiment described in FIG. 5 allows base station equipment that does not have a radar detection function to comply with the revised technical requirements, while maintaining effective startup and ongoing operations when operating in an area that requires DFS. On the other hand, if the base station has a radar detection function, then in step 506, the base station may be configured to allow special communication and perform effective scanning, although such a base station does not require this. Therefore, even a base station capable of detecting radar can be configured to perform passive scanning in step 506.

Base station assistance for radar detection functions In one embodiment, the access point can further reduce startup time by obtaining the help of one or more relevant base stations that are capable of performing radar detection functions. 6A and 6B describe an exemplary start-up operation in which the base station can assist the access point in the radar detection function. In one embodiment, the access point can use steps 101-103, as described with reference to FIG. 1, for combining the radar idle channel table.

Subsequently, in step 601, the access point can perform a 60-second radar scan on a channel selected at random in the radar free channel list. This scan can continue until the access point finds the first channel without radar. Any scan channel containing radar will be deleted from the radar free channel list. In step 602, the access point can select the first radar idle channel as the current channel, delete this channel from the radar idle channel list, and in step 603, start sending beacons on this channel. Note that because the spare channel has not been selected, the startup time can be significantly reduced compared to the processing procedure described with reference to FIG. 1.

Next, in step 604, the access point allows the base station to authenticate and associate. If at least one new base station is already associated with the access point, as determined in step 605, before sending or receiving the data frame, in step 606, the access point immediately sends a basic measurement request frame to the associated base station. In one embodiment, the access point can send the frame to the base station requesting the association for the first time. This signal requires the base station to scan the radar on a candidate channel in the radar free channel list. The access point may request that the scan be limited to a short duration (e.g., 2-5 seconds). In one embodiment, the access point can identify the channel to be scanned by the base station. For example, the channel identification process may include steps substantially similar to steps 312-319 (ie, instead of setting the standby channel as the current channel, the radar of the candidate channel should be scanned).

However, the base station may be unable or unwilling to accept and perform the measurement requirements on the designated channel, as determined in step 607. If the base station responds that it does not perform radar measurements (e.g., rejects, does not accept, or does not respond to the request at all), then in step 605, the access point determines whether there is another newly related base station. If it exists, then the process continues at step 606. If there is not another newly related base station that will perform the measurement, then in step 609, the access point will indicate to any related base station that the communication will be delayed for a predetermined period of time, and at the same time the access point will switch to the alternative channel by itself, and perform short-term Radar scanning.

During the normal operation of step 609, in order to minimize the waiting time and interruption of the data processing process, especially when many base stations need to access the network, the access point scanning period should be short enough to ensure continued network operation, such as 100 to 200 ms. In one embodiment, the point coordination function (PCF) of the access point combined with the appropriate contention-free period value can notify the base station to interrupt the transmission in this period (plus the access point disconnects and returns to the current channel of operation). Short buffer time), resulting in a long NAV value in the base station. Note that PCF, provided by the IEEE 802.11h standard, allows the access point to control the transmission timing from its related base station. In particular, by using PCF, the access point will poll each base station for broadcasting: and the base station will only respond after being polled. The NAV value indicates the minimum time that the base station must wait until it attempts to communicate with the access point again. After scanning, the access point returns to the current channel of operation, resets NAV and PCF, and continues to listen to the radar on the current channel. After approximately the same time period for scanning, for example, 100 to 200 ms, the access point can repeat its scanning on that spare channel. Assuming a 200ms scan/reset time, the access point will repeat the process 300 times to meet the revised technical requirements for the 60-second scan of the standby channel.

If the candidate backup channel is idle by the radar, the access point can set the refresh time of the backup channel. On the other hand, if the candidate alternate channel contains radar, the access point can delete the channel from the radar free channel list. Note that if the access point detects the radar on the current channel before it has identified any of the alternate channels in the radar free channel list from the candidate alternate channels, then the access point usually restarts (i.e., returns to Figure 1 or Figure 6A Step 101).

In step 610, the access point can determine from the radar idle channel table whether another candidate spare channel should be scanned (ie, if the number of spare channels not used by the radar is less than a predetermined number, such as 2). If not, then normal WLAN operation can be started in step 612. If so, then in step 611, the access point can return to the current channel, send a beacon, and listen to any other base stations requesting association. Then the processing procedure returns to step 605.

Note that even if no base station attempts to associate after startup, this process should follow. In this case, after determining in step 605 that there is no associated channel available, the access point should directly transfer to step 609, instead of performing short-term detection for radars on different channels, and sending beacons running on the current channel Until a full scan has been performed.

Referring back to step 607, if the newly associated base station responds that it will perform radar, then in step 613, the base station switches to performing radar measurement. Note that if the base station detects a radar, the access point can delete the channel from the radar free channel list, and can use another unscanned candidate alternate channel from the radar free channel list in the next scan. In step 614, the access point can determine whether another backup channel is needed. If it is not required, then normal operation starts at step 612. If necessary, then in step 605, the access point determines whether another newly associated channel is available. The steps described with reference to FIG. 6B can be repeated until a full scan has been performed for each spare channel.

Assuming that multiple associated base stations are available and can perform radar measurements, a scan of the desired number of channels is quickly performed in the radar idle channel table. For example, an access point may request an associated base station to scan one channel, and may also request another base station to scan another channel. In this way, assuming that a full scan takes 60 seconds, if the two base stations that can perform measurements are immediately associated with the access point once they start the access point operation on the initial channel, then within 60-120 seconds, the desired one can be found Two spare channels are added to the radar free channel list. Note that if the two scans can be performed in parallel at the same time, and only an additional spare channel is required, then the first channel that is found to have no radar is designated as the spare channel (then thereafter, the access point can immediately interrupt the scan of the second channel) ).

Advantageously, by requesting the base station to perform a radar scan, the startup of the WLAN network can be several minutes faster than when the access point scans 3 or more channels by itself during startup. Note that if available, the base station is allowed to associate after only one channel is found to be an idle channel of the radar, thereby ensuring that the communication between the access point and the base station starts immediately when it completes its scanning cycle. Moreover, because the associated base station is not allowed to start sending or receiving data, until the base station indicates the following: it cannot perform measurement, has performed a full scan of a channel in the radar free channel list, or provides the expected number of spare channels. The base station will not start any applications that may be interrupted.

In an embodiment of step 613, the access point may repeatedly request a base station to perform short scans (for example, 2-5 seconds) on an alternate channel (using basic measurement requests), until these short scans cumulatively meet the revised radar scan technique Requirements (e.g. 60 seconds on each spare channel). In this way, during normal operation and assuming that the predetermined expected number of spare channels has not been scanned, the access point can continue to request: each newly associated base station performs a short on one of the unscanned candidate channels in the radar free channel list. scanning. Importantly, by obtaining the help of any newly associated base station to perform the radar scan function, the start-up delay of any new base station can be limited to the requested cumulative scan time (e.g., 60 seconds). In one embodiment, if the base station is re-associated with the access point, the access point no longer makes this request, thereby avoiding interruption of queued data reception or interruption of any latency-sensitive applications running on the base station.

Other Embodiments Although the illustrative embodiments have been described in detail herein with reference to the accompanying drawings, it should be understood that the present invention is not limited to these precise embodiments. They are not intended to be exhaustive or to limit the invention to the precise format disclosed. As such, many modifications and changes are obvious to skilled professionals.

For example, the scan time related to the access point and the base station is only exemplary and not restrictive. Other embodiments of the present invention include different (e.g., longer or shorter) scan times depending on device systems and/or policy considerations. In particular, although DFS has been described in detail with reference to European revision requirements, other countries will soon announce DFS rules. For example, certain organizations, such as the International Telecommunication Union (currently including 144 member states around the world), are working around the world to coordinate the 5GHz spectrum. As part of this coordination work, many regional regions in the world will adopt rules that require DFS technology. Although not all regional areas will specify the same implementation details.

In one embodiment of the present invention, instead of being limited to the radar scanning of one channel (see FIG. 6B), the base station equipment can be periodically requested to scan radars on multiple channels. For example, the access point can implement some other defined scan period and sequence. During these defined scanning periods, a series of base stations with this function can search for radars in parallel across multiple channels, which can significantly change the performance of the access point.

In another embodiment of the present invention, instead of providing refresh time for all spare channels, each spare channel may have its own refresh time.

In another embodiment shown in FIG. 7, the startup process 700 can consider whether the adjusted area can exempt the WLAN device from detecting the radar in the 5150-5250 MHz sub-band. If it is exempt, the WLAN device can adopt a quick start procedure, which also meets the technical requirements of entrusted management extension. In one embodiment, the process 700 can execute steps 101-103 (described with reference to FIG. 1) after being called by the main program.

In step 701, a channel can be randomly selected from the radar idle channel list. Note that this random selection needs to be able to maintain compatibility with the extended rules (ie, prevent the WLAN device from selecting channels within 5150-5250 MHz for each start-up time). If the selected channel is in the 5150-5250 MHz sub-band, and the adjustment area is exempt from scanning the channels in the sub-band, then in step 703, a quick start procedure can be performed. In this quick start procedure, the processing 700 can easily bypass all start radar scans (generally, it will last 60 seconds or more). Importantly, because the adjustment area that can avoid scanning in the 5150-5250MHz sub-band does not allow radar to be in this sub-band, this bypass is acceptable. Therefore, if the selected channel is in the 5150-5250MHz sub-band, then the selected channel has a fairly high probability of not carrying radar, and the selected channel can be designated as the current operating channel.

In one embodiment, in step 703, the spare channel may also be randomly selected from the 5150-5250 MHz sub-band. Advantageously, because the spare channel resides in the 5150-5250 MHz sub-band, the need for radar scanning (generally lasting 60 seconds or longer) for selecting the spare channel can also be eliminated. Note that selecting a spare channel from the 5150-5250MHz sub-band (not selected randomly from all available channels) will not conflict with the expansion rule, because the use of the spare channel is practically impossible.

In other words, assuming normal operation, the alternate channel will not be called, because the radar is usually not detected in the current channel. In this way, the selection of the alternate channel can be performed in step 703, taking into account the rare event of radar detected on the current channel, although it is prohibited in this sub-band. After the backup channel is selected, normal operation on the current channel can be started in step 107.

If the current channel is not the 5150-5250MHz sub-band, or if the adjusted area cannot be exempt from scanning in this sub-band, then in step 704, start radar scanning can be performed on each channel in the radar free channel list. Note that if the adjusted area is exempt from scanning, the scanning of any channel in the radar idle channel table on the 5150-5250MHz sub-band can be skipped. Subsequently, steps 105-108 can be performed as described with reference to FIG. 1. Therefore, the scope of the present invention can be defined by the following claims and their equivalents.