MEASURING CHANNEL PERFORMANCE IN WIRELESS LOCAL AREA NETWORKS

Abstract

There is herein disclosed a method of measuring the performance of a first communication channel at a first access point in a WLAN, in which the first access point is associated to a client device so that the first access point can send data to, and/or receive data from, the client device on a second communication channel, the method including disassociating the first access point to the client device and associating a second access point to the client device; switching the working channel of the first access point from the second communication channel to the first communication channel; and making one or more performance measurements in respect of the first communication channel at the first access point.

Claims

1. A method of measuring performance of a first communication channel at a first access point in a wireless local area network (WLAN), in which the first access point is associated to a client device so that the first access point can send data to, or receive data from, the client device on a second communication channel, the method comprising: disassociating the first access point from the client device and associating a second access point to the client device; switching a working channel of the first access point from the second communication channel to the first communication channel; and making one or more performance measurements in respect of the first communication channel at the first access point.

2. The method as claimed in claim 1, further comprising determining the working channel for the first access point.

3. The method as claimed in claim 2, wherein determining the working channel for the first access point takes the one or more performance measurements into account.

4. The method as claimed in claim 1, further comprising identifying the first communication channel.

5. The method as claimed in claim 4, further comprising determining which channel of the first access point has provided the least recent performance data.

6. The method as claimed in claim 1, wherein the one or more performance measurements include measurements of noise on a channel.

7. The method as claimed in claim 1, wherein the one or more performance measurements include measurements of a contention level on a channel.

8. The method as claimed in claim 1, the method further comprising determining an appropriate access point in the WLAN to use as the second access point.

9. The method as claimed in claim 8, further comprising determining which access points in the WLAN are capable of communicating with the client device.

10. The method as claimed in claim 8, further comprising determining which access point in the WLAN is most capable of communicating with the client device.

11. The method as claimed in claim 1, wherein communicating between the second access point and the client device comprises the client device sending data to the second access point, the data being intended for transmission beyond the second access point.

12. The method as claimed in claim 1, the method further comprising switching the working channel of the first access point back from the first communication channel to the second communication channel.

13. The method as claimed in claim 1, the method further comprising making performance measurements on the first communication channel at a proxy access point.

14. A non-transitory computer-readable storage medium storing machine-readable instructions for, when loaded on a computer and executed thereby, performing the method as claimed in claim 1.

15. A wireless local area network (WLAN) comprising: a first access point and a second access point, the first access point being configured to send data to, or receive data from, a client device on a second communication channel, the second access point being configured to send data to, or receive data from, the client device; means for switching a working channel of the first access point from the second communication channel to the first communication channel; and means for making performance measurements in respect of the first communication channel.

Description

BRIEF DESCRIPTION OF THE FIGURES

[0028] For illustration only, a specific embodiment of the disclosure will now be described in detail with reference to the accompanying drawings, in which:

[0029] FIG. 1 is a schematic drawing of the network of the disclosure.

[0030] FIG. 2 is a schematic drawing of the network of the disclosure after certain method activities according to the disclosure have been performed.

[0031] FIG. 3 is a schematic drawing of the network of the invention after other method activities according to the disclosure have been performed.

[0032] FIG. 4 is a flow chart showing the method of the disclosure.

DETAILED DESCRIPTION OF EMBODIMENTS

[0033] FIG. 1 shows a Wi-Fi network in accordance with the disclosure. There is a master access point 2 (which will be referred to as master AP 2) which can communicate wirelessly with three slave access points 3,4,5 (which will be referred to here as slaves 3,4,5). Slave 3 can communicate wirelessly with three client devices labelled 6 in FIG. 1. Slave 4 can communicate wirelessly with three client devices labelled 7 in FIG. 1. Slave 5 can communicate wirelessly with three client devices labelled 8 in FIG. 1. In the specific embodiment described here, each of the three slaves has one radio which it uses for communicating with its respective client devices.

[0034] Each slave 3,4,5 can communicate with its respective devices on one of several different channels within the frequency band of operation of that slave 3,4,5. The quality of each channel is primarily affected by two factors: interference (e.g. from microwave ovens and analogue TV senders); and other Wi-Fi traffic. These two factors reduce the quality of communication on a given channel. However, the extent to which the two factors reduce the quality of the channel varies from one channel to another and also varies over time. It is desirable for each slave to communicate with its clients on the best-performing channel in terms of signal quality.

[0035] To this end, each slave selects a particular channel as its working channel for a period of one week, for example. Therefore, for one week the slave 3 communicates with its clients 6 on that working channel. At regular intervals over the course of the week, the slave 3 measures the level of noise on the working channel, the level of neighbor contention on the working channel and its “own utilization ”. By neighbor contention level it is meant the proportion of the time that slaves other than the slave 3 and its associated clients are putting sufficient energy onto the channel that the channel appears busy and transmission between the slave 3 and its clients 6 is not possible. The “own utilization ” measurement is the amount of time that data is being transmitted from or received by the slave 3. Specifically, this is the proportion of each minute that the slave 3 is sending or receiving data.

[0036] The slave 3 sends the measured data to the master 2. The master 2 uses the measured data to determine the average noise level and neighbor contention level for the slave 3. When the week has elapsed, the slave 3 switches its radio to a different channel and proceeds to communicate with its clients 6 on that channel for one week. Over the course of that week, the slave 3 makes regular measurements of the noise and contention levels and sends them to the master 2. This process repeats using each of the slave's available channels as the working channel in accordance with a schedule. The master 2 compares the determined average noise level for each channel and also compares the determined average contention level for each channel from these comparisons determines the worst performing channel. The slave 3 then cycles through each of the channels again, each for one week as before, except for the determined worst performing channel which is left out of this cycle. As in the first cycle, the slave 3 makes noise and contention measurements for each working channel which are averaged and compared as before. As before, the worst performing channel is excluded from the subsequent cycle.

[0037] If, at the end of each week-long period, a working channel is in use (i.e. not idle) then the change in working channel is delayed. Furthermore, if a channel has been excluded from the cycle for some time, the master 2 will not have any recent performance data in respect of that channel. For these reasons, it may be that the master has received insufficient recent noise and contention data in respect of one or more channels to enable it to make a meaningful performance comparison between all the channels. Therefore, the master 2 monitors the amount of performance data it has received from the slaves and identifies one channel of one slave in respect of which it has received the least recent performance data. This channel will be referred to as the “target channel” and the slave as the “target slave”. In this description, slave 3 will be considered to be the target slave.

[0038] The master 2 asks the target slave 3 whether or not its radio is idle. If the radio is idle, the master 2 instructs the target slave 3 to switch its radio to the target channel for a period of time, and to make performance measurements on the target channel over the course of that period. When the period has elapsed, the target slave 3 returns to its normal channel schedule.

[0039] The target slave 3 sends the measured performance data to the master 2 which uses it in the channel performance comparison process referred to above.

[0040] If the target slave's radio is not idle, the master 2 instructs the target slave 3 to test whether the target slave's clients would be able to switch to communicating with one or more of the other slaves 4,5 rather than the target slave 3. The target slave 3 does this by sending an 802.11k measurement request to each client 6, asking it to measure and report the signal strength at which it sees each of the other slaves 4,5. The target slave 3 reports this information to the master 2. If the signal strength is sufficient for the clients 6 to switch to another slave(s), the master instructs the target slave 3 to switch its clients 6 to that other slave(s) using 802.11v BSS transition requests. The target slave 3 does so, then switches its radio to the target channel for a period of time and makes performance measurements on the target channel over the course of that period. This situation is shown in FIG. 2. The target slave 3 sends the measured performance data to the master 2 which uses it in the channel performance comparison process referred to above. When the period has elapsed, the target slave 3 returns to its normal channel schedule. The master 2 then performs a load-rebalancing procedure which may involve switching the clients which were moved to a different slave back to the target slave 3. That procedure will not be described in detail here.

[0041] If it is not possible to switch the clients 6 to a different slave 4,5, the master determines which of the other slaves 4,5 is located nearest to the target slave 3. The master 2 does this by analyzing RSSI measurements for signals sent between the target slave 3 and neighboring slaves 4,5. The slave determined as the nearest to the target slave 3 will be referred to as the proxy slave. In the presently described embodiment, slave 4 is the proxy slave.

[0042] The master 2 asks the proxy slave 4 whether or not its radio is idle. If the radio is idle, the master 2 instructs the proxy slave 4 to switch its radio to the target channel for a period of time, and to make performance measurements on the target channel over the course of that period. When the period has elapsed, the proxy slave 4 returns to its normal channel schedule. The proxy slave 4 sends the measured performance data to the master 2 which uses it in the channel performance comparison process referred to above.

[0043] If the proxy slave's radio is not idle, the master 2 instructs the proxy slave 4 to test whether the proxy slave's clients would be able to switch to communicating with one or more of the other slaves 4,5 rather than the proxy slave 4. The proxy slave 4 does this by sending an 802.11k measurement request to each client 7, asking it to measure and report the signal strength at which it sees each of the other slaves. The proxy slave 4 reports this information to the master 2. If it is possible for the clients 7 to switch to another slave(s), the master instructs the proxy slave 4 to switch its clients to that other slave(s) using 802.11v BSS transition requests. The proxy slave 4 does so, then switches its radio to the target channel for a period of time and makes performance measurements on the target channel over the course of that period. This situation is shown in FIG. 3. In FIG. 3, the proxy slave 4 has switched its client to slave 5. When the period has elapsed, the proxy slave 4 sends the measured performance data to the master which uses it in the channel performance comparison process referred to above.

[0044] As proxy slave 4 is located in the vicinity of target slave 3, the communication conditions experienced by proxy slave 4 are similar to those experienced by target slave 3. Therefore, the averaged noise and contention measurements that the master 2 obtains using the proxy slave's performance data will be similar to those which it would have obtained had it used the target channel's performance data. Therefore, the master 2 uses the averaged noise and contention measurements obtained using the proxy slave's data as if they had been obtained using the target slave's data in the channel performance comparison process referred to above.

[0045] When the period has elapsed, the proxy slave 4 returns to its normal channel schedule. The master 2 then performs a load-rebalancing procedure which may involve switching the clients which were moved to a different slave(s) back to the proxy slave 4. This procedure will not be described in detail here.

[0046] If it is not possible to switch the clients to a different slave, the master determines which slave other than the proxy slave 4 is the next nearest to the target slave 3. The master 2 does this by analyzing RSSI measurements for signals sent between the target slave 3 and its neighboring slaves. The process described above is then repeated with this newly identified slave as the proxy slave.