WO2005099174A1

WO2005099174A1 - Dynamic frequency selection for a wireless system

Info

Publication number: WO2005099174A1
Application number: PCT/IB2005/051085
Authority: WO
Inventors: David H. Evans; Adrian W. Payne
Original assignee: Koninklijke Philips Electronics N.V.
Priority date: 2004-04-08
Filing date: 2005-03-31
Publication date: 2005-10-20
Also published as: GB0407966D0

Abstract

A wireless device (110, 160) which selects a frequency on which to operate, predicts interference from information received about possible operating frequencies of a neighbouring wireless device (120, 130). If interference is predicted, a number of actions are possible. The frequency selection could be tailored to avoid such frequencies, a search for available frequencies could avoid them, or the neighbouring wireless device could be instructed to alter its transmissions. Additionally, to aid the selection of a frequency, the device may listen to see if there is interference before using a frequency.

Description

DESCRIPTION

DYNAMIC FREQUENCY SELECTION FOR A WIRELESS SYSTEM This invention relates to a wireless device, in particular to a wireless device having frequency selection, to an integrated circuit for use in a wireless device, to software for use in a wireless device, to methods of operating a wireless device, and to systems for predicting interference between wireless devices.

Most wireless systems operate in isolation and it is uncommon to find dissimilar systems operating in very close proximity to one another. The principal reason for this is that the transmissions from one system can affect the performance of another, usually in a detrimental way. Care is invariably taken in the definition of the spectrum regulations and technical specifications to ensure that unwanted transmissions are kept below functional levels so that operation under normal circumstances is not affected. Likewise, where the allocated frequency bands are shared by dissimilar systems, additional mechanisms are mandated as part of the specification to avoid mutual interference. A good example is the 5 GHz Wireless Local Area Network (WLAN) bands. WLAN systems have to share these bands with the mobile satellite service and radar systems. WLANs pose a potential interference source to the satellites and radars could interfere with wireless LANs. Consequently mechanisms are mandated for the WLAN systems to minimise the likelihood of interference. There is now an interest in operating dissimilar systems in very close proximity to one another, typically within a single mobile terminal. An example is the combination of Bluetooth in a cellular phone, and another example is the integration of a wireless LAN also within a cellular terminal. However when these wireless systems are operated in very close proximity the parameters of the basic specifications may be insufficient to avoid interference and also the additional mechanisms may have limiting features that may restrict good operation. US patent application 20003/0198200 A1 discloses two separate devices operating in the same frequency band. It uses measurement of interference between the two devices. It shows the management of the spectrum utilization of a frequency band that is shared, both in frequency and time, by multiple devices. At one or more devices operating in the frequency band, pulses associated with signals occurring in the frequency band are detected by sampling part or all the frequency band for a time interval. From the detected signal pulses, the signals can be classified. In addition, overall spectrum activity can be measured. Using classification information for signals detected in the frequency band, policies can be executed so that a device may take certain actions in order to avoid interfering with other signals, or to optimize simultaneous use of the frequency band with the other signals. WO 03/001742 A1 shows listening to and measuring interference between separate wireless devices using the same frequency band. There remains a need for improved handling of these issues.

It is an object of the invention to provide improved apparatus or methods. According to a first aspect of the invention, there is provided a wireless device having a frequency selection means arranged to select an operating frequency within a first band of frequencies, and to receive information about possible operating frequencies of a neighbouring wireless device, the frequency selection means being arranged to use the information to predict if there could be interference between one or more frequencies in the first band and the possible operating frequencies. A consequence of such prediction is that the chance of interference can be reduced. Compared to relying only on listening to determine interference, such prediction has a number of advantages. Listening takes time and can lead to long connection set up delays. Secondly, such listening may miss completely an intermittent interfering signal. Receiving information about possible frequencies of a neighbouring wireless device helps enable such advantages. Conventionally, such information is not made available. The interference can be in terms of the wireless device being susceptible, or, in principle, in terms of the wireless device causing the interference to the neighbouring device. If interference is predicted, a number of actions are possible, such as for example, action to reduce the predicted interference, for example by mitigating or pre-empting the interference or causing another device to do so. For this aspect, it is not essential to take any such action, as even if no such action is taken, simply generating the prediction can be useful, for example for informing a user about wireless conditions, or for diagnosing or testing purposes. Other uses can be envisaged. The frequency selection could be tailored to avoid such frequencies, the search for available frequencies could avoid them, or in principle, the neighbouring wireless device could be instructed to alter its transmissions to reduce interference. As an additional optional feature, the frequency selection means is also arranged to listen for interference. This combination of listening and prediction has a number of advantages. The combination can be made in various ways, including the use of listening to enable confirmation of the predictions, or use of the prediction to avoid or reduce the time spent listening to those frequencies predicted to cause interference. As another additional feature, the search by listening is not carried out for the frequencies for which interference is predicted. This enables the search to be made more quickly, and can reduce the overall time for establishing a connection. As another additional feature, the possible operating frequencies comprise frequencies outside the first band, and the predicting comprises determining harmonics within the first band. This is useful as harmonics are a principal type of interference, more than broad band noise for example. As another additional feature, the information comprises an indication of all the possible operating frequencies of the system operated by the neighbouring wireless device. This can enable the frequency selection means to avoid them all. This has the advantages of simplicity since little information is needed and it need not be updated often. As another additional feature, the information comprises an indication of any of: frequency bands of operation of the neighbouring wireless device, a type of cellular system of the neighbouring wireless device, and extent of the first band. This can help determine more accurately which frequencies are available. As another additional feature, the information comprises an indication of a frequency currently used by the neighbouring wireless device. This refinement can be useful where it is desirable to determine more accurately where interference is predictable so that more of the spectrum can be searched or made available. This can enable more capacity to be used more effectively while still enabling reduced interference. As another additional feature, the information comprises an indication of transmitted power levels of the neighbouring wireless device. This is a further refinement enabling more of the spectrum to be searched or made available.

There are a number of ways of using this type of information, including influencing the frequency selection directly to avoid higher powered of the possible frequencies, or indirectly by influencing which frequencies are listened to, to reduce the amount of listening, or to refine predictions of interfering channels for example, as desired. Such refinement can involve determining if some of the possible frequencies have sufficiently low transmission power that interference can be tolerated or that one can be confident the harmonics will not cause interference. If combined with listening, a suitable threshold of transmission power can be determined dynamically. Alternatively, a threshold can be predetermined by simulation or measurement. As another additional feature, the wireless device may incorporate the neighbouring wireless device. This can enable operation of the device in more than one communication system, and can facilitate communication of the information. As another additional feature, the wireless device is incorporated in a portable consumer device. In such devices, the proximity problem is more severe. In principle, it can also be of use in fixed terminals, for example base stations, but interference is usually less of an issue because the antennae can usually be spaced apart. As another additional feature, the neighbouring wireless device incorporated into the wireless device may be arranged to operate in a cellular system. As another additional feature, the wireless device may be arranged to operate in a WLAN or Fixed Wireless Access (FWA) system. These are some of the most commercially significant types of wireless device for short range mobile users. Another aspect provides an integrated circuit for use in the wireless device set out above, and arranged to receive information about possible operating frequencies of a neighbouring wireless device, and having frequency selection means arranged to use the information to predict if there could be interference between one or more frequencies in the first band and the possible operating frequencies. Another aspect provides software for use in the above wireless device, for receiving information about possible operating frequencies of a neighbouring wireless device, and using a frequency selection means to use the information to predict if there could be interference between one or more frequencies in the first band and the possible operating frequencies. Another aspect provides a corresponding method of operating such a wireless device, by receiving information about possible operating frequencies of a neighbouring wireless device, and using a frequency selection means to use the information to predict if there could be interference between one or more frequencies in the first band and the possible operating frequencies. Another aspect provides a wireless device arranged to use an operating frequency within a first band of frequencies, and to receive information about one or more possible operating frequencies of a neighbouring wireless device outside the first band, having a means for using the information to predict if there could be interference between one or more frequencies in the first band and the possible operating frequencies, and means for controlling the wireless device or the neighbouring wireless device to reduce the interference if it is predicted. This is not limited to devices having a frequency selecting system, and is intended to cover any way of avoiding interference involving passing information on current frequencies in use or to be used, between two wireless devices and using such information to reduce interference. Another aspect provides a system for predicting interference between a wireless device arranged to use an operating frequency within a first band of frequencies, and a neighbouring wireless device having one or more operating frequencies outside the first band, the system having means for receiving information about the frequencies in use or to be used and for calculating harmonics from the information, and comparing the harmonics with the frequencies to predict interference. This is not limited to frequency selecting systems and extends to prediction being carried out externally to the wireless devices. Any of the additional features can be combined together or with any of the aspects of the invention, as would be apparent to those skilled in the art. Other advantages may be apparent to those skilled in the art, especially over other prior art not known to the inventors.

Embodiments of the invention will now be described by way of example, and with reference to the accompanying drawings, in which: Figure 1 illustrates a known listening process for using in dynamic frequency selection; Figure 2 shows an arrangement of a wireless device according to an embodiment of the invention; Figure 3 shows another embodiment of the invention; Figure 4 shows steps according to another embodiment of the invention, for use in the arrangement of Figure 2 or other embodiments; Figures 5 and 6 show a flow chart according to another embodiment; and Figure 7 shows another embodiment of a wireless device. By way of introduction to the embodiments, some analysis of the interference problem will be described. A first point is the overlap of frequencies. The harmonics of some cellular transmissions fall within the 5 GHz WLAN bands. These are usually third harmonics and the systems and bands that are of interest are given in Table 1 below.

Table 1 : Cellular systems and their frequencies and the affected WLAN frequencies

Clearly the whole of the lower 5 GHz band (5150 to 5353 MHz) could be affected by either GSM (Global System for Mobile Communications) transmissions or UMTS FDD band III (UMTS= Universal Mobile Telecommunications System, FDD=Frequency Division Duplex) transmissions and parts of the upper band could be affected by either UMTS FDD band II or UMTS TDD (Time Division Duplex). Therefore the operation of a GSM or a UMTS cellular transmitter in close proximity to a WLAN receiver could seriously degrade the performance of the latter, especially when they form part of an integrated system. The sixth harmonics from the 900 MHz cellular bands may also pose an interference threat. A second point concerns the intermittent nature of interference. In the case of a WLAN, dynamic frequency selection (DFS) operates by randomly selecting a channel from an appropriate number of channels. The WLAN receiver then listens for transmissions within the channel for a period of 60 seconds. If nothing is detected in the channel, it is assumed to be clear and normal operation can commence. Otherwise, if the channel is found to be occupied the receiver is tuned to another channel and repeats the listening process. This process clearly takes time and it is possible that an associated cellular system may not be operating during the listen period, causing the WLAN to choose a channel that may at a latter time be subject to interference from the cellular system. To address these issues, the embodiments described below use knowledge of the cellular system to assist the WLAN's DFS mechanism. Some of the embodiments relate to a wireless transceiver employing dynamic frequency selection (DFS) and operating in close proximity (for example in same product) to another transceiver operating in a lower frequency band. The DFS search algorithm uses information supplied by the lower-frequency transceiver about its operating frequencies to avoid searching frequencies that may suffer interference from harmonics generated by the lower-frequency transceiver. The types of information about the lower-frequency transceiver that may be used can include: a) Frequency band - in which case the DFS avoids selecting a sub- band where harmonics may be generated. b) Frequency channels - in which case the DFS avoids selecting specific frequencies where harmonics may be generated. c) Transmit power - in which case the DFS avoids selecting specific frequencies where harmonics are likely to be generated at a damaging level. These can be used as alternatives or in combinations. Faster channel selection can result, and better management of the operating frequency of a WLAN radio system so that it can avoid the interference and resultant degradation. By making use of such prior knowledge of what cellular system and/or radio channels are being used, the dynamic frequency selection (DFS) mechanism of the WLAN can either select one or other of the two 5 GHz frequencies band or it can deliberately avoid channels where the cellular interference will occur. This can save considerable time when a WLAN initiates a connection and has to execute the DFS procedure. It can be applied to a wireless transceiver employing dynamic frequency selection (DFS) and operating in close proximity (for example in same product) to another transceiver operating in a lower frequency band. The DFS search algorithm uses information supplied by the lower-frequency transceiver about its operating frequencies to avoid searching frequencies that may suffer interference from harmonics generated by the lower-frequency transceiver. If applied to GSM1800 and UMTS, it can reduce the interference. from the third harmonic of their operating frequencies that sometimes occurs in the 5 GHz WLAN band. This is useful where cellular and WLAN systems are operated simultaneously and in close proximity within for example a combined single mobile terminal. Figure 1 shows a conventional listening process, according to ETSI

(European Telecommunication Standards Institute) specification EN301893V1.2.3 (2003-08). In this example of a dynamic frequency selection process, the upper line of the graph shows a power-up signal varying with time. The middle line of the graph shows a channel availability check (CAC) occurring, and the third line shows an interference signal, in this case in the form of radar signals from a distant source. At time To, through to T_{1 t} indicated as Tpowe_mp in Figure 1 , the wireless device is powering up. At time Ti for a duration of T_C avail chec_k, the wireless device listens to interference on a given channel. If no interference occurs, then that channel would be selected for use by the wireless device. If, as shown, interference is detected within the CAC time, then a second CAC process at a new frequency within the first band, is tried, as shown in dotted lines. The first channel to be checked is selected at random, to enable approximately uniform use of the channels of the first band. For further details of this DFS process, the reader is referred to the above mentioned ETSI standard. Figure 2 shows an example according to an embodiment of the invention. A wireless device 40 has an antenna 50 coupled to a wireless transceiver 10. This is fed by a frequency selector 20. This in turn is fed by a prediction element 30 which predicts interfering frequencies from information provided on an input 60 about possible operating frequencies of a neighbouring wireless device. These operating frequencies can be inside or outside the first band. As mentioned above, by predicting interfering frequencies, frequency selection to avoid interference can be managed better than cases where listening only is used. Prediction can be combined with listening. Figure 3 shows another embodiment of the invention. In this case, a dual mode product 100 is shown. This has a first wireless device comprising a WLAN wireless transceiver 110 coupled to a processor 160. A neighbouring wireless device comprises a GSM and UMTS wireless transceiver 120 coupled to a control processor 130. The transceivers can be implemented following established principles. Typically for small mobile handheld products, they can be implemented on a single RF chip, carrying out digital and analogue RF signal processing, and including the interface for driving the antenna. The processor 160 is arranged to carry out frequency selection functions, and other control and coding/decoding functions represented by element 220. These other functions can include digital demodulation, link layer MAC protocol processing, and synchronisation functions for example. The frequency selection function is typically part of the link layer, layer 2 processing. This is shown in more detail in Figure 3, and includes a prediction element 200, for receiving information from the neighbouring wireless device. The prediction element 200 can produce a coarse prediction of interfering frequencies based on minimal information such as the cellular system and bands used by the neighbouring wireless device. An example of this prediction is shown in Figures 4 and 5, described below. For a finer prediction, the prediction element 200 can use information on the frequency currently in use by the neighbouring wireless device. If outside the band, then the prediction element 200 can calculate harmonics to generate a list of frequencies which are predicted to have interference. The frequency selection function also includes a random initial frequency select function 210. This feeds a selected frequency to a comparator 190 for comparing the candidate frequency with the list of frequencies predicted to have interference. This in turn feeds a listening element 180 arranged to use a CAC process to listen for interference on that given frequency. According to the output of this, a selected frequency may be set for use by the wireless transceiver 110. This selected frequency may be used directly by the transmitter of the transceiver 110, or may be sent to the far end for use by the transmitter at the far end of the link. Figure 3 also shows parts of the control processor 130 of the neighbouring wireless device. This control processor 130 includes other control and code/decode functions 150, and a store 140 of parameters indicating the system in use, and the frequency in use. Optionally this can also include an indication of transmitted power levels at one or more of the frequencies. This information is all made available to the prediction element 200 for predicting interference. Optionally, the prediction element 200 can be part of the control processor 130, if information on the proposed frequency for use by the WLAN wireless transceiver 110 is passed to the control processor 130. The link between the processors 160 and 130 can be implemented using conventional techniques, such as shared memory, or a serial link, for example. Optionally the two processors 160, 130 can be implemented on a single chip, in which case the parameters would be available without any special measures. The processors can be implemented using conventional hardware; in particular the functions of the two processors 160, 130 can be implemented in a single processor. In a typical implementation, demodulation and modulation of signals passing to and from the transceivers (respectively 110 or 120) is carried out by specialised DSP (Digital Signal Processing) circuitry, while other functions having less demanding processing speed requirements can be carried out by general purpose microprocessor circuitry. All these are typically implemented as modules of an Application Specific Integrated Circuit (ASIC). Software for carrying out the various individual functions described above can be implemented using conventional programming languages for execution on the processing hardware described above. Optionally, the interference prediction can be carried out for devices where the WLAN wireless transceiver is in a separate product to the GSM and UMTS wireless transceiver. In this case an external communications link is needed between the two devices. This could take the form of a Bluetooth link, or other short range radio link. If both devices are not portable, then a wire line link could be used. An example of the information available from the cellular system is as follows: 1. Operating frequency band. This is defined by the type of cellular systems that is being used as shown in table 1. 2. Operator's frequency channel. This identifies the specific part of the operating band that the cellular system for a given network provider is operating in. For example, a channel is 5 MHz wide for UMTS. 3. Transmit power control settings. The information from points 1 and 2 can be matched against the channel plan for the 5 GHz band and these channels are then excluded from search or from use. The information from point 3 may assist in determining how much of a threat the harmonic interference is likely to be. If the power setting is high, then the channels that may be subject to harmonic interference must be avoided. If the power is low, then it may be possible to use them. The information listed in points 1 , 2 and 3 above, can be provided by the cellular system to the DFS mechanism of the WLAN. Point 1 is relatively simple, as shown in Table 1. Based on knowledge of which cellular system the WLAN is operating in coexistence with, the DFS mechanism can be made to avoid either the upper or lower bands or to avoid parts of the upper band. The information in point 2 enables a refinement of this selection since use of point 1 alone would mean that large parts of the WLAN bands would have to be avoided which might be unacceptable. A cellular phone contains a frequency map of the base stations where it can get access to the network. This map is built up from previous searches and enables the phone to get rapid access without having to repeat this process each time that it is turned on. Supplying this information (cellular handset transmit frequency settings) to the WLAN DFS mechanism will enable the DFS to avoid searching on the WLAN channels where the harmonics of these frequencies occur. Only selected WLAN channels therefore need to be avoided rather than complete parts of the WLAN bands. The cellular transmitter power control settings (point 3) may give an indication of how strong the cellular interference could potentially be. If the transmit power setting is low, then it is possible that the channel containing the likely interference could be usable since the interference may be low. This is possible for two reasons. Firstly since the transmit power is reduced, the power of the harmonics will be correspondingly reduced and secondly under this situation the transmitter amplifier may not be driven so hard and in turn will produce even fewer, and lower power, harmonics. By making use of the knowledge of the cellular system, an algorithm can be created that can assist the DFS to avoid channels where interference from the cellular system may occur and hence possibly reduce the search time for the DFS to find clear channels. Furthermore in some situations such as where the cellular transmit power is known to be low, use of those channels may be possible. Therefore by use of the available information, only a minimum amount of WLAN channels may need to be avoided. This approach can avoid the third harmonic interference better than simple DFS. Figure 4 shows a flowchart indicating how the frequency selection functions can work according to an embodiment of the invention. At step 300 information is received from the neighbouring wireless device. This information can include frequencies in use, and if appropriate, transmitted power levels of the frequency or frequencies in use. At 310, a list of frequencies is generated, which are predicted to have interference. This can involve calculating harmonics of frequencies below the first band. At step 320 a proposed next frequency for use by the wireless device is generated. At step 330, if the proposed frequency generated at step 320 is in the list generated at step 310, then the process jumps to step 350. If not in the list, then listening is carried out at step 340 by a CAC (Channel Availability Check) process, to listen for interference. At 345 it is tested whether the interference exceeds a given threshold. If yes, then the process moves to step 350, where it is tested if there are any more frequencies to listen to in the first band. If at step 345 the interference does not exceed the threshold, then at step 380 the selected frequency is output for use by the wireless device or for sending to the far end. The above mentioned step 350 of testing for whether there are any more frequencies to listen to can cause the process to return to earlier steps of generating the proposed next random frequency in the band at step 320, and at step 330, testing whether this proposed frequency is in the list of those predicted to have interference. If at step 350 it is found that there are no more frequencies to listen to, then at step 360 the information on transmitted power levels is used to find a frequency with the lowest power harmonic interference. At step 370 it is tested to see if this is below a predetermined threshold at which interference can be tolerated. If not, then at step 390 it is output that no frequency is available. Higher protocol levels can be alerted, and the process can be retried later as desired. If at step 370 the frequency is found to be below the predetermined threshold of interference, then the process returns to step 340, and a CAC listening process is carried out at this frequency. Figures 5 and 6 show a flowchart for predicting which frequencies could have interference, based on information about the cellular mode, band and operating frequencies, and the band used by the WLAN device. In Figures 5 and 6, references to 802.11a are intended to refer to devices using derivations of that IEEE802.11a standard which make use of DFS, or equivalent frequency selection methods. In this flowchart, the process boxes that describe modification of the DFS algorithm, refer to avoiding the 3rd harmonic of the relevant 1.8GHz cellular system that could be present at 5GHz. The decision boxes for checking if the particular cellular system is operating in a certain part of the band are checking whether the allocated channel for that system is within the specified band. At step 400, it is determined whether the WLAN device operates in an upper band. If so, the process moves to step 401 , otherwise it goes to Figure 6, described below. At step 401 it is determined whether the cellular mode is either of GSM-1800 or UMTS FDD band III. If so at step 402, it is concluded that all of the upper band is available for use. Otherwise, at step 403 it is tested to see whether the cellular mode is UMTS FDD band II. If not, then the process goes to step 405. If so, at step 404, it is tested whether the UMTS FDD is operating between 1850 and 1908 MHz. If not, then the process goes to step 407, where it is concluded that all of upper and lower bands of the 802.11a standard can be used. Otherwise at step 408, it is determined whether the WLAN device can operate in the lower band. If not, at step 410 it is concluded that the DFS should be modified to avoid the predicted interference on frequencies within the 1850-1908 MHz band. If the WLAN device can operate in a lower band, then at step 411 it is concluded that all of the lower band is available. Considering step 405, mentioned above, this involves testing whether the cellular mode is UMTS TDD (Time Division Duplex). If so, then at step 406 it is tested whether the device is operating between 1900 and 1908 MHz. If not then the flowchart moves to step 407 described above. If so, at step 409 it is determined whether the device can operate in the lower band. The flowchart goes to above mentioned step 411. Otherwise at step 412 it is concluded that the DFS should be modified to avoid the interference predicted for frequencies within any harmonics of 1900-1908 MHz. Finally, considering above mentioned step 405, if the cellular mode is not UMTS TDD, then the flowchart moves to above mentioned step 407. Figure 6 shows a similar flowchart for the event that the 802.11a device does not operate in the upper band. In this case, at step 500, it is tested whether the cellular mode is GSM1800 or UMTS FDD band III. If not, then at step 502 it is concluded that all of the lower band is available. If so, at step 501 it is tested whether the cellular mode is UMTS FDD band II. If yes, at step 503 it is tested whether it is operating between 1716 and 1783 MHz. If yes, the conclusion at step 504 is that the DFS should be modified to avoid frequencies between 1716 and 1783 MHz, where interference is predicted. If step 501 or step 503 test negative, then the flowchart moves to step 505. Here it is tested whether the GSM 1800 system is operating between 1716 and 1783 MHz. If so at step 506 the conclusion is to modify the DFS to avoid the predicted interference at frequencies between 1716 and 1783 MHz. Otherwise, at step 507 it is concluded that all of the lower band is available. The flowcharts of Figures 5 and 6 are intended to give a coarse prediction of interfering frequencies, based only on information about the WLAN device bands, the cellular mode and the bands used by the cellular device, without knowing the actual frequencies of operation. This enables a prediction based on a minimum of information, which need not be updated regularly. The flowchart of Figures 5 and 6 can be used as an alternative to the flowchart of Figure 4, or in combination. In combination, the coarse prediction might be used first, and if insufficient frequencies are found to be available, the more precise prediction using individual frequencies, shown in figure 4 could be used. The arrangement of Figures 5 and 6 can be adjusted appropriately to be applicable to different cellular systems or other systems used by a neighbouring wireless device at different frequencies, as appropriate. Clearly it can be modified to apply to different WLAN or FWA systems using different bands and different frequencies to the 5 GHz frequencies used by the 802.11a type WLAN systems. With regard to the flowcharts, alternative embodiments exist for the case when GSM-1800 and UMTS FDD in Band III are both simultaneously operating. One alternative embodiment would be to use a metric to determine whether the 3rd harmonic interference from UMTS is greater than GSM, rather than assuming UMTS will be worse. This metric could be based on transmit power settings. A further alternative embodiment would be to modify the 802.11a DFS such that it avoids interference from both GSM and UMTS. Figure 7 shows an arrangement similar to that shown in Figure 2, and similar reference numerals have been used where applicable. In this case the wireless device 40 differs in not necessarily using frequency selection. The wireless device 40 still has an input 660 to feed information on possible operating frequencies of a neighbouring device outside the first band. Optionally this neighbouring device can be integrated in the same product as shown in Figure 3. A means for predicting interfering frequencies 630 uses this information, as well as information on the frequencies in use or proposed for use by the wireless device 40 as stored in storage element 620. The interference prediction can be carried out following the flowcharts set out in Figures 4, 5 or 6, or in other ways. The predicted frequencies where interference can occur are fed to a controller 635, which reduces the interference between the wireless device 40 and the neighbouring wireless device in ways other than selecting appropriate frequencies. For example, the transmissions from the respective devices could be timed so as to avoid interference, or power levels could be adjusted to reduce interference when necessary. This could be applied to wireless transceivers which use multiple frequencies simultaneously. For example, frequency hopping systems, spread spectrum systems or OFDM type systems could be tailored to reduce or minimise their transmissions on frequencies which are predicted to have interference, either being susceptible to interference, or causing interference. As before, the interference may be from harmonics of lower frequency signals. The controller 635 may be coupled to exert control over the neighbouring wireless device by means of an output 670. The fields of application of the embodiments of the invention include combined cellular and WLAN systems, such as dual mode products e.g. GSM or UMTS mobile phone with WLAN capability. Optionally, the prediction and control can be carried out in an external system, coupled to the wireless devices to receive frequency information and send control information. As has been described above, a wireless device, 40 or 1 10, 160, selects an operating frequency within a first band of frequencies, and listens to see if there is interference before using the frequency. It also predicts interference from information received about possible operating frequencies of a neighbouring wireless device 120, 130. Compared to relying only on listening only to determine interference, such prediction can reduce connection set up time and can handle an intermittent interfering signal. If interference is predicted, a number of actions are possible. The frequency selection could be tailored to avoid such frequencies, the search for available frequencies could avoid them, or the neighbouring wireless device could be instructed to alter its transmissions. Other variations and examples within the scope of the claims will be apparent to those skilled in the art. In the present specification and claims the word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. Further, the word "comprising" does not exclude the presence of other elements or steps than those listed. The inclusion of reference signs in parentheses in the claims is intended to aid understanding and is not intended to be limiting.

Claims

1. A wireless device (110, 160) having a frequency selection means (160) arranged to select an operating frequency within a first band of frequencies, and to receive information about possible operating frequencies of a neighbouring wireless device (120,130), the frequency selection means (160) being arranged to use the information to predict if there could be interference between one or more frequencies in the first band and the possible operating frequencies.

2. The device of claim 1 , frequency selection means (160,180) also being arranged to select the operating frequency based on the prediction of interference.

3. The device of claim 1 or 2, the frequency selection means

(160,180) also being arranged to listen for interference, and to select the operating frequency based on the listened interference.

4. The device of claim 3, the listening being adapted based on the prediction of interference.

5. The device of any preceding claim, arranged to reduce interference based on the prediction of interference.

6. The device (1 10,160) of any preceding claim, wherein the frequency selection means (160) is arranged to predict whether there could be interference from harmonics of the possible operating frequencies of the neighbouring wireless device (120, 130) operating outside of the first band.

7. The device (1 10,160) of any preceding claim, the information comprising as indication of all the possible operating frequencies of the neighbouring wireless device (120, 130).

8. The device (110,160) of any preceding claim, the information comprising an indication of any of: frequency bands of operation of the neighbouring wireless device (120, 130), a type of cellular system of the neighbouring wireless device (120, 130), and extent of the first band.

9. The device (110,160) of any preceding claim, the information comprising an indication of a frequency currently used by the neighbouring wireless device.

10. The device (110,160) of any preceding claim, the information comprising an indication of transmitted power levels of the neighbouring wireless device (120, 130).

11. The device (110,160) of any preceding claim, further comprising the neighbouring wireless device (120, 130).

12. The device (110,160) of claim 11 , wherein the neighbouring wireless device (120, 130) is arranged to operate in a cellular system.

13. The device (1 10,160,) of any preceding claim, arranged to operate in a Wireless Local Area Network (WLAN) or Fixed Wireless Access (FWA) system.

14. An integrated circuit for use in the wireless device (110, 160) set out in any preceding claim, and arranged to receive information about possible operating frequencies of a neighbouring wireless device (120, 130), and having frequency selection means (160) arranged to use the information to predict if there could be interference between one or more frequencies in the first band and the possible operating frequencies.

15. Software for use in the wireless device (110, 160) as claimed in any preceding claim, for selecting an operating frequency within a first band of frequencies, for receiving information about possible operating frequencies of a neighbouring wireless device (120, 130), and for predicting if there could be interference between one or more frequencies in the first band and the possible operating frequencies.

16. A method of operating a wireless device (110, 160) arranged to use an operating frequency within a first band of frequencies, having the steps of receiving information about possible operating frequencies of a neighbouring wireless device (120, 130), and of using the information to predict if there could be interference between one or more frequencies in the first band and the possible operating frequencies.

17. A wireless device (10, or 630, 635) arranged to use an operating frequency within a first band of frequencies, and to receive information about one or more possible operating frequencies of a neighbouring wireless device outside the first band, having a means for using the information to predict if there could be interference between one or more frequencies in the first band and the possible operating frequencies, and means for controlling the wireless device or the neighbouring wireless device to reduce the interference if it is predicted.

18. A system (30 or 200 or 630) for predicting interference between a wireless device arranged to use an operating frequency within a first band of frequencies, and a neighbouring wireless device having one or more operating frequencies outside the first band, the system having means for receiving information about the frequencies in use or to be used and for calculating harmonics from the information, and comparing the harmonics with the frequencies to predict interference.