GB2399999A - Mitigating interference between wireless systems - Google Patents

Abstract

Two separate radio frequency networks (e.g. WLAN and Bluetooth) may be operated within interference distance from one another in a way which mitigates the possibility of interference. A device node receives a modulated noise signal and identifies a characteristic in the noise signal without demodulating the signal. The identified characteristic is used to identify the noise signal, (e.g. as a Bluetooth signal) and predict the behaviour of the signal without demodulating the signal. Being able to identify and predict the behaviour of the noise signal without demodulating the signal is useful as it removes the need for a complaint receiver, e.g. demodulation of a Bluetooth signal would require a Bluetooth complaint receiver. The received signal strength indicator (RSSI) signal may be used and the timing and characteristics of the interfering signal may be determined, e.g. a Bluetooth signal includes a telltale repeat interval or pattern and information about the synchronization reference point and slot occupancy probability can be used to develop a statistics package that indicates a probability of transmission from the interferer at any given time. Each node in a network can transmit information about its local noise source to each of the other nodes in the network so that a first node wishing to transmit to a second node can use the information on the noise at the second node to determine the best transmission time.

Description

MITIGATING INTERFERENCE BETWEEN WIRELESS SYSTEMS

Background

This invention relates generally to wireless systems Including wireless local area network devices.

Pacicet-based wireless local area network (LAN) devices enable a plurality of clients to be coupled together.vih a server without the need for extensive wiring. The ISLE 802.11 family of standards (ISLE Standard 802.11 available from the Institute of Electrical and Electronics Engineers, New York, New York) describes a standard for wireless LAN systems. TO involves the use of either 2.4 -Hz Industrial, Scientific and Medical (ISM) or 5 GH- communication frequency bands. These bands are minimally regulated and, as a result, other interfering wireless devices (that do not comply with the IEEE 80.11 standard) may be crarsmitting in the same area in the same band.

As an example, within a given office than is util ring a system compliant with the IEEE 800.11 standard, other individuals may utilize devices compliance with the Bluetooth specification (V.1.0, December 1, 1999) for JO wireless devices. Like the IEEE 80Q.11 standard, the Bluetooth devices also operate in the.4 GHAN ISM band.

Interference may result between one B7uetooth and packet-based wireless LAN devices. Generally, in the case of Bluetooth devices, their power output is relatively small relative to the wireless LAN devices. However, a proximate Bluetooth device may adversely affect and interfere with the reception of a local wireless LAM device. Another device in the LAN may transmit to a local LAN device proximate a Bluetooth transmitter. The remote LAN transmitter may have no idea that a lower power Bluetooth transmi Her may also be transmitting. As a result, interference may occur wh ch varies depending on the receiver that is receiving the signal.

Proposals for mit-gating the effects of interference between Bluetooth and packet-based wireless LANs operating in the same frequency band genera!' y have relied upon - fequency orthogonality. However, such te_nniques may be ineffective when the Bluetooth and wireless LAN devices are in close proximity which, of course, is when the incererence is most substantial.

Thus, there is a need for a way to mitigate interference between wireless devices operating on different standards within the same frequency band. In addition, there is a need for a system that accommodates for the problems that arise when various devices ire a wireless network are not aware that receivers in that network may be promulgate to non-compliant transmitters operating within the same frequency band.

Brief Description of the Drawings

Figure 1 is a schematic depiction of one embodiment of the present invention; Figure 2 is a block diagram of a portion of the mitigation module shown in Figure 1 in accordance with one embodiment of the present invention; Figure 3 is a depiction of a statistics package format utilized to transmit information between nodes in accordance with one embodiment of the present invention; Figure 4 is a block diagram for another component of the mitigation module shown in Figure 1 in accordance with one embodiment of the present invention; Figure 5 shows a hypothetical statistics package waveform in accordance with one embodiment of the present invention and further illustrates how the statistics package may be utilized to determine when to transmit information to a receiver in accordance with one embodiment of the present inventions Figure 6 is a schematic depiction of a wireless LAN network with proximate Bluetooth transmitters in accordance with one embodiment or che present invention; and Figure 7 is a flow chart for sof ware in accordance with one embodiment or the present invention.

Detailed Description -

Referring to Figure 1, a node 10 in a wireless local area network (LAN) may be positioned proximate to a Bluetooth picoret 11. The Bluetooth piconet 11 may operate in accordance with the Bluetooth specification. The wireless - LAN node 10 may operate in accordance wich one of the wireless LAN standards such as che IEE 80.11 standard. The node 10 and the piconec 1' may operate in the same frequency band such as the A, 4 GHz Industrial, Scientific and Medical (ISM) band which is minimally regu' aced. The node 10 includes a mitigation module 16 that is responsible for 2C mitigating potential interference between the Bluetooth piconec 11 (which is not part of the wireless LAN that includes the node 10) and the node 10 itself.

The wireless Led] node 10 also includes a physical layer s In such as a modulator/demodulator or modem and a medium access control unit (MAC) l4. The physical layer may receive a received signal strength indication (RSSI) signal from the physical layer 12. The RSSI signal is conventionally utilized in association with what is known as - a channel access control.

The raw RSSI data, received from the physical layer 12, - is also uc,li7ed by the mitigation module 15. The - mitigation module 15 uses the RSSI data to detect transmission of any devices that are nor part or the LAM, such as transmission from a Bluetooth piconec. The mit gation module 16 subsequently develops scat- sties about the operation or such Bluetooth plconets. The statistics may then be used to predict when any device in the nlue_oo.h piconet may be transmitting. This prediction information may then be utilized to modify the transmission time of a transmitter within the HAN to avoid transmitting when a potentially interfering Bluetooth p'conet s more lively to also be transmitting.

While Bluetooth and 802.11 embodiments are described, the present invention is not limited to such e:.amcles.

Embodiments may be implemented to avoid interference between wireless transmitters in a variety or circumstances.

The statistical data developed by the mitigation module 16 is provided to the MAC 14. The MAC 14 then provides that informa,icn to owner LAN network transmitters wirelessly coupled to the node 10. In addition, the MAC i4 may use data received from other nodes in the LEN network to determine when to operate its own physical layer 12 in a transmission mode so as to reduce the likelihood of interfering with transmissions by Bluetooth piconets proximate to the internal wireless LAN receiver. Thus, the mitigation module 16 includes a statistics generating unit 18 and a collision probability estimator 44.

The Bluetooth specification compliant piconel 11

transmits data in regularly occurring bursts. These bursts may appear as relatively rectangular signal blocks that occur at regular intervals. Thus, in accordance with one embodiment of the present invention, when the node lO is neither sending or receiving wireless LAN signals, in is assumed that any background noise received by the antenna 15 is the result of a Bluetooth transmission signal. A Bluecooth signal includes z telltale 625 mic_osecQn repeat interval or pattern. Each 625 microsecond interval is called a "slot". The pattern of slot occupancy repeats with a period that is at most six slots and is always a -actor of six. However, any given sloe may or may nor be occupied w.h a transmission depending on the particular protocol uti iced by che proximate Bluetooth piconet 11.

The Bluetooth piconet 11 transmits in recurring sloes starting from a synchronization reference pain,. That is, each 625 microsecond slot begins at a synchronia.:on reference point. Information about the synchroni-atlon reference point, the slot occupancy probability, and the nature or the b25 microsecond transmission intervals may be collected over time. A probability may then be developed to determine the likelihood of interference between a transmission received by the node 10 and the noise received from the Bluetooth piconet 11.

In accordance with one embodiment of the present invention, it is not necessary to actually demodulate the RSSI data. This may be important in some embodiments because to do so may require that the node 10 include a Bluetooth compliant receiver. By identifying the Bluetooth signal and the background RSSI noise without demodulating the signal, sufficient information may be obtained, in some embodiments, about the nature of the proximate Bluetooth transmitter to decrease the likelihood of interference.

As mentioned above, not all of the slots of a Bluetooth transmission may be occupied. Different Bluetooth Protocols (such as HV1) may occupy or use different ones of The recurring set of six slots. For example, the HV1 protocol transmits data in every othe' slot. Thus, that Bluerooch protocol sends bursts of data in alternating 625 microsecond intervals with a six slot repeat. Tn general, the empty slots occur in a regular pattern in each six slot sequence.

By ro_lowing the sequence of six Lions, even without inicial2; knowing which slot IS the first slot or the sequence, the node 10 can find the empty slots and can determine the periodicity of those empty slots.

mi he stat sties generating unic 18 may sample the RSSI dare received from the physical layer 1 al regular intervals. Since the slot is 625 microseconds in length, 1Q advantageously the sample rate of the unit 18 is integrally dividable info 625. One such advantageous sampling rate is microseconds. Thus rate may be sufficiently fine to locate the start arid stop of Bluetooth transmission within a given slot without unreasonably increasing the design requirements for the node 10.

The statistics generating unit 18, shown in Figure 2, includes an inhibit line 32 coupled to the MAC 14. When the ARC 14 is operating the physical layer 12 to transmit or receive data, the inhibit line 32 terminates the generation of statistic packages. This inhibition avoids generating stat sties packages when the data may be obscured by the ongoing receipt or transmission of wireless LAN data (not pursuant to the competing protocol such as the Bluetooth protocol). Therefore, the analysis may be simplified and the results may be improved in some embodiments, by inhibiting the statistics package generation during internals when the node 10 itself is either transmitting or receiv fig. A synchronization estimate is achieved using an integrator 20, an offset removal unit 22, a shift register 24, a Bluetooth slot pattern correlate 36 and a Bluetooth slot pattern correlate 40. The synchronization eat-mate is based on a known pattern that repeats wild known periodiclty. The integrator 20 integrates the RSI data over each sample interval and develops an average level for the RSSI data. The DC offset remove' unit 0 tastes the average measurements and resolves them co zero over an extended time period. Thereby, the unit 23 removes any DC otrset in the RSSI dare.

The shift register 2 accumulates the integrated sample levels over a period of time. In one embodiment of the present invention, with a twentyf-ve microsecond sample rate, the shift register 04 may be capable of scoring twenty-five samples and re-circulating those samples. That is, in order to analyze the 605 microsecond slot pattern, successive sets of twenty-ive samples are stored one on top of the other in the twenty-five locations within the shift register 4. Feriodicallv, data is shifted our of the shift register 34 to the Bluetooth slot pattern correlate 40.

The unit 18 likely begins its analysis at an indeterminate point within the sequence or slots transmitted by the Bluetooth piconet 11. That is, the unit 18 initially has no way to know whether the slot it first receives happens to be the first slot in a sequence of six slots generated by the piconet 11. The correlate 40 finds the start point of the sequence of six slots. When the correlate 40 sees a peak in the data received from the shift register 94, the correlate 40 knows where the Bluetooth transmission pattern starts. Thus, by progressively overlaying the data in the shift register 4 over a sufficient period of time, She start of the slot sequence may be identified based on the t me location of the peak level.

The correlate 36 determines whether there is a transmission in a given slot. The correlate 40 winds where each 635 microsecond slot is, averaged over time.

When the inhibit line 32 is active, the shift register 24 simply recycles or re-circulaces without new input data to maintain synchronization with its previous analyses.

Thus, data is sniffed from the shift register 24 to the accumulator 25 and then summed with new data in the summer 28 during non-lnhibitd operation. In inhibited operations, the data simply circulates back to the shift register 24 through One combiner 30 that has been operated by the inhibit line 32 signal to block new input data and to simply circulate the current data residing in the shirt register 24.

The slot occupancy estimation unit 38 coordinates the start of each slot and determines, based on the data from the magnitude and synchronization unit 40 and the correlate 36, where the slot begins using the local time base. The magnitude and synchronization unit 4Q determines i- there is! any Bluetooth transmitter that has been recognized based on the RSSI data and determines if there is a peak in the data I from the Bluetooth slot pattern correlate 40. The magnitude and synchronization unit 42 tells the slot occupancy estimation unit 38 that a Bluetooth signal has been identified (or not) and provides a reference or start point for the first slot.

The estimation unit 38 then figures out if there is anything in each of the six slots. The output from the slot occupancy estimation unit 38 may be of the Format shown at in Figure 5. It may in the form of high pulses 62 and low pulses 64 that provide estimated Bluetooth transmission probabilities at given times. This information is a compilation of the timing of the slots of the local Bluetooth piconet 11 and the slot occupancy probability.

Thus, the pulse 62 indicates a higher probability of a luetooth transmission occurring while the pulse 64 indicates a lower probability.

The combination of data including a syncnroniacion reference point and a slow occupancy prooabilit-, escimaticn function may be represented as condensed data set with which to estimate che probability of a future time frequency collision with a detected Bluetooth piconec. For example, th s data may be compacted into a single 32-oit word that constitutes the statistics package communicated to one or more other nodes in a wireless LAN network.

Referring to Figure 3, the 32-bit word, in accordance with one embodiment of the present invention, may include a six tuple containing six probability estimates of two bits each, one for each Bluetooth slot. In addition, the 39-bit word may include a timing synchronization function (TSF) reference that provides time information than is correlated to the recognized tome base within the wireless LAN network.

The TSF data may, for example, be in accordance with the TSF standard set forth in the IEEE 802.11 specification. The TSF reference may be the least significant bits from a TSF timer, divided by twenty-five at the start of the first slot.

By providing the statistics package in a compact format, the statistics package may be readily and conveniently transmitted to all the nodes in a given network to advise them of the local conditions at each node. If each node transmits it own package during slack intervals, it is advantageous to provide the packages in a compact format to avoid any significant overall reduction of network bandwidth.

Each node 10 mitigation module 16 may also include a collision probability est mator 44 (Figure 1). The estimator 44 receives the statistics package from a unit 38 of a node to which the node 10 intends to transmit data.

Thus, in effect, the received statistics package provides information about the local interference condl.ions proximate co fine intended recipient node.

The estimator 44 receives a transmit request 66, shown in Figure 5. The estimator 44 compares the transmit request 66 to the statistics package bO. It initiates a transmit holdafl signal be that causes the transmission of the transmit request 6b to be shifted in time to a time when the probability of a collision is lower. Thus, if a request seeks a transmission at time 66 which would overlap with a higher probability Pulse 6, the transmission may be held off so that at most it overlaps with a pulse 64 indicating a lower probability of an overlap with a Bluetoolh piloted transmission.

The estimator 44, shown in Figure 4, expands the data contained in the statistics package 60. Based on a t-mer, the estimator 44 knows what time it is. The estimator 44 takes the statistics package (such as the package 60 in Figure S) including the time data received from the local sample interval unit 52 and the local slot number 50 and maps that data against the current time. The sample interval unit 52 supplies the sample interval information (e.g., 05 microseconds). The local slot number 50 may supply the slot interval (e.g., 625 microseconds). The global sample interval 54 aligns the statistics data JO the correct time by calculating the time relet ve to the statistics package. Based on the current time, the probability estimator 44 determines the occupancy probability for the next six Bluetooth slots.

The probability estimator 45 provides the ability to predict what a Bluetooth piconet 11 will do in the future based on the statistic package 60 developed from analyzing the Bluetooth transmissions over a period of time. A collision probability calculator 48 receives the Bluetooth occupation probability estimation from the estimator 45 and the packet length for the packet intended to be transmitted by a node 10. This information may be provided in the Transmit request 66. The wireless LAN noae's intended trarsml: characteristics are expanded and compared over the next six slots and data for each slot is provided to the collision probability calculator 48. Thus, the calculator 48 receives slot by slot data from the occupation calculator 46 and slot by slot data from the es imator 45.

The output of the calculator 48 is provided to a threshold comparator 56. The comparator So compares the transmit request 66 to the estimated Bluelooth Eransmission probability indicated at 60 and determines whether IO initiate a holdoff 68. The holdaff 68 moves the proposed transmission to a period of time of acceptably low collision probabilities.

A hypothetical local area network, shown in Figure 6, may include a node or transceiver 72, a node or transceiver and a node or transceiver 76. In addition, a Bluetooth/LAN transceiver or access point 74 may also be included in the network. An access point is a bridge connected on one side of one network and on the other side to another network for forwarding packets between the two networks. In addition to the wireless local area network including transceivers 7, 74, 76 and 80, a plurality of Bluetooth piconets 70, 78 and 82 may be proximate to one or more of the transceivers 70 through 80. For example, the piconec 70 may have a range 70a which encompasses the transceiver 72. Likewise, the piconet 78 may have a range 78a that encompasses the access poinl 74 and the piconet 89 may have a range 82a that encompasses the transceiver 76.

In this example, the Bluetooth piconets 70, 78 and 82 may operate in the same frequency band as the wireless LAN transceivers 72, 74 and 76. Thus, the possibility of interference ex sts between a locally proximate 31uetooch cicone: such as the piconer 70 and the transceiver 72. In contrast, the transceiver 80, which is not proximate to any of the Bluetooth piconecs, may not have any Bluetooth interference problems.

The access point 7a may transmit dale to the transceiver 72 as indicated in 84. However, the access point 74 may be far enough away from the Bluetooth piconec that the access point 74 may have no way to directly determine that its transmission may be interfered with by che Bluetooth piconet 70.

Instead, each transceiver 72, 74, 76 and 80 of the wireless LAN network does its own local evaluation o any potential interferers. Thus, the transceiver 7 analyzes the transmission from the Bluetooth piconet 70 within the range 70a and prepares a statistics package. The statistics package developed by the transceiver 7 and particularly by its unit 18, may then be transmitted to the access point 74.

In one embodiment of the present invention, a relatively compact transmission such as the 32-bit word illustrated in Figure 3, may be utilized.

Similarly, each node, such as the transceivers 72, 74, 76 and 80, transmits its statistics package information to all the other nodes in the wireless LAN network. As a result, any node wishing to transmit data to any other node can then take into account the local interference conditions with respect to the intended receiving station.

A transmitter, such as the access point 74, then uses a statistics package that it received from the transceiver 79 to rime its transmission 84 to the transceiver 72. This is done through the collision probability estimalor 44 local co the access point 74. More particularly, the statistics package may be generated by a unit 18 in the Transceiver 72 and transmitted co all of the other networks. nodes. The collision probability estimator 44 In the access point 74 may use the statistics package from the transceiver 7_ to make collision avoidance decisions and to control the timing of one transmission of data from the access point 74 to the t^anscei-rer 72.

Re_erring back to Figure 1, the MAC 14 may Include a processor 110 and a storage tin the' stores interference mitigation sor-are 90, in accordance wilh one embodiment of the present invention.. The software 90 may control the operation of the mitigation module 16 itself including the urn' 18 and the estimator 44. In some embodiments of the present invention, that control may be implemented in software and in ether embodiments, the control may be implemented in firmware or hardware. Similarly, the unit 18 and estimator 44 are Illustrated as being implemented in n O hardware but in other embodiments, they may be implemented in software.

Referring to Figure 7, the interference mitigation software 90 begins by preparing a local statistics package for any local Bluetooth piconet as indicated in block 92.

n 5 The statistics package is prepared in the unit 18. A check at diamond 94 determines whether an open channel exists. If an open channel exists, wherein no ongoing transmissions or receptions are occurring in a particular node 10, that node may transmit its local statistics package to all the other nodes in a wireless LAN network as indicated in block 96.

When the transmission request is received at a node 10, as indicated in diamond 98, a statistics package that was previously received from the intended target receiver is acquired as indicated in block 100. The collision avoidance calculation is implemented as indicated in block 102 using the estimator 44 for example.

A check at diamond 104 determines whether the collision probability threshold probability is exceeded. If so, the transmission is heldoff as indicated in block 106. When the transmission threshold is no 'anger exceeded, as determined in diamond 104, the data is transmitted as indicated In iO8.

While the present invention has been described with respect to a limited number or embodiments, those skilled in the art will appreciate numerous modifications and variations therefrom. It is intended that the appended claims cover all such modifications and variations as fall within the true spirit an] scope of this present invention.

Whar is claimed is:

Claims (13)

1. A method comprising: receiving a modulated noise signal; identifying a characteristic in said noise signal without demodulating said signal; and using said characteristic to identify said noise signal.

2. The method of claim 1 wherein receiving a noise signal includes receiving a noise signal having a characteristic identifiable without demodulating said signal and using said characteristic to predict the behavior of said signal without demodulating said signal.

3. The method of claim 2 wherein identifying the characteristic includes identifying a time characteristic in said noise signal without demodulating said signal.

4. The method of claim 3 wherein identifying a characteristic includes identifying a periodicity in said noise signal and using said periodicity to predict the future behavior of said noise signal.

5. A device comprising: a receiver that receives a modulated noise signal and identifies a characteristic in said noise signal without modulating said signal; and a unit that uses said characteristic to identify said noise signal.

6. The device of claim 5 including a transmitter that controls transmissions to reduce the likelihood of interference at an intended transmission recipient.

7. The device of claim 5 wherein said receiver includes a circuit that develops a statistical estimation of the likelihood of the occurrence of the noise signal based on the nature of said characteristic.

8. A method comprising: receiving a modulated noise signal having a characteristic identifiable without demodulating said signal; and using said characteristic to predict the behavior of said signal without demodulating said signal.

9. The method of claim 8 including receiving a slotted noise signal and determining the probability that a given slot is occupied.

10. The method of claim 8 wherein receiving a signal having a characteristic includes receiving a signal having a time characteristic and using said time characteristic to predict the behavior of said signal at a future time.

11. A device comprising: a receiver that identifies a modulated noise signal without demodulating said signal based on a characteristic of said noise signal; and a unit that predicts the behavior of said signal based on said characteristic without demodulating said signal.

12. The device of claim 11 wherein said unit identifies a slotted noise signal and determines the probability that a given slot is occupied.

13. The device of claim 11 wherein said receiver develops a statistical package indicating the probability that a noise signal will occur at a given time instance.