Currently, short-range radio communication systems (e.g. WLAN 802.11, Bluetooth, ZigBee, Z-Wave, etc.) use a bi-directional data exchange. These systems are based on connections that are controlled by higher-layer applications. To achieve short latencies, radio receivers need to scan frequently or are locked into low-duty cycle connections. Low latency and low power are difficult to achieve simultaneously.
In accordance with an embodiment of the present invention, a radio system includes a master device. The master device may include a wake-up transmitter to send a wake-up message to a slave device to wake-up the slave device. The master device may also include a short-range transmitter to communicate with the slave device once the slave device has been woken up. The master device may be a mobile phone that communicates with the slave device (e.g., Bluetooth accessory devices) over a Bluetooth network.
In accordance with another embodiment of the present invention, a radio system includes a slave comprising a short-range transmitter and a wake-up radio receiver, wherein the slave is configured to have a sniff sub-rating with a master, the master being configured to wake up the slave via the wake-up radio receiver.
In accordance with another embodiment, a method is directed to communicating a master device with a slave device over a short-range network. The method may include establishing a sniff period from the slave to the master; establishing a subrating period from the master to the slave, wherein the subrating period is greater than the sniff period; and allowing the slave device to wake up during the subrating period so that at the next sniff period the slave communicates with the master.
Other aspects and features of the present invention, as defined solely by the claims, will become apparent to those ordinarily skilled in the art upon review of the following non-limiting detailed description of the invention in conjunction with the accompanying figures.
The following detailed description of embodiments refers to the accompanying drawings, which illustrate specific embodiments of the invention. Other embodiments having different structures and operations do not depart from the scope of the present invention.
Embodiments of the present invention may take the form of an entirely hardware embodiment that may be generally be referred to herein as a “module”, “device” or “system.”
Embodiments of the present invention will be described more fully hereinafter with reference to the accompanying drawings, in which some embodiments of the invention are shown. This invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to persons skilled in the art.
Generally, low power consumption in radio systems is achieved by applying very low duty cycle's on the link therebetween. If a radio system only operates for 1 ms every second, a 1000× power advantage may obtained. Further power reduction can be achieved by extending the interval between the wake-ups. However, this poses two issues: (1) the latency increases as the devices cannot be reached during the sleep interval; and (2) the transmitter receiver may get out of sync, which is especially an issue in frequency hopping (FH) systems. The latter issue is addressed below in the section titled “AUTOMATIC RECONNECTION OF A PREVIOUSLY LOST CONNECTION IN A FREQUENCY HOPPING COMMUNICATION SYSTEM.” As is discussed later, this concept keeps Bluetooth devices (or other short-range radio communication devices) loosely synchronized even if no data was exchanged.
Additionally, as discussed in the section titled “LOW POWER RADIO SYSTEM,” a low-power uni-directional radio system, which has an extremely low power receiver, can be applied as a wake-up radio for other more power hungry systems.
As described in the section titled “WAKE-UP ASSISTED SNIFF LINK WITH SUB-RATING,” the low-power uni-directional wake-up radio can be used with a Bluetooth's sniff link to achieve an overall system that combines low latency with low power.
As a general overview, some concepts disclosed herein apply sub-rating on a Bluetooth link combined with a wake-up radio to maintain a short response time in an accessory of a Bluetooth system or other short-range system. With sub-rating, the accessory can sleep for extensive times (up to hours if we apply the automatic reconnect concepts as described below in the section titled “AUTOMATIC RECONNECTION OF A PREVIOUSLY LOST CONNECTION IN A FREQUENCY HOPPING COMMUNICATION SYSTEM”) while keeping a short latency from the accessory to a mobile device (e.g., the mobile device keeps a sniff/reconnect interval of 1.28 s). To obtain a short latency also from the mobile device to the accessory while the accessory Bluetooth radio sleeps for hours, a wake-up radio is used with which the mobile device can wake-up the accessory.
Thus, embodiments of the present invention combine sub-rating (possibly extended with the auto-reconnect features) with a wake-up radio to obtain both low latency and a low standby current.
I. Automatic Reconnection of a Previously Lost Connection in a Frequency Hopping Communication System
An example of a frequency-hopping transceiver unit 200 is shown in
If a connection 120 between the master unit 100 and the slave unit 110 is lost during Bluetooth sniff mode the units 100, 110 would, in the known prior art, return to the well-known page and page scan modes to recover the connection 120. The master unit 110 would return to the page scan mode (with a duty cycle of about 0.9-1%) while the former master unit 100 would try one page lasting for about 10 seconds. If the connection 120 would not be recovered, the former master unit 100 would not page the slave unit 110 again. This recovery procedure may be costly with respect to power consumption. It would also not result in a successful recovery if the units were out of the coverage range for more than 10 seconds, i.e. the duration of the page. All in all, this means that the connection 120 would not be recovered automatically when the units come in range again, unless the interruption lasts less than said 10 seconds. Thus, in the known prior art, a user action would normally be required to re-connect the units again once they have lost their previous connection 120.
Embodiments of the present disclosure propose a recovery procedure, which allow for a more automatic recovery of a previously lost connection 120 between two units 100, 110 in a FH system, e.g. Bluetooth sniff mode, as compared to the known prior art. When a previous connection 120 between two units 100, 110 has been lost (e.g. due the fact that the two units 110, 110 have moved out of the coverage range of each other) a recovery of said connection 120 may be established and at the same time any excessive power consumption in the two units 100, 110 may be limited while acceptable latencies may be provided. Some embodiments of the present invention provide a substantially automatic recovery procedure for recovering a previously lost connection 120 between two units 100, 110 in a FH system, such as Bluetooth sniff mode. A recovery procedure is presented herein, which may have both a low-duty cycle page and a low-duty cycle page scan. This may be achieved by using clock information that was present during the last connection 120 between the two units 100, 110. The clock information may be available from free-running clocks in the respective units 100, 110. As time elapses, this clock information may be less reliable, e.g. due to clock drifts, which in turn requires a higher duty cycle. To this end, a recovery procedure having different recovery states is proposed herein. The recovery procedure may, preferably, be divided into a first recovery state (referred to as the fast recovery state in the following disclosure) and a second recovery state (referred to as the slow recovery state in the following disclosure). In the fast recovery state, the units 100, 110 may reconnect to each other within e.g. one second once the units come into the coverage range of each other again. If the units have not been reconnected (and, hence, re-synchronized) within a maximum time period TFASTRECOVERY of e.g. 30 minutes, the units may enter the slow recovery state. In the slow recovery state, the reconnection may take more time as compared to the fast recovery state, e.g. up to 30 seconds. It should be noted that the accuracy of the clock estimate may depend on the relative drift of the clocks in the two units 100, 110 as well as on the time elapsed since the two units exchanged their respective internal clock values during a connection. The larger the drift and the longer the time, the larger is the uncertainty in time and frequency and, hence, the longer the paging process may take. In the proposed system, the clocks are free-running. Clock offsets may be utilized in the estimation process. In this way, a unit may have a list of clock offsets with respect to one or several other units that it has been connected with in the past.
In the following, two exemplary embodiments of the inventive recovery procedure will be presented; more details can be found in the following. The building blocks used in the fast and slow recovery procedures may be based on the conventional paging scheme used in Bluetooth. For example, only ID packets may be used to exploit their robustness; the Device Access Code (DAC) of the slave unit 110 may preferably be used. Indeed, an ID packet comprises an direct-sequence (DS) code. This code may be related to the identity of the device in question (Device Access Code, DAC). Since it is a DS code, it may provide processing gain like in a direct-sequence spread spectrum system. In the receiving unit, the received ID packet is compared (correlated) with an exact replica of the code. If sufficient bits match, a reception may be announced. Depending on the threshold that is set in the system, it may be possible to accept the packet already when 80% or 90% of the bits match. This may provide the extra robustness to the system. Furthermore, the page hopping sequence belonging to the slave unit 110 may be used. These choices enable the master unit 100 to reach the slave unit 110 even using a conventional paging procedure. During recovery, the slave unit 110 will only scan periodically on a single hop frequency according to the page hopping sequence, similar to a conventional page scan. The recovery scan window may, however, be considerably shorter than the conventional page scan. For the slave unit 110, the difference between fast and slow recovery is only in the interval between scans, which may be somewhat longer in slow recovery than in fast recovery. The master unit 100 may transmit ID packets on different frequencies, but compared to conventional paging, the duty cycle may be much smaller.
In the following, two exemplary embodiments are presented in greater detail. It goes without saying that the various features described with reference to the two embodiments may also be combined in the same embodiment.
During fast recovery, the time elapsed since the previous or last synchronization between master unit 100 and slave unit 110 is still sufficiently small (a couple of minutes up to e.g. maximum 30 minutes) to be able to predict the timing in the both units 100, 110 (see
1. Fast Recovery Scan
When a loss of link 120 (see
During scan, the slave unit 110 may correlate the incoming signals against a known 68-bit access code related to the BD-ADDR of the slave unit 110. This may be an ID packet corresponding to the Device Access Code (DAC) also used in conventional page scan. When the core layer output exceeds a threshold, the slave unit 110 may be configured to enter a recovery response sub-state which may be identical to the Bluetooth page response sub-state. The slave unit 110 may be configured to return an identical ID packet at the proper timing and await the reception of a FHS packet. The FHS packet may then re-synchronize the timing and frequency hopping of the slave unit 110. The slave unit 110 may remain in recovery scan until the link 120 is re-established or when a timeout TFASTRECOVERY has exceeded indicating the end of the fast recovery state. Assuming a worst case mutual drift of e.g. 40 ppm, the timeout TFASTRECOVERY may, preferably, be set to about 30 minutes. It should be appreciated that the fast recovery scan is similar to the conventional Bluetooth page scan. A difference is, however, in the clock used for determining the hop frequencies and wake up timing. In the prior art scheme, it is the native clock of the slave unit 110 that sets the timing; in the fast recovery procedure disclosed here, it is the slave unit's estimate of the master clock that sets the timing.
2. Fast Recovery Page
When a loss of link 120 has been detected, the former master unit 100 enters the fast recovery page state. The initial timing of the fast recovery page may be based on the anchor point timing experienced during sniff mode. If an ID packet is sent at the anchor point, the slave unit 110 will receive this (provided the units 100, 110 are within range). The ID packet comprises the DAC of the slave unit 110 which is to be reconnected to the master unit 100. The frequency may be selected from the slave unit's page hopping sequence f(k), where k is the current clock of the master.
However, due to drift, the timing of ID transmissions, i.e. transmissions of an at least one ID packet, sent by the master unit 100 and that of the scan window in the slave unit 110 will drift; after 750 μs/2 y (ignoring the second ID transmission) there will not be an overlap anymore. For y=20 ppm, this may happen already after 18 seconds. Therefore, in the recovery page state, the transmission window may need to be increased both before and after the anchor point, see
The increase may be done gradually as time progresses as is illustratively shown in
3. Latency in Fast Recovery
The latency in the fast recovery procedure may preferably, but not necessarily, be determined only by the scan interval. When the units are in range and no errors occur, the maximum response time may be 1.28 s. If a uniform distribution of the scan timing once the units come into range is in, the average latency may be 640 ms.
4. Compatibility with Bluetooth Page
It may be the case that after link loss but before the slow recovery begins, one of the units 100, 110 is reset and enters the disconnected mode. For example, the master unit 100 may be reset while the slave unit 110 is still in slow recovery state. The master unit 100 would then not be able to link to the slave 110. This may be solved in two different ways. In a first variant, a special page can be constructed which is similar to the original, standardized, Bluetooth paging method with A and B trains. A difference is in the repetition time of the A and B trains. In the standardized method, the use of A and B trains is alternated every 1.28 s. The slave unit 110 in recovery scan will miss its frequency always when it is misaligned and its scanning frequency is not in the batch with overlapping frequencies (this is because the scan window is smaller than 10 ms, the time needed to cover all 16 frequencies of a conventional page train). However, since the slave unit 110 in recovery scan has exactly the same periodicity of 1.28 s, it will continuously miss the proper frequency. To compensate for this, the special page from the master unit 100 (during the fast recovery mode when it cannot do a page recovery as discussed above) is configured to have a slightly different periodicity such that the slave unit 110 scanning is going to time slide through the A and B trains of the standard Bluetooth page. Accordingly, the period of alternation between A and B trains may change from 1.28 s to 1.28±dt where dt is dependent on the scan window length used by the slave in recovery scan. This scheme will also work even if the slave unit 110 had left the slave recovery mode and entered the standby state. A second variant involves using a 11.25 ms scan window for the slave recovery process. This may be identical to the original Bluetooth page. The slave unit can then be accessed both via the standard Bluetooth page and via the new recovery page. This may have some impact on the way the page recovery scheme will select the position of additional Nrp retransmissions and the time schedule when new retransmissions are added. Also the duty cycle of the slave unit in recovery mode will increase (from 0.1% to about 0.9%) whereas the duty cycle of the master recovery scheme can be reduced (it will probably never have to reach to 5% but may end at about 1.5%). The latter scheme is also described in more detail in the second embodiment hereinbelow.
5. Slow Recovery
It is possible to continue increasing the number Nrp of recovery page slots as time progresses even beyond TFASTRECOVERY, TFASTRECOVERY may be set to e.g. 30 minutes. However, the duty cycle of the master unit will then potentially increase above 5% and may, hence, drain the battery for higher duty cycles. Therefore, when the fast recovery states have lasted for about TFASTRECOVERY=30 minutes, the units 100, 110 may be adapted to enter a slow recovery state. In this state, the number of recovery slots Nrp is frozen to 127 in this example; the master duty cycle therefore remains at a level of approximately below 5%. Since the time drifting still may take effect, the recovery scan state is adapted; not by increasing the scan window length, but by changing the scan repetition period such that a time sliding effect is enforced (but much faster than is caused by the drift from the clocks).
6. Slow Recovery Scan
When the units 100, 110 enter the slow recovery state, they both have a low duty cycle (about 5% for the recovery page and about 0.1% for the recovery scan). In a worst case situation, the recovery paging may occur just in between two scans as is shown in
If the scan repetition period is increased from 1280 ms to 1280 ms+ΔT, the scanning instances may start to slide with respect to the former anchor points. The maximum ΔT may be determined by the page window or:
ΔT≦Nrp
When in range and error-free conditions, the scanning window may overlap with the page window in maximally eight intervals or about 10 s. On average, it would then take 5 s before overlap occurs. If a smaller step ΔT is chosen, it may take longer before an overlap occurs. However, once an overlap occurs, there may be several overlap occasions in a row which increases the robustness.
It may be important that for the selection of the scan frequency, the scanning device adheres to the former anchor points of the previous connection based on the master clock information. For the scanning frequency selection, an update may be made every 1280 ms e.g. right in between the two anchor points. This is shown in
7. Slow Recovery Page
So far, it has been assumed that the paging in the slow recovery page state is identical to that in the last stage of the fast recovery page state. At instance k, two ID packets where sent at a fixed carrier frequency f(k) with a maximum repetition number of Nrp
Sending on three frequencies can be done in two different ways, see
ΔT=n×2.5 ms+1.25 ms
But since two overlaps in a row may be needed, there should be ΔT<ΔTmax/2.
8. Latency in Slow Recovery
Using a scan repetition, which is compatible with both the slow recovery page schemes (shown in
9. Compatibility with Bluetooth
Preferably, but not necessarily, the slave unit 110 in slow recovery scan state is also susceptible to normal Bluetooth paging. This can be accomplished by choosing a proper scan repetition interval. Normal Bluetooth paging applies page trains A and B. A single page train may comprise ID packets sent at 16 different frequencies during a 10 ms window. This page train A may be repeated for 1.28 s after which a different page train B may be used for 1.28 s, etc. The 1.5 ms recovery scan window may be configured to slide properly through the trains. The separation between two scans on the same page train may now be nominally 2.56 s (page trains alternate with a 1.28 s period). Therefore, the scan repetition period should be:
ΔT=m*10 ms+0.625 ms
where m is an integer
In the previous section, it was required that two consecutive scans would slide over a time length of 2.5 ms in total. The same coverage is also be obtained by four consecutive scans with a time stagger of:
ΔT=n×2.5 ms+0.625 ms
where n is an integer
Because of the reduced duty cycle in the slow recovery page state (as opposed to the 100% duty cycle during a Bluetooth page), in order to get overlaps in four consecutive instances, it may be needed that ΔT<ΔTmax/4.
The preferred slow recovery scan repetition period Tscan may be 5
obtained with n=12, which leads to m=3. The scan repetition period may then become:
Tscan=1280+30+0.625 ms=1310.625 ms
An alternative would be to use the original Bluetooth scan procedure. In that case, the slave unit could both be accessed by the standard Bluetooth paging, and by the new slow recovery page (which is the fast recovery page at maximum Nrp). For further details, the reader is referred to the second embodiment described hereinbelow.
During fast recovery, the time elapsed since the previous or last synchronization is still sufficiently small (a couple of minutes up to e.g. maximum 30 minutes) to be able to predict the timing in the both units 100, 110 (see
1. Fast Recovery Scan
When a loss of link 120 (see
During scan, the slave unit 110 may correlate the incoming signals against a known 68-bit access code related to the BD-ADDR of the slave unit 110. This is an ID packet corresponding to the Device Access Code (DAC) also used in conventional page scan. When the correlator output exceeds a threshold, the slave unit 110 may be configured to enter a recovery response sub-state which may be identical to the Bluetooth page response sub-state. The slave unit 110 may be configured to return an identical ID packet at the proper timing and await the reception of a FHS packet. The FHS packet may then re-synchronize the timing and frequency hopping of the slave unit 110. The slave unit 110 may remain in recovery scan until the link 120 is re-established or when a timeout TFASTRECOVERY has exceeded indicating the end of the fast recovery state. Assuming a worst case mutual drift of 40 ppm, the timeout TFASTRECOVERY may, preferably, be set to about 30 minutes. It should be appreciated that the fast recovery scan is similar to the conventional Bluetooth page scan. A difference is, however, in the clock used for determining the hop frequencies and wake up timing. In the conventional scheme, it is the native clock of the slave unit 110 that sets the timing; in the fast recovery procedure disclosed herein, it is the slave unit's estimate of the master clock that sets the timing.
2. Fast Recovery Page
When a loss of link 120 has been detected, the former master unit 100 enters the fast recovery page state. The initial timing of the fast recovery page may be based on the anchor point timing experienced during sniff mode. If an ID packet is sent at the anchor point, the slave unit 110 will receive this (provided the units 100, 110 are within range). The ID packet comprises the DAC of the slave unit 110 which is to be reconnected to the master unit 100. The frequency may be selected from the slave unit's page hopping sequence f(k), where k is the current clock of the master.
However, due to drift, the timing of ID transmissions, i.e. transmissions of at least one ID packet, sent by the master unit 100, and that of the scan window in the slave unit 110 will drift; after 5.6 ms/2 y there will not be an overlap anymore. For y=20 ppm, this may happen after e.g. 140 seconds. Therefore, in the recovery page state, the transmission window may need to be increased both before and after the anchor point, see
3. Compatibility Between Fast Recovery Scan and Bluetooth Page
Since the fast recovery scan state uses substantially the same parameters as the conventional Bluetooth page scan, a slave unit 110 in fast recovery scan could always be connected via the conventional Bluetooth page procedure.
4. Latency in Fast Recovery
It should be appreciated that the latency in the fast recovery procedure is only determined by the scan interval. When the units 100, 110 are in range and no errors occur, the maximum response time is 1.28 s. The average latency will be 640 ms, if a uniform distribution of the scan timing once the units come into range is assumed.
5. Slow Recovery
It is possible to continue increasing the number Nrp of recovery page slots as time progresses even beyond Nrp
6. Slow Recovery Scan
When the units 100, 110 enter the slow recovery state, they both have a low duty cycle (about 0.3% for the recovery page and 0.9% for the recovery scan). In a worst case situation, the recovery paging may occur just in between two scans as is shown in
If the scan repetition period is increased from 1280 ms to 1280 ms+ΔT, the scanning instances may start to slide with respect to the former anchor points. The 20 maximum ΔT may be determined by the page window or:
ΔT≦Nrp_max*10 ms=15*10 ms=150 ms
It should be noted that this sliding corresponds to a mutual drift of about 17% which is much larger than the mutual clock drift of several tens of ppm. Whether this clock drift is positive or negative may therefore be immaterial.
It should be noted that the selection of the scanning frequencies may still be based on the former anchor points and may be updated every 1280 ms (which is, however, not necessarily identical to every new scan event). Because of the time sliding of the scanning window with respect to the former anchor points, once in a while a scanning frequency in page hopping sequence may be skipped.
When in range and error-free conditions, the scanning window may overlap with the page window in maximally eight intervals or about 10 s. On average, it would then take 5 s before overlap occurs. If a smaller step ΔT is chosen, it may take longer before an overlap occurs. However, once an overlap occurs, there may be several overlap occasions in a row which increases the robustness.
7. Slow Recovery Page
So far, it has been assumed that the paging in the slow recovery page state is identical to that in the last stage of the fast recovery page state. At instant k, at least one ID packet is sent at a fixed carrier frequency f(k) with a maximum repetition number of Nrp
Sending on three frequencies may be done in two different ways, see
For option 1, a single scan window of 11.25 ms may cover 3 frequencies. In that case, the initial interval increases of ΔT=150 ms may be kept. If option 2 is applied, a single scan window only covers a single frequency. Three scans are required to cover 3 frequencies, see
1. Compatibility with Bluetooth Page
Preferably, the slave unit 110 in slow recovery scan state should also be susceptible to normal Bluetooth paging. This may be automatically guaranteed if ΔT is an integer multiple of 10 ms. In that case, the scan window will slide over the conventional Bluetooth page trains.
1. Latency in Slow Recovery
Using a scan repetition period which is compatible with both the slow recovery page schemes (shown in
In this disclosure, some embodiments have been described which may allow for improvements to the current Bluetooth sniff mode. A recovery procedure has been defined in order to improve the user experience of Bluetooth enabled units 100, 110 considerably. If the link is lost during Bluetooth sniff mode, the units 100, 110 may try to reconnect automatically using a low duty cycle recovery procedure. The recovery procedure may reuse the page frequency hopping sequence and the DAC ID packets as used in the conventional Bluetooth page procedure. A fast recovery may reconnect the two units 100, 110 within an average delay of about 640 ms when they are brought into range again. The fast recovery state may last for about 30 minutes (assuming a worst-case mutual drift of 40 ppm). During fast recovery, the duty cycle of the former slave unit 110 may be increased to 0.9%, the same as for a conventional page scan. The slave unit 110 may apply conventional page scan techniques (same page hopping scheme, same page scan window of 11.25 ms) but may, instead, use the clock information of the former master unit 100. In fast recovery scan state, the slave unit 110 may also be susceptible to conventional page messages using the standard Bluetooth page procedure. The duty cycle of the former master unit may increase over the 30 minutes window starting at 0.02% and rising to 0.3% at maximum. The master unit 100 may send ID packets including the slave's DAC. As time passes, the master unit 100 intensifies the repetition of ID transmissions. If no reconnection happens within e.g. 30 minutes, the units enter a slow recovery state. The duty cycle of the former master unit 100 is not further increased but remains at 0.3%. The duty cycle of the slave unit may remain at 0.9%; however, the scan period of the slave unit 110 may be increased to force a time sliding effect with the master recovery page transmissions. This may result in a longer delay when the units 100, 110 are within range (average delay in the order of 12 seconds) but will keep the units 100, 110 at low duty cycle. In slow recovery scan state, the slave unit 110 will also be susceptible to a conventional Bluetooth page messages. The slow recovery may carry on for more than 13 hours before FH synchronization is lost (again, assuming a worst-case mutual drift of 40 ppm). If the link has not been reconnected within this time, the units 100, 110 may return to conventional Bluetooth page scan (or turn off themselves, requiring a user action to start up again). In practical applications, the units would leave the slow recovery scheme much earlier (like 8 hours). The procedures described in this disclosure are quite general. However, if different clock accuracies are assumed, the final values will change. Larger inaccuracies will result in longer delays and/or larger duty cycles which translate into more power consumption.
II. Low-Power Radio System
As described in more depth herein, embodiments of the present invention relate to a Transmit Reference Spread Spectrum (TRSS) system which applies a frequency offset to separate the reference signal from the information signal. In contrast to conventional Direct Sequence Spread Spectrum (DSSS) systems where the spreading reference needs to be recreated in the receiver, in the TRSS system, the reference is embedded in the transmitted signal. Because the transmit signal contains the information and reference signals, acquisition and synchronization as required in DSSS systems are not necessary, and thus, the signal can be de-spread instantaneously irrespective of the processing gain. In conventional DSSS systems, a lengthy acquisition time is needed to synchronize the locally generated reference signal with the received signal, which also requires a larger processing gain. Moreover, in the TRSS system, the reference signal does not have to be extracted from the received signal, but de-spreading can be achieved directly by a mixing procedure as is later described. Finally, since the reference does not have to be recreated or extracted, the reference can be anything, including wideband noise. In these respects it is quite different from a pilot signal which could be embedded in a DSSS system.
The following Figures illustrate exemplary embodiments of TRSS systems, TRSS transmitters and TRSS receivers.
TRSS systems according to embodiments of the present invention may be used in uni-directional radio systems, including uni-directional short-range radio systems. One example of a uni-directional short-range radio system is a wake-up radio system 2055. A wake-up radio system includes a wake-up receiver 2200 and a transmitter 2100 communicable together via a wireless message. At reception of this message by the wake-up receiver 2200, which is transmitted by the transmitter 2100, the wake-up receiver 2200 will activate its host or other electronics associated with the wake-up receiver 2200. For example, referring back to
Another example of a uni-directional short-range radio system is an indoor positioning estimation system 2060 where one or more beacons 2090 are spread out in a building 2070 and broadcast positioning transmit messages to a recipient, which may be the cell phone 2050, other mobile devices 2050′, a controller 2080, or any other type of processing device. The beacons 2090 may include a transmitter 2100 of the present invention. The recipient (e.g., cell phone 2050′) receives the positioning messages via a receiver 2200 of the present invention that may be embedded in the recipient. Based on these positioning messages, the recipient can determine the transmitter's location inside the building 2070. For example, after receipt of the beacon signal, the recipient may retrieve information from the transmitted signal which indicates the beacon position (e.g., maps of the building, location of beacons, closest beacon position, etc.) or any other data desired to be transmitted to the recipient. In one embodiment, the beacon 2090 may optionally, include a receiver of the present invention (not shown) so that the recipient can transmit a reply message to one or more beacons 2090 upon recipient of the broadcast of the positioning messages or other messages from the beacons 2090.
Other applications are also realized with the present invention and the wake-up system 2055 and indoor positioning systems 2060 are only meant to be two exemplary applications of the present invention.
It should be noted that the transmitter and receivers presented in
In one embodiment, the reference signal can be generated at baseband or intermediate frequency (IF) and then be up-converted to RF or other desired frequency. The bandwidth (e.g. RF band) of the reference signal 2112 can be any desired bandwidth. In one embodiment, the reference signal 2112 can be any RF band, such as any industrial, scientific and medical (ISM) band (e.g., 2.45 GHz). In another embodiment, the reference signal 2112 can be any lower band, such as the FM band from 88 to 101 MHz. It should be understood that the reference signal 112 can be any band of frequencies and the present invention is not limited to only an RF band or FM band.
The reference signal 2112 is modulated by the information-bearing data signal, b(k), 2120, at multiplier 2125, resulting in a first modulated signal 2127. This data signal b(k) can use any modulation scheme, such as BPSK, QPSK, 16-QAM, etc. The modulated signal 2127 is then multiplied with signal 130 (e.g., cos (ωrft)) by multiplier 2140 where ωrf is the RF carrier frequency. Additionally, a frequency offset signal 2152 (e.g., a(t)*cos(ωrf+Δω)t) is created by multiplying signal 2150 (e.g., cos(ωrf+Δω)t) with reference signal a(t) 2112 by multiplier 2155, where Δω is the transmitted offset frequency. This resulting signal 2152 is then is combined with a signal 2142 (e.g., a(t)*b(k)*cos(ωrft)) by adder 2160, resulting in a transmit signal s(t) 2170. The transmit signal 2170 is represented by:
s(t)=b(k)·a(t)·cos(ωrft)+a(t)·cos(ωrf+Δω)t
where ωrf is the RF carrier frequency and Δω is the offset frequency. Typically, the RF frequency ωrf is in the order of 100 MHz to a few GHz, whereas the offset frequency Δω is in the order of a few kHz or MHz.
It is noted that the bandwidth BWa of the reference signal 2112 is much broader than the bandwidth BWb of the information-bearing data signal 2120 so that a spectrum spreading results. In one exemplary embodiment, the reference bandwidth BWa is some tens of MHz. Since the offset frequency is much smaller (e.g., in the order of 1 MHz or less), the spectra of the reference signal 2112 and combined data-reference signal almost completely overlap.
After the transmit signal s(t) 2170 is generated, the transmit signal s(t) 170 may then be transmitted through an antenna 2180 into surrounding space, which, in turn, may be received by a receiver 2200, which is discussed below with regards to
Compared with the transmit signal s(t), the received signal r(t) at the receive antenna 2205 will likely be attenuated because of the radio propagation. Furthermore, the transmit signal may be distorted due to multipath phenomena encountered on the radio propagation path. The received signal (or “received transmitted signal”), as referred to herein, relates to the propagated transmit signal, which may have been distorted.
In the receiver 2200, 2200′, the received signal (r(t)) 2207 proceeds to at least two multipliers, 2210 and 2230, for de-spreading and, optionally, demodulation. The exact location and configuration of these multipliers can be variable. For example,
x(t)=r(t)·cos(Δωt+φ)=={b(k)a(t)·cos(ωrtt)+a(t)·cos(ωrt+Δω)t} cos(Δωt+φ)
The frequency-shifted signal x(t) 2235 is multiplied with the received transmit signal r(t) 2207 by multiplier 2230 resulting in the de-spread signal (y(t)) 2240. It should be noted that de-spread signal 2240 (y(t)=r(t)2 cos(Δωt+φ)) produced by the receiver 2200 is a square of the received signal (r(t)2) multiplied by the frequency offset signal 2220 (e.g., cos(Δωt+φ)).
It should be further noted that the RF frequency (ωrf) does not occur in the receiver circuit, but instead, only the offset frequency (Δω). As such, there is no high-power RF local oscillator (LO) included or required in the receiver. Furthermore, the reference signal a(t) does not need to be regenerated in the receiver 2200, 2200′ for de-spreading or demodulation of the received signal 2207.
If only squaring is applied, the desired de-spread information-bearing signal 2120 will be located at the offset frequency Δω and this signal can be retrieved at IF. This may be advantageous since greater gains at IF can be obtained. In addition, the unknown or variable phase φ does not need to be retrieved.
The receiver 2200, 2200′ squares the received signal r(t) 2207. After squaring, the resulting signal 2232 is calculated as follows:
As shown in the equation above, the resulting DC component at the carrier frequency is:
and the component at the offset frequency (Δω) is b(k)·a2(t). Note that the signal component at the offset frequency (IF) is the information bearing signal including b(k). The signal at DC can be considered a self-interference signal. The components that are located at twice the RF carrier frequency (˜2ωrf) may be ignored and thus, can be filtered away (or integrated and dumped) using a filter or like device.
To prevent inter-carrier interference (e.g. from the self-interference signal located at DC), the spectrum of the squared reference a2(t) should resemble a Dirac impulse. To accomplish this, the reference signal 2112 (a(t)) should produce a constant amplitude after squaring. This can be achieved by using a constant envelope function, e.g. a binary function. In one embodiment, if the reference signal 2112 (a(t)) and the information-bearing signal 2120 (b(k)) are binary signals (e.g., +1, −1), the resulting square will be a constant: a2=1, b2=1. In the frequency domain, the DC component
of the demodulated data signal 2232 is fixed, whereas the de-spread information-bearing signal 2120 (b(k)) (i.e. after de-spreading in the receiver) arises at the offset frequency Δω. This information-bearing signal is thus extracted from the transmitted signal 2170 without having to generate a reference signal or via the use of a high-frequency local oscillator. Nonetheless, since the squared reference signal at DC is a spike, there is no cross-interference between the information-bearing signal 2120 and the reference signal 2112. Subsequent mixing with the offset frequency Δω will move the intermediate frequency (IF) portion of the signal to baseband where the information-bearing signal 2120 (b(k)) can be retrieved.
In one embodiment, the symbol rate of the de-spread information-bearing signal 2120 b(k) and the frequency offsets Δωi are based on 32 kHz (or other low frequency) which is also used for the real-time clock. The receiver then only needs a low power oscillator (LPO) with a 32 kHz reference from which all clocks in the receiver are derived. The low frequency of the oscillator allows for a low power oscillator to be employed and thus, the receiver becomes a low powered device. In one embodiment, the power of the low power oscillator allows for the peak power consumption of the receiver to be fully operated at 10-100 μW. Thus, applications, such as wake-up radios, do not need to be based on amplitude shift keying (ASK) or on-off keying, and can still apply spectrum spreading to obtain robustness in a multi-path fading and interference-prone environment.
It is noted that, in
In determining the transmit signal s(t) 2370 for the multiple channel transmitter 2300, a signal source 2310 first generates the reference signal 2312.
The reference signal 2312 is then sent to multiple different multipliers 2320, 2316 and 2318. At multiplier 2320, the reference signal 2312 is multiplied by the carrier frequency signal (ωrf) 2314, resulting in a carrier reference signal 2336. At a first channel branch 2322, the reference signal 2312 is multiplied by a first information-bearing signal (b1(k)) 2305 by a multiplier 2316 and the resulting signal 2326 is then multiplied by a first offset frequency signal (cos(ωrf+Δω1)) 2308 by multiplier 2321. At a second channel branch 2328, the reference signal 2312 is multiplied by a second information-bearing signal (b2(k)) 2307 by multiplier 2318 and the resulting signal 2330 is then multiplied by a second offset frequency signal (cos(ωrf+Δω2)) 2309 by multiplier 2323. The modulation schemes for b1(k) and b2(k) may not necessarily be the same. For example, the modulation scheme for b1(k) may be BPSK while the modulation schemes for b2(k) may be QPSK. Nonetheless, the signals 2332 and 2334 resulting from each channel branch 2322 and 2328 are combined with the carrier reference signal 2336 by adder 2340 resulting in the transmit signal (s(t)) 2370. The transmit signal (s(t)) 2370 is thus:
s(t)=a(t)cos(ωrft)+b1(k)·a(t)·cos(ωrf+Δω1)t+b2(k)·a(t)·cos(ωrf+Δω2)t
This transmit signal 2370 is transmitted through an antenna of the transmitter 2300 into space.
The optimal signal-to-noise ratio (SNR) is obtained when (Δωi)=πn/Tb where Tb is the symbol period of the data signal b(k) and n an integer (e.g., n=1, 2 for 2 channels).
Because of the non-linear, squaring operation of the received signal r(t), self-interference will arise due to the inter-modulation mixing of different components of r(t). To avoid inter-modulation products to end up in viable channels, combinations of additions and/or subtractions of the offset frequencies should not be equal to any of the offset frequencies themselves (i.e., Δωi±Δωj≠Δωk where i, j, k=1, 2, 3, . . . n for n parallel channels). This can, for example, be realized by selecting odd harmonics (e.g., 1 MHz, 3 MHz, 5 MHz . . . 2 m+1 MHz) for the offset frequencies for the channels. After squaring, the inter-modulation products due to self-interference will then end up at even harmonics (e.g., 0 MHz, 2 MHz, 4 MHz, 6 MHz, . . . 2 m MHz) which are not on any of the viable channels. Other combinations are possible that equally prevent inter-modulation.
As an example, a TRSS system operating in the FM broadcast spectrum (88-101 MHz) could have a RF center frequency of ωrf=98 MHz and a spreading bandwidth (BW) of 16 MHz. Assuming an information rate (R) of R=32 kb/s (based on the typical frequency of 32 kHz of a Real-Time clock), the offset frequencies could be chosen to be Δω1=5R=160 kHz, Δω2=8R=256 kHz, and Δω3=11 R=352 kHz. Inter-modulation products due to self-interference as the square thereof will arrive at f=3R=96 kHz, f=6R=192 kHz, and f=10R=320 kHz, each of which is adjacent to the desired signals. Furthermore, inter-modulation products caused by strong FM broadcast signals may arrive at f=200 kHz, f=300 kHz, f=400 kHz, and so on. The latter is based on the fact that the FM channel spacing is 100 kHz with at least a minimum separation of 200 kHz between adjacent FM channels. Also these inter-modulation products will be outside the bands of interest.
As another example, a TRSS system operating in the 2.4 GHz ISM spectrum could have a RF center frequency of ωrf=2441 MHz and a spreading bandwidth of 80 MHz. Assuming the same information rate of R=32 kb/s, the same offset frequencies can be selected, as indicated in the above example. All radio standards operating in the 2.4 GHz ISM band have a channel grid and spacing of at least 1 MHz. The first inter-modulation product after squaring will be at 1 MHz which is well above the offset frequencies presented.
For a wake-up system or other systems, a single channel may suffice. The channel will send a specific bit sequence that will wake-up the receiver. Only if this specific bit sequence is received will the receiver wake-up its host. A pilot channel could be added to support the synchronization in the receiver. Note that this pilot will be generated at baseband and follows the same modulation and combination with offset carriers as the information-bearing signals. Preferably, the data stream bp(k) for the pilot uses a very simple modulation scheme like BPSK.
In one embodiment, the pilot channel is self-decoding. The pilot is obtained using the correct offset frequency between the reference and the pilot channel. As such, the pilot is obtained immediately and with minimal power. For example, to obtain the pilot, there is no need for a local oscillator at the RF frequency and the pilot does not need to be generated in the receiver.
In an indoor positioning system or other systems, multiple of channels could be added that provide different kinds of data. For example, we could have one pilot channel at Δω1 which indicates that a beacon is present; a second channel at Δω2 may carry positioning information; a third channel at Δω3 may provide local maps that can be downloaded; and Δωn providing other information; and so on. A receiver for receiving multiple channels is shown in
One exemplary embodiment, however, may only contain a single mixer that can be tuned to each of the different offset frequencies Δω1, Δω2 and Δω3 For example, first, the receiver would tune to Δω1 to look for a pilot signal. Once found, the pilot signal can give important information for fine synchronization and timing. Then, the receiver would tune to the second offset frequency Δω2 to retrieve its position signal. Only in case the proper maps are not already in the host may the receiver tune to Δω3 to download one or more maps. Although three channels 2414, 2416, 2418 are illustrated in
The pilot signal 2408 may carry a simple one-zero sequence. This sequence should be easy to detect and can be a presence indication of an indoor beacon or a wake-up signal. The pilot 2408 can also provide symbol and/or frame timing information to the receiver 2400. Once found, this information can then be used by the receiver 2400 to demodulate one or more channels 2416, 2418.
Further, the pilot signal 2408 can be used to obtain the proper phase and frequency of the offset frequency Δω at the receiver 2400. At the transmitter 2300, an offset carrier of cos(Δωt) is applied. In the receiver 2400, a signal cos((Δω+δ)t+φ) can be recreated and for proper demodulation, δ=0 and φ=0. We could obtain this by applying an IQ mixer (i.e., multiplying the signal with cos((Δω+δ)t+φ) and sin((Δω+δ)t+φ) and perform frequency and phase tracking in the digital domain to compensate for δ and φ.
In addition to the phase and frequency synchronization, the pilot signal 2504 can also provide a reference for the symbol timing and the frame timing on the other channels. The rising and falling edges of the zero-one pattern can be used for bit timing purposes. For frame timing, the one-zero sequences, whose length corresponds to the frame length, can be inverted and alternated. For example, for a frame length corresponding to 6 pilot symbols (note that a pilot symbol may be longer than the data symbols on the other channels; the pilot rate may be 32 kb/s whereas the data rate may be 320 kb/s) two sequences would be needed: 101010 and 010101. By alternating the sequences, we obtain a frame sync at the boundary of two sequence: 101010, 010101, 101010, etc. Alternatively, the frame sync may be embedded on the information-bearing channels itself, i.e. a specific bit pattern on the information-bearing channel may indicate the start of a frame. In another embodiment, the frame timing may be indicated by a simple duplication at the frame boundary of a 1 or 0 bit in the alternating 1-0 sequence of the pilot channel.
The circuit results in a very low-current receiver that can operate below 1 mW levels. By properly dimensioning the system (selection of binary data and reference signals, off harmonic frequency offsets, all based on 32 kHz), a high-performance, robust system results. Self-synchronization is achieved by including a one-zero pattern as pilot channel.
III. Wake-Up Assisted Sniff Link with Sub-Rating
In a Bluetooth system, connection establishment can take quite some power since two devices need to find each other both in time and in frequency. A Bluetooth device in standby wakes up once every 1.28 s at a hop carrier frequency which changes every new wake-up event. If another device wants to make a connection, it shall page a unique ID continuously on a number of (up to 32 different) carrier frequencies until the standby unit responds. Therefore, most Bluetooth accessories, like a voice headset, a stereo headset or a watch, are placed in a sniff mode once the user turns on the devices. In the sniff mode, there is no time or frequency uncertainty. Instead, the slaves (“accessories”) are time and frequency locked to the master (i.e., mobile device, phone, etc.). Both master and slave wake up at pre-defined times and frequencies, which is very power efficient.
A conventional sniffed link 2600 is shown in
According to various embodiments, sub-rating refers to a standard feature described in the Bluetooth specifications. By applying sub-rating in the slave, the power consumption in the slave can be reduced considerably. Sub-rating maintains the latency from the slave to the master; that is, although the slave typically wakes up every N sniff events, if the slave needs to contact the master, the slave can do so at the next coming sniff event. As such, the latency from slave to master is maximally Tsniff although the sub-rate interval, Tsubrate, can be much longer. Yet, the latency from the master to the slave has been increased from Tsniff to Tsubrate. In order to get the same latency, the slave would need to listen every Tsniff, and which is the same the sniff situation that we started with. However, a combination of Bluetooth with the dfTRSS wake-up radio, which is described above in the section titled “LOW POWER RADIO SYSTEM,” solves this issue. As previously described in some embodiments, the low power radio system is a system that uses a Transmit Reference scheme with a LF frequency offset between the information signal and the reference signal and only this offset frequency, which is in the order of a few kHz to a few MHz, has to be recreated accurately in the receiver. The RF signal can be mapped directly to baseband by self-mixing. With this dfTRSS system, the peak current consumption can be lowered well below 0.1 mA. With a duty cycle of ˜1% (for example), the wake-up radio can operate on an average current close to 1 μA.
Referring to
The slave 2804 will respond to a master-initiated wakeup message 2704 containing a wake-up signal 2706 by contacting the Bluetooth master 2802 at the next coming sniff event 2705. Note that the auxiliary wake-up radio 2808 only triggers the slave 2804 to wake up. The auxiliary wake-up radio 2808 does not send the information that needs to be sent from the master to the slave, according to some embodiments.
An exemplary hardware configuration 2800 of the current disclosure is shown in
The power in the slave 2804 eventually depends on the sub-rate interval (ignoring the standby current in the wake-up radio). This sub-rate interval may be limited by the accuracy of the clocks used by the master and the slave. The clocks may drift somewhat. As a result, the timing of the master and slave “slides” or becomes out of sync relative to each other (“sliding effect”). Usually, this sliding effect is compensated periodically at the sniff events when the slave resynchronizes, but the interval must be small enough to allow for this compensation (i.e., the slave may not be time drifting out of the window where it would completely miss the master). The robustness may be improved by increasing the scan window, but this will result in a higher duty cycle (and thus, more power consumption) and eventually puts a floor to the achievable duty cycle. With a 40 ppm mutual drift between low power oscillators (LPOs) of the devices, for scan window of 100 μs, the maximum interval may be limited to 2.5 s. With Tsniff=1.28 s, this would limit the sub-rating factor, N, to 2.
For many accessories (e.g., a remote display, remote user interface (UI), etc.), an active connection only is needed if a triggering event occurs (e.g., an incoming call, a UI interaction by pushing a button, or other triggering occurrence). In these event-driven applications, a much longer interval may be employed in order to reduce the power consumption in the slave. In the above section entitled “AUTOMATIC RECONNECTION OF A PREVIOUSLY LOST CONNECTION IN A FREQUENCY HOPPING COMMUNICATION SYSTEM,” a method was described how to keep a master and slave virtually synchronized while no communication can happen to compensate for the mutual LPO drift. In the fast auto-reconnect mode, the slave is in a low duty cycle mode scanning during a short window at a fixed period (e.g. 1.28 s), whereas the master transmits paging messages at the same fixed period, but increasing the number of paging messages as time elapses and no reconnection is established. After 30 minutes in the fast auto-reconnect mode, the master will move to the slow auto-reconnect mode where the master stops increasing the number of paging messages, but stays at a fixed duty cycle.
Referring now to
This concept is very similar to the concept with sniff and sub-rating as described before and visualized in
A timing diagram 2900 of the combination of Bluetooth auto-reconnect with the wake-up radio is shown in
The Figures illustrate the architecture, functionality, and operation of possible implementations of systems and methods according to various embodiments of the present invention. It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by a human or special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used herein, the singular forms “a”, an and the are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms “comprises” and/or “comprising,” when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
Although specific embodiments have been illustrated and described herein, those of ordinary skill in the art appreciate that any arrangement which is calculated to achieve the same purpose may be substituted for the specific embodiments shown and that the invention has other applications in other environments. This application is intended to cover any adaptations or variations of the present invention. The following claims are in no way intended to limit the scope of the invention to the specific embodiments described herein.
This application claims benefit of priority as a continuation-in-part to the filing date of U.S. patent application Ser. No. 12/501,053, as filed on Jul. 10, 2009, and as a continuation-in-part to the filing date of U.S. patent application Ser. No. 11/954,106, as filed on Dec. 11, 2007, both of which are incorporated herein by reference in their respective entireties.
Number | Name | Date | Kind |
---|---|---|---|
5940431 | Haartsen et al. | Aug 1999 | A |
6061388 | Saulnier et al. | May 2000 | A |
6823186 | Salokannel et al. | Nov 2004 | B2 |
7340215 | Yokoshi et al. | Mar 2008 | B2 |
7454171 | Palin et al. | Nov 2008 | B2 |
7539457 | Lim et al. | May 2009 | B2 |
7809012 | Ruuska et al. | Oct 2010 | B2 |
20010024474 | Rakib et al. | Sep 2001 | A1 |
20020176445 | Melpignano | Nov 2002 | A1 |
20030058808 | Eaton et al. | Mar 2003 | A1 |
20030060161 | Park | Mar 2003 | A1 |
20040106424 | Yoshizawa | Jun 2004 | A1 |
20040147267 | Hill et al. | Jul 2004 | A1 |
20040229569 | Franz | Nov 2004 | A1 |
20050055374 | Sato | Mar 2005 | A1 |
20050160301 | Disser | Jul 2005 | A1 |
20060009240 | Katz | Jan 2006 | A1 |
20060111187 | Miyazaki | May 2006 | A1 |
20060128308 | Michael et al. | Jun 2006 | A1 |
20070047506 | Froehling et al. | Mar 2007 | A1 |
20070140154 | Chun | Jun 2007 | A1 |
20070140253 | Daigle | Jun 2007 | A1 |
20070173270 | Block et al. | Jul 2007 | A1 |
20070184880 | Frank | Aug 2007 | A1 |
20070242026 | Julian et al. | Oct 2007 | A1 |
20070275746 | Bitran | Nov 2007 | A1 |
20070287381 | Hulvey | Dec 2007 | A1 |
20070287542 | Miyazaki et al. | Dec 2007 | A1 |
20080070632 | Obuchi et al. | Mar 2008 | A1 |
20080102861 | Linsky et al. | May 2008 | A1 |
Number | Date | Country |
---|---|---|
1545069 | Jun 2005 | EP |
2820597 | Aug 2002 | FR |
WO 9828926 | Jul 1998 | WO |
WO 02089410 | Nov 2002 | WO |
03034645 | Apr 2003 | WO |
03065289 | Aug 2003 | WO |
WO 2004038938 | May 2004 | WO |
WO 2006013310 | Feb 2006 | WO |
2007069901 | Jun 2007 | WO |
2007127884 | Nov 2007 | WO |
2008006077 | Jan 2008 | WO |
2009033826 | Mar 2009 | WO |
2009078575 | Jun 2009 | WO |
2009127859 | Oct 2009 | WO |
Entry |
---|
Hall, E.S. et al., RF Rendez-blue: reducing power and inquiry costs in bluetooth-enabled mobile systems, Computer Communications and Networks, 2002. Proceedings. Eleventh International Conference on Oct. 14-16, 2002, Piscataway, NJ, USA, pp. 640-645. |
Bluetooth Sig: “Specification of the Bluetooth System: Master Table of Contents and Compliance Requirements; Covered Core Package Version: 2.1+EDR”, Internet Citation. Jul. 26, 2007, pp. 1-11. |
International Search Report; Feb. 15, 2011; issued in International Patent Application No. PCT/EP2010/065788. |
Written Opinion of the International Searching Authority; Feb. 15, 2011; issued in International Patent Application No. PCT/EP2010/065788. |
Dennis L. Goeckel et al., Slightly Frequency-Shifted Reference Ultra-Wideband (UWB) Radio, IEEE Transactions on Communications, Mar. 1, 2007, vol. 55, No. 3, IEEE Service Center, Piscataway, NJ. |
Zhang, Q. et al., Multiple Access Slightly Frequency-Shifted Reference Ultra-Wideband Communications for Dense Multipath Channels, Communications, 2007. ICC '07, IEEE International Conference, Jun. 1, 2007, pp. 1083-1088. |
International Search Report, corresponding to International Patent Application No. PCT/US2009/067288, dated May 6, 2010. |
Written Opinion, corresponding to International Patent Application No. PCT/US2009/067288, dated May 6, 2010. |
International Search Report; Dec. 23, 2010; issued in International Patent Application No. PCT/US2010/052223. |
Written Opinion of the International Searching Authority; Dec. 23, 2010; issued in International Patent Application No. PCT/US2010/052223. |
International Search Report; Dec. 29, 2010; issued in International Patent Application No. PCT/US2010/052221. |
Written Opinion of the International Searching Authority; Dec. 29, 2010; issued in International Patent Application No. PCT/US2010/052221. |
J.C. Haartsen, X. Shang, J.W. Balkema, A. Meijerink and J.L. Tauriz, “A new wireless modulation scheme based on frequency-offset”, 12th Annual Symposium of the IEEE/CVT, Nov. 3, 2005, pp. 1-7. |
Jing Wang and Japp C. Haartsen, “Performance of Transmit-Reference Radio System in Frequency-selective Fading Channels”, 12th Annual Symposium of the IEEE/CVT, Nov. 3, 2005, pp. 1-7. |
International Search Report and Written Opinion from Application No. PCT/EP2008/051142 mailed Jun. 5, 2008. |
U.S. Appl. No. 12/616,854, filed Nov. 12, 2009. |
U.S. Appl. No. 12/617,097, filed Nov. 12, 2009. |
International Preliminary Report on Patentability; Jan. 19, 2012; issued in International Patent Application No. PCT/US2009/067288. |
International Preliminary Report on Patentability; May 15, 2012; issued in International Patent Application No. PCT/US2010/052221. |
International Preliminary Report on Patentability; May 15, 2012; issued in International Patent Application No. PCT/US2010/052223. |
Number | Date | Country | |
---|---|---|---|
20100112950 A1 | May 2010 | US |
Number | Date | Country | |
---|---|---|---|
Parent | 12501053 | Jul 2009 | US |
Child | 12617213 | US | |
Parent | 11954106 | Dec 2007 | US |
Child | 12501053 | US |