WO2007102113A2 - Low power master-slave radio links. - Google Patents

Abstract

A wireless transceiver comprising a wideband receiver and a narrowband transmitter, the wideband receiver being adapted to receive a first signal from a slave transceiver and to determine a reference frequency from a centre frequency of the first signal, the narrowband transmitter being automatically adjustable to transmit a second signal at the reference frequency for reception by the slave transceiver.

Description

LOW POWER MASTER-SLAVE RADIO LINKS

The invention relates to the field of wireless communication systems, and in particular that of asymmetric wireless communication systems utilising both wideband and narrowband radio technologies.

Broadly speaking radio transceivers can be designed to operate in either "narrowband" or "wideband" modes. Definitions of the term can vary, but for the purposes of this disclosure it should be understood that, in general, narrowband transceivers are those where the transmission bandwidth is approximately equal to the modulation bandwidth, and wideband transceivers are those where the transmission bandwidth is much greater than the modulation bandwidth. More generally, the expressions "wideband" and "narrowband" used in an asymmetric communication system context herein are intended to be interpreted in a relative sense, i.e. that the "wideband" transmitter or receiver operates over a wider bandwidth than the co-operating or complementary "narrowband" transmitter or receiver.

In terms of currently available radio systems, narrowband solutions tend to find their way into applications where low cost and low complexity are key requirements (e.g. 433MHz radio car key fobs for arming / disarming alarm systems, or devices for transferring modest quantities of data over short distances like radio microphones or baby monitoring devices etc). On the other hand, where higher data rates or greater user densities are anticipated, and where a more complex transceiver implementation can be justified, wideband solutions can often be the better choice.

In either case, irrespective of whether a narrowband or a wideband approach is adopted for a given application, two-way radio based communication systems in use today generally adopt a completely homogeneous approach. That is to say, within a given system design, the same transmission bandwidth and type is used at both ends of the link and in both communication directions.

In a wireless communication system consisting of a stationary base station and one or more mobile devices, the level of electrical power consumption is much more critical for the mobile devices than for the base station, since the mobile devices need to be able to function for long periods on battery power alone. The base station can, however, operate on mains electrical power. It is known that the power requirements for wideband wireless transmissions are in general much less than those for equivalent (i.e. equivalent data rate) narrowband transmissions. Conversely, power requirements for narrowband receivers are much less than those for wideband receivers, given the increased complexity of wideband reception and decoding.

To reduce the power consumption requirements of mobile devices in a wireless communication system, it would be advantageous for the mobile devices to each comprise a wideband transmitter and a narrowband receiver, while the base station comprises a wideband receiver and a narrowband transmitter. Such a system is known for example from US 6853835, which discloses such an asymmetric wireless communications system using both narrowband and wideband radio technologies.

In general, narrowband solutions are more compact and simpler to implement than wideband systems, and because of this simplicity usually operate at lower power levels. This makes them much more attractive than wideband systems for applications where very small size, ultra low power, or very low cost are key requirements. However, when designing or specifying a narrowband system, it is vital to ensure that anticipated offsets or drift in the frequency references that are invariably present in both transmitter and receiver, are very much less than the transmission bandwidth of the signal. If this criterion is not met then there is the very real danger that transmitted signals will not fall exactly within the pass band of the designated receiving device, and bit error rates (BER) will increase dramatically. When this happens most, or all, of the transmitted data can effectively be lost. The standard solution to this problem is to either make the receiver bandwidth sufficiently wide to accommodate the worst case frequency drift, or to ensure that only high accuracy, temperature compensated references are used throughout. The disadvantages with these approaches are that either the system becomes very sensitive to interfering signals, or the frequency reference components end up bulky and very expensive. In contrast, wideband systems, by their very nature, are not so susceptible to errors or drifts in their frequency reference components, making the need for expensive, highly accurate components unnecessary. However, wideband transceivers, especially on the receive side, do have a tendency to require a high degree of complexity including fast, high power correlators and code synchronisation loops for despreading. These features make current wideband solutions much less attractive for low power, low cost applications compared to their narrowband equivalents.

Prior art solutions where slave transceivers comprise a wideband transmitter and a narrowband receiver are not therefore amenable to the use of low cost components, due to the problems associated with frequency drift. Instead, highly accurate and stable (and therefore more costly) frequency reference elements would be required in the slave transceivers.

The object of the invention is to provide a master-slave configuration radio system in which narrowband transmissions are passed in one direction and wideband transmissions passed in the other direction. Lower cost frequency reference components can be comprised in the mobile slave transceivers without affecting the functionality of the system.

In a first aspect, the invention provides for a wireless transceiver comprising: a wideband receiver; and a narrowband transmitter, the wideband receiver being adapted to receive a first signal from a slave transceiver and to determine a reference frequency from a centre frequency of the first signal, the narrowband transmitter being automatically adjustable to transmit a second signal at the reference frequency for reception by the slave transceiver.

In a further aspect, the invention provides for a method of wireless communication between a slave transceiver and a master transceiver, comprising the steps of: transmitting a wideband signal from the slave transceiver; receiving the wideband signal at the master transceiver; determining by the master transceiver a reference frequency from a centre frequency of the wideband signal; transmitting by the master transceiver a narrowband signal for reception by the slave transceiver at the reference frequency.

Embodiments of the present invention will now be described by way of example and with reference to the accompanying drawings in which:

Figure 1 illustrates schematically an asymmetric narrowband / wideband system;

Figure 2 illustrates schematically a more detailed view of a master- slave wireless communication system of the invention; and

Figure 3 is a flow diagram of a two-way communication process between the master and slave transceivers.

Disclosed herein is a master-slave wireless communication system that utilises an asymmetric communication channel, passing narrowband transmissions in one direction and wideband transmissions in the other.

This approach removes the need for a highly accurate and stable frequency reference element in the "slave" end of the link where cost and physical space are at a premium, at the expense of a slight increase in complexity in the master.

In figure 1 a schematic illustration of an asymmetric narrowband / wideband system is shown, comprising a master transceiver 1 and a slave transceiver 2. The master transceiver 1 comprises a wideband receiver 3 and a narrowband transmitter 4, while the slave transceiver 2 comprises a narrowband receiver 5 and a wideband transmitter 6. Narrowband transmissions 8 from the narrowband transmitter 4 of the master transceiver 1 , communicated via a diplexer 10 and an aerial 14, are received by the narrowband receiver 5 at the slave transceiver 2, via an aerial 15 and a diplexer 1 1. Spread spectrum transmissions 9 are transmitted from the wideband transmitter 6, via the diplexer 1 1 and the aerial 15, and received by the wideband receiver 3, via the diplexer 10 and aerial 14. A frequency reference source 13 in the slave transceiver 2 provides a signal to both the wideband transmitter 6 and the narrowband receiver 5. A frequency reference source 12 in the master transceiver 1 is synchronised with the frequency reference source 13 of the slave transceiver such that the wideband receiver 3 decodes the spread spectrum transmission 9 received from the wideband transmitter 6 correctly. This frequency reference source 12 is then used by the narrowband transmitter 4 to generate a narrowband transmission 8 with a centre frequency that is matched to that of the narrowband receiver 5.

The use of a common frequency reference source 13 in the slave transceiver 2 for the wideband transmitter 6 and the narrowband receiver 5 ensures that any drift in frequency of the source 13 equally affects frequencies in the receiver 5, transmitter 6 and the chipping clock for PN code sequence generation.

To illustrate the operation of the system, a situation where data is to be sent from slave to master will now be described. The slave wideband transmitter 2 may comprise a very simple wideband CDMA transmitter in which incoming data is first spread using a known pseudo noise (PN) code sequence, and then modulated onto an rf carrier and transmitted. The nominal centre frequency of this wideband "spread spectrum" transmission 9 will depend upon the accuracy of the slave frequency reference 13, which may not be particularly accurate. The slave transceiver 2 preferably uses a common frequency reference 13 to generate both the rf carrier frequencies for up and down mixing and the chip clock for generating the PN code sequence. In other words, any change in rf carrier frequency due to frequency reference drift must also cause a corresponding change in the chip clock rate. The spread spectrum signal 9 is received by the wideband receiver 3 in the master transceiver 1. The input bandwidth of this receiver 3 can be arranged to be sufficiently wide to allow for the worst-case frequency drift or offset generated by the wideband transmitter 6 in the slave.

Using the known PN sequence for the system, the master can now synchronise its chipping clock, using this to demodulate and de-spread the incoming signal in the usual way. Since, as stated above, the PN code generator in the slave and its rf carrier frequency are related, by implication the master 1 can now determine the centre frequency of the spread spectrum transmission from the value of its own chipping clock, as adjusted after synchronisation. The master then determines a reference frequency value from this centre frequency. The reference frequency value is preferably stored in memory 28 in the master for later use. When the master needs to transmit data back to the slave it uses this stored reference frequency value to set the centre frequency of its own narrowband transmitter 4. Since this narrowband transmission 8 is now guaranteed to be on the right rf centre frequency to be detectable by the slave's narrowband receiver 5, data is successfully demodulated in the slave with a low bit error rate (BER). In a system that comprises multiple slave devices, separate reference frequency values for each slave in the network will need to be stored in the master's memory 28 for later access, and each slave transceiver 2 may be identified by unique identification codes stored along with each reference frequency value.

Preferably, the centre frequency of the slave as measured by the master transceiver from adjustment of the master PN code chipping clock rate is equal to the reference frequency value subsequently used for transmission by the master transceiver. However, the reference frequency value may alternatively differ from the centre frequency according to a predetermined relationship, being for example a predetermined frequency shift, provided that there is a defined and fixed relationship between the transmit and receive frequencies of the slave transceiver.

The operation of the communication system relies on the fact that wideband systems are more robust against a carrier frequency offset than equivalent narrowband systems. This is essentially a result of the fact that as the transmitted signal bandwidth increases, the proportion of uncertainty due to frequency errors reduces. As an example take a situation where we want to send a 5kb/s data signal across a radio link. In a narrowband system, even using a high modulation index we may only need to use say 2OkHz of transmission bandwidth to convey the data with a very low BER. However, if we are operating at a carrier frequency of say 868MHz and we are using a frequency reference with an overall accuracy of for example 25ppm then this translates to a possible frequency deviation from nominal of around 22kHz, or approximately 1 10% of the total transmission bandwidth. Since this drift is large compared with the transmission bandwidth, and also in all probability enough to shift the signal completely outside the channel bandwidth of the receiver, it is likely that in the worst case almost no data will be successfully decoded at the receiver.

However, let us now consider the wideband case. Let's start again with the same 5kb/s data stream, now mixed with a PN code sequence running at a chip rate for example 20 times greater than the basic data rate. This results in a total transmitted signal bandwidth of around 100kHz. With the same frequency reference accuracy as before (i.e. 25ppm), a 22kHz drift only represents around 22% of the total transmission bandwidth. Since we are operating with a wideband CDMA system anyway, we can now easily afford to increase the bandwidth of our receiving system by 22% to say 122kHz to fully accommodate even the worst-case drift. To help matters further, if we are considering a master- slave topology, then it is probable that the master, unlike the slave, can afford to employ a highly accurate, temperature compensated frequency reference. This being the case, the majority of the frequency error within the system may safely be accommodated within the slave transceiver 2. The communication system described above also benefits from the fact that wideband transmitters are much simpler to design and implement than wideband receivers. This is because the transmitter only has to carry out three relatively simple operations: mix the incoming data with a chipping code, use the result to modulate an rf carrier and transmit the resulting signal. In contrast, the receiver has to not only receive the rf signal, but also demodulate it and then feed the result into a set of correlators to compare the match with its own locally generated PN code. The receiver has to ensure that the chip rate is adjusted to match the chip rate of the transmitter and implement synchronisation loops to ensure that the phase of the PN code generator is in-phase with the incoming data stream. This additional complexity inevitably adds to the cost and power consumption of such a wideband receiver.

A more detailed representation of a communication system according to the system described above is shown in figure 2. In this, components comprising the wideband transmitter 6 and wideband receiver 3 are further illustrated.

The frequency reference source 13 in the slave transceiver 2 generates a signal that is directed to a frequency divider 22, which provides a clock frequency for a PN code generator 23. A mixer 24 spreads the incoming data 34 using the PN code, which is then sent to the modulator 21. The modulator 21 generates a spread spectrum signal with a centre frequency provided by the frequency reference 13, the signal being transmitted to the master transceiver 1 via a diplexer 1 1 and aerial 15. The spread spectrum signal is received by the aerial 14 of the master transceiver, and input to the wideband receiver 3 via a diplexer 10. The signal is demodulated by the wideband demodulator 25, under control of the control gate 26. The sync circuit 29, PN code generator 27, control gate 26, correlator 30 and demodulator 25 work together to demodulate and correlate the signal, as is known in the art.

Using a frequency reference source 12, which preferably has a higher accuracy and/or stability than the frequency reference source 13 of the slave transceiver, the master receiver adjusts the frequency and phase of the PN code generator 27 to de-spread the incoming signal. The centre frequency of the incoming signal is calculated based on the adjustments to the PN generator. The centre frequency may be stored in memory 28, together with an identifying code corresponding to the particular slave transceiver the frequency relates to. This stored frequency may be updated periodically, based on transmissions received from the corresponding slave transceiver.

The synchronisation circuit 29 and correlator 30, fed by the adjusted PN code 27 frequency, generate a data packet and feed this to the data out line 32.

When the master transceiver 1 needs to transmit a signal back to the slave transceiver 2, for example in response to a received signal, the centre frequency of the narrowband transmitter 4 is adjusted to match that of the stored value for that slave transceiver 2. A data packet from the data input 31 is encoded, for example by frequency shift key encoding, and the modulated signal is transmitted via the diplexer 10 and aerial 14 for reception by the narrowband receiver 5 of the slave transceiver 2.

Shown in figure 3 is a flow chart diagram of the process whereby the master and slave transceivers may be configured to communicate. Firstly, at step 35 the slave transceiver 2 powers up its (relatively low accuracy) reference oscillator, which generates the frequency reference 13. This frequency reference 13 is used, at step 36, to clock an internal PN code generator 23. At step 37, the slave transceiver generates a network ID packet plus data and spreads the signal using the PN sequence generator 23. At step 38, the resulting wideband signal is modulated on to a carrier from the reference oscillator and transmitted. At step 39, the slave goes into a narrowband receive mode, where the reference oscillator act as the receiver local oscillator. The slave transceiver 2 determines, at step 40, whether an incoming signal is detected. If so, the incoming narrowband signal 8 is demodulated and decoded by the narrowband receiver 5 at step 41. The slave transceiver then goes to a sleep mode, at step 42, until the next wake-up period. If no signal is detected at step 40, the slave transceiver directly enters the sleep mode at step 42.

The master transceiver 1 , at step 43, listens for a wideband transmission from a slave transceiver 2. At step 44, if an incoming signal 9 is detected, the master transceiver proceeds to step 45 to deal with the signal; if not, the master transceiver carries on listening (step 43).

At step 45, the master transceiver adjusts the frequency and phase of its own internal PN code generator 27 to de-spread the incoming signal 9. Only when the frequency and phase are matched to that of the PN code generator 23 in the slave transceiver can the incoming signal 9 be de-spread accurately. At step 46, the master transceiver calculates a centre frequency of the incoming signal 9 based on these adjustments to the PN code generator 27. This centre frequency, defined as the slave reference oscillator value, will ideally match that generated by the frequency reference 13 in the slave transceiver. At step 47, the slave reference oscillator value is stored in memory 28 (step 49).

If the master transceiver needs to transmit a signal back to the slave transceiver, decided at step 48, a data packet is first generated and encoded ready for transmission (step 50). At step 51 , the master transmitter 4 centre frequency is adjusted to match the slave reference oscillator value held in memory 28. At step 52, the data packet is modulated on to the narrowband transmitter 4 carrier signal, and at step 53 the resulting narrowband signal 8 is transmitted for reception by the slave transceiver 2.

The invention could be used in any application where a two-way radio communication system is needed that can exploit the advantages of very small, low cost slave units. Although there are many applications that could take advantage of this idea some specific examples might include: medical/fitness control and monitoring; domestic/office control and monitoring; logistics / track and trace; automatic ticketing; and automotive monitoring.

Other embodiments are intentionally within the scope of the appended claims.

Claims

1. A wireless transceiver (1 ) comprising: a wideband receiver (3); and a narrowband transmitter (4), the wideband receiver being adapted to receive a first signal (9) from a slave transceiver (2) and to determine a reference frequency from a centre frequency of the first signal, the narrowband transmitter being automatically adjustable to transmit a second signal (8) at the reference frequency for reception by the slave transceiver.

2. The wireless transceiver (1 ) of claim 1 in which the first signal (9) is a wideband spread spectrum signal.

3. The wireless transceiver (1 ) of claim 2 in which the first signal (9) is modulated according to a Code Division Multiple Access scheme.

4. The wireless transceiver (1 ) of claim 1 in which the reference frequency is dependent upon and varies in a predetermined relationship with a chip rate frequency of a code generator (27) comprised within the wideband receiver (3).

5. The wireless transceiver (1 ) of claim 4 in which the code generator is a pseudo noise generator (27).

6. The wireless transceiver (1 ) of any one of claims 4 and 5 in which the wideband receiver (3) is adapted to automatically adjust the chip rate frequency to alter the reference frequency value towards a slave reference frequency value generated by the slave transceiver (2).

7. A method of wireless communication between a slave transceiver (2) and a master transceiver (1 ), comprising the steps of: transmitting (38) a wideband signal (9) from the slave transceiver; receiving (44) the wideband signal at the master transceiver; determining (46) by the master transceiver a reference frequency from a centre frequency of the wideband signal; transmitting (53) by the master transceiver a narrowband signal (8) for reception by the slave transceiver at the reference frequency.

8. The method of claim 7 in which the wideband signal (9) is encoded according to a Code Division Multiple Access scheme.

9. The method of any of claims 7 and 8 in which the narrowband signal (8) is frequency shift key modulated.