WO2007093937A2

WO2007093937A2 - Radio system, master transceiver, radio transceiver and method of transmitting data in a network

Info

Publication number: WO2007093937A2
Application number: PCT/IB2007/050414
Authority: WO
Inventors: Mihai A. T. Sanduleanu; Neil C. Bird
Original assignee: Koninklijke Philips Electronics N.V.
Priority date: 2006-02-13
Filing date: 2007-02-08
Publication date: 2007-08-23
Also published as: WO2007093937A3

Abstract

The present invention relates to a radio system, radio transceiver device, master transceiver device and method for transmitting data between a plurality of the radio transceiver devices, wherein the radio transceiver devices are arranged as nodes of an asymmetrical network controlled by the functionality of the master transceiver device. Data is transmitted between the radio transceiver devices via a control master functionality by using a magnetic coupling. Thereby, an ultra low-power radio architecture can be achieved which satisfies simultaneous requirements of ultra low-power operation, small size and low cost.

Description

Radio system

FILED OF THE INVENTION

The present invention relates to a radio system, master transceiver device, radio transceiver device, and method of transmitting data between a plurality of such radio transceiver devices. In particular, the present invention relates to an ultra low-power radio system for Ambient Intelligence applications.

BACKGROUND OF THE INVENTION

Concepts that provide a vision for smart homes of the future, such as Ambient Intelligence, require an underlying wireless infrastructure suitable to transport a wide range of data types and content around the environment. The ideas of Ambient Intelligence are described for example in E. Aarts et al., "The New Every Day", 010 Publishers, ISBN 90- 6450-502-0 and relate to digital environments sensitive and responsive to the presence of people. Characteristic features are context awareness (environment has knowledge of what is happening), personalization (actions and responses are tailored to user needs), and adaptivity (environment will change in response to events). Specifically, an environment is described, in which the technology is embedded hidden in the background and demands the presence of a multitude of invisible distributed devices. The requirements of such an infrastructure are diverse and many thereof can be met with existing wireless standards and systems, while for others, on-going research at both system and circuit level is needed to complete the required broad portfolio of wireless links. The technical challenges stem not only from the large range of wireless data rates, but also from the combination of these with the varying distances over which these links will be required to operate, e.g., in-room and room-to-room. A specific requirement is for ultra low power wireless links to interconnect a multitude of transceiver (e.g. sensors) that will populate the environment.

In Ambient Intelligence applications, high data rate links of the order of Gb/s are required for video data streaming to an uncompressed HDTV (High Definition Television) display. Other applications are those where bulk data transfer is concerned. At the low end of the scale wireless links may be required, which support ubiquitous sensors of the environment. Typically, a sensor will transmit data occasionally, a few minutes per day, and each message will be in the order of a few hundred bits of information, e.g., a 64-bit address, a 128-bit data packet and/or a protocol/error correction overhead. Furthermore, there is a strong need to minimize energy per message.

In terms of new technologies to support the whole range of wireless data rates for Ambient Intelligence, two gaps can be identified. The first gap relates to a technology for providing multi-Gb/s data links in an in-room WPAN (Wireless Personal Area Network) system, and the second gap relates to a solution for ultra low power wireless links for various sensor applications. Thus, for many applications in the Ambient Intelligence and similar environments, further developments are required to meet cost, size and power consumption requirements.

SUMMARY OF THE INVENTION

It is therefore an object of the present invention to provide a radio system and method of transmitting data between a plurality of radio transceiver devices, by means of which low power operation can be combined with small size and low cost.

This object is achieved by a radio system as claimed in claim 1, a master transceiver device as claimed in claim 7, a radio transceiver device as claimed in claim 8, and a transmission method as claimed in claim 15.

Accordingly, a wireless infrastructure can be provided, which is capable of supporting features and applications of ultra low power radios with small form factor and low costs, such as wireless sensor networks. The provision of an asymmetrical network controlled by the master transceiver device and using magnetic coupling for radio transmission provides the advantage of minimum transmission powers at high frequency accuracy and match on the transmitter as well as the receiver side. Furthermore, the asymmetry caused by different types of modulations for transmission and reception leads to a reduced complexity of the transceiver circuits which can be based on simple timer functions in combination with receiver and transmitter functions.

The use of the same oscillator frequency for modulation and demodulation further reduces complexity, while the provision of a fixed oscillator frequency provides for short turn-on times, small size and low power consumption. The radio transceiver devices may comprise acoustic wave resonator means for providing a time reference. The use of acoustic wave resonator means does not require any phase-locked loop or other complex circuitries for ensuring accurate time references, and thus serves to keep implementation size small and enable short turn-on times at low power consumption.

A first type of modulation may be used for transmission from the radio transceiver devices to the master transceiver device, and a second type of modulation may be used for transmission from the master transceiver device to the radio transceiver devices. As an example, the first type of modulation may be an amplitude keying modulation, such as an on-off keying (OOK) modulation, and a second type of modulation may be a frequency shift keying modulation, such as FSK or OFSK (Orthogonal Frequency Shift Keying). Thereby, transmitter function can be kept simple at the radio transceiver devices.

The master transceiver device may have a higher sensitivity of reception than the radio transceiver devices, so that power requirements can be kept low at the radio transceiver devices.

The demodulation means of the radio transceiver devices may comprise a passive poly-phase shifter means for generating two quadrature output signals from the received signal. Due to the fact that the passive poly-phase shifter means do not consume any substantial power, low-power consumption can be achieved. Furthermore, the modulation means of the radio transceiver devices may be arranged to switch on and off a transmission frequency derived from the fixed oscillator frequency. In particular, the transmission frequency may be derived by frequency multiplier means as a multiple of the fixed oscillator frequency. Thereby, the RF (Radio Frequency) frequency necessary for the carrier can be generated at low circuit complexity. Furthermore, the demodulation means may comprise mixer means arranged to be operated in a subharmonic mode. This provides the advantage that internal disturbance or interference can be kept low.

As a specific, the radio transceiver devices may be implemented as tag devices. Further advantageous developments are defined in the dependent claims. BRIEF DESCRIPTION OF THE DRAWINGS

The present invention will now be described based on a preferred embodiment with reference to the accompanying drawings, in which:

Fig.l shows a schematic radio transceiver architecture according to the preferred embodiment;

Fig.2 shows a schematic functional diagram of a frequency detector of the preferred embodiment;

Fig.3 shows a schematic circuit diagram of a low-noise amplifier and automatic gain control stage of the preferred embodiment; Fig.4 shows a schematic circuit diagram of a current generation circuit for digital control of the preferred embodiment;

Fig.5 shows a schematic circuit diagram of a BALUN circuit of the preferred embodiment;

Fig.6 shows a schematic circuit diagram of a combination of low-noise amplifier, automatic gain control and BALUN circuit, as provided in the preferred embodiment;

Fig.7 shows a schematic circuit diagram and signal waves of a subharmonic mixer of the preferred embodiment;

Fig.8 shows a schematic circuit diagram of a frequency doubler and power amplifier circuit as provided in the preferred embodiment;

Fig.9 shows a schematic block diagram of a master-slave configuration with asymmetrical links, according to the preferred embodiment;

Fig.10 shows a schematic diagram indicating receiver sensitivity; and Fig.11 shows a schematic diagram indicating required noise figure values for different transmit power levels.

DETAILED DESCRIPTION

In the following, the preferred embodiment will be described on the basis of an ultra low-power (ULP) radio system comprising asymmetrical wireless links for an indoor environment.

There are two main categories for RFID (Radio Frequency Identification) systems available. These are near field communication systems (NFC) which employ inductive coupling of the transponder tags or Smart Label to the reactive energy circulating around the reader antenna, and far field communication systems (FFC) which couple to the real power contained in free space propagating electromagnetic plane waves. Near field coupling techniques are generally applied to systems operating in the low frequency (LF) and high frequency (HF) bands with relatively short reading distances well within the radian sphere defined by λ/2π (wherein λ corresponds to the wavelength of the transmission signal), while radiative far- field coupling is applicable to potentially longer read range UHF (Ultrahigh Frequency) and microwave RFID systems.

Active RFID tags are aimed at applications where a larger distance between the tag and the reader is required. By augmenting the passive RFID tag with a battery, the return range of the tag will increase. Nevertheless, the problem of data transfer initiation at the tag still remains. For other applications where the tag should initiate the data transfer, another ultra low power approach is required. The NFC approach is a symmetrical system that does not allow a duplex communication initiated at both ends. The 13.56MHz based NFC solutions can have a small form factor (few mm²) when the magnetic antenna is integrated on a SOI (Silicon-on-Insulator) wafer together with the active components.

The transfer of energy and data in NFC is based on vicinity inductive coupling. In near-field inductive coupling, the reader antenna loop and the tag coil windings establish a loosely connected "space transformer" resulting in power transfer across short bi- directional reading distances that are comparable to the actual physical dimensions of the field creating or transmitting antenna loop. Such coupling is deemed to be predominately magnetic if the physical dimensions of the transmitting loop are small compared to the systems operating wavelength. The space surrounding any antenna can be divided into two parts, depending on the properties of the created radiation. Due to the fact that changes in electromagnetic fields occur gradually, the boundary is not exactly defined. The primary magnetic field begins at the antenna and induces electric field lines in space. This area is termed the near field of an antenna. The zone where the electromagnetic field separates from the antenna and propagates into free space as a plane wave is termed the far field where the ratio of electric field E to magnetic field H has the constant value of 120π oder 377Ω. The approximate distance where this transition zone happens is λ/2π. For inductively coupled RFID systems when the far field has begun, inductive coupling is no longer possible. The radius r = λ/2π around the antenna therefore represents an insuperable range limit for inductively coupled systems. In the near field, the magnetic field strength attenuates according to the relationship 1/d³ (where d denotes the distance from the source of the field), i.e. the magnetic field intensity decays rapidly as the inverse cube of the distance between the reader antenna and the tag. In power terms, this equates to a dramatic 1/d⁶ reduction with distance of the available power to energize the tag. The magnetic field strength is thus high in the immediate vicinity of the transmitting coil, but a very low level exists in the distant far field. Hence, a spatially well-confined interrogation region or localized tag reading zone is created. This rapid attenuation of the energizing and data communication field with increasing distance is the fundamental reason why 13.56MHz passive RFID systems have a maximum reading distance in the order of Im. For the transmission of data at distances of about 10 m, powers in the order of Watts are needed.

Consequently, for sensor networking, where the communication can be initiated from either and of a link, and where distances up to 10m need to be covered, a conventional radio solution is required. Nevertheless, the radio transceivers should be initially configured in a simple way and ideally without wires. Therefore, a radio transceiver, e.g. tag, based on magnetic coupling is a cheap solution for such a purpose. As the power source is one of the factors which determine size and form factor of the devices of the radio system, low power operation should be achieved. At the circuit level, a complete radio transceiver with a power consumption of a few mW in the active mode is conceived. At system level, the protocol can be optimized for applications where the sensor needs to transmit data a short amount of time and for the rest of the time the radio transceiver can be in a power-down mode. However, if the metric for low power is optimum data rate, from the energy-per-bit perspective, a high data rate may be more energy efficient.

In the preferred embodiment, a message length of 288 bits and a turn-on time of about 200μs is assumed to be needed in general for synchronization purposes. The energy used to send the messages dominates at low data rates, whereas the turn-on time energy is dominant for higher transmission speed. The optimum may be a transceiver capable of data rates in the order of 50kb/s at active power consumption below 3mW. In this case however, a phase locked loop (PLL) radio transceiver with conventional technologies is almost impossible with this power consumption. From the cost and form factor perspectives, a PLL based radio transceiver needs a reference crystal and, therefore, the cost and the volume of the system increases. Therefore, the solution according to the preferred embodiment is achieved without a PLL. The absolute frequency accuracy can be acquired from a crystal oscillator with a SAW (Surface Acoustic Wave), BAW (Bulk Acoustic Wave) or resonant cavity type resonators. The oscillator can be always on and the only turn-on time required is the switching on of the receiver part (which may be about 2μs). This reduces the turn-on energy with respect to message energy and places the optimum at a higher data rate, e.g., 10Mb/s. This unconventional radio architecture can work with a resonator without high absolute accuracy but requires a frequency band were frequency errors can be tolerated. Then, as an example, the power consumption of the radio transceiver may be about 18 mW at a frequency of 17GHz.

The robustness of the system, the availability of a free band and the extra requirements of power, cost and size determine the choice of the frequency band of operation. Operation in the ISM (Industry, Science, and Medical) bands with approved devices does not require any license. By waving licensing requirements, these bands have been made generally accessible to virtually anyone. This is mainly why the ISM bands are so important for commercial and consumer applications. The availability of a "clean" band significantly simplifies the design of the radio transceivers (e.g. in terms of linearity requirements), and consequently allows low- power operation. As an example, the wavelength at 17GHz is 17.6mm, so that a λ/4 patch antenna could have dimensions in the order of 4 x 4mm².

In a conventional radio solution, the power consumption of the PLL is about 40% of the total consumption and turn-on times are in the order of tens of μs. In contrast thereto, the proposed ULP radio architecture according to the preferred embodiment has a short turn-on time. A time reference based on BAW resonators with a crystal oscillator will have a turn-on time of a few clock cycles. For example, in the 1 - 10GHz band, BAW resonators can operate with an absolute accuracy of 0.2 - 0.3% in a production process. This accuracy is good enough to allow "legal" transmission (e.g. in the 17GHz band) when frequency multiplication or sub-harmonic mixing is used.

Fig.l shows a schematic block diagram of the ULP transceiver architecture according to the preferred embodiment. The solution may be a System in Package (SiP) solution. The transceiver operates as a ULP radio transceiver in the nodes of an asymmetrical network controlled by a master device. The network handles the frequency accuracy problem and the different modulation formats for transmitter and receiver of the asymmetrical network. An IQ (in-phase/quadrature phase) demodulator uses a fixed oscillator 20 (at e.g. 8.575GHz). This fixed oscillator 20 comprises at least one BAW resonator and the expected absolute frequency accuracy can then be +/- 26MHz in the middle of the exemplary 17GHz band. An RF signal received via an antenna is filtered at a band pass filter 16 and supplied to a single-ended low-noise amplifier (LNA) 17 where it is amplified before being applied to an active BALUN 21 that provides power gain and conversion from a single-ended signal to a differential signal. In general, the term "BALUN" means balance-unbalanced, because the device is used to adapt a balanced device to an unbalanced one. In a balanced device the same voltage with respect to ground potential is applied to both terminals, while in an unbalanced device, a signal is only applied to one terminal and the other terminal is at ground potential. In the BALUN 21, the pair of output terminals is balanced, i.e., the currents are equal in magnitude and opposite in phase. The pair of input terminals connected to the output of the LNA 17 is unbalanced, i.e. one side is connected to electrical ground and the other carries the output signal of the LNA 17. The antenna may be arranged as a patch antenna and the band pass filter 16 may have a matching capability for providing impedance matching between the antenna and the input and output, respectively, of the transceiver.

The balanced output of the BALUN 21 is supplied to a passive poly-phase phase shifter or poly-phase filter 22 which generates two quadrature outputs (I, Q) required by subsequent complex mixers 23-1 and 23-2 in the different branches. The passive poly- phase filter 22 does not consume any power, since it is built solely with passive devices. The fixed oscillator 20 generates a differential signal (e.g. at 8.575GHz) and controls the local oscillator (LO) inputs of the complex subharmonic mixers 23-1 and 23-2. The subharmonic mixers 23-1 and 23-2 generate internally the 17GHz dyne required by the present homodyne architecture with constant oscillator frequency. According to the preferred embodiment, the transmitter part of the radio transceiver is supplied with the same oscillator frequency and thus operated with the same oscillator 20. In particular, the transmitter may be based on a simple turn-on/off in dependence on the transmission data. A frequency doubler 19 generates the RF frequency necessary for the transmission carrier. Thereby, a simple OOK (On/Off Keying) transmitter is achieved. OOK modulation is a special case of amplitude shift keying (ASK) modulation where no carrier is present during the transmission of one binary state, e.g., the zero state. OOK modulation can be implemented at high simplicity and low implementation costs. It has the advantage of allowing the transmitter to idle during the transmission of one binary state, therefore consuming less power. As a result of using this type of modulation for transmission, at least a 50% reduction in battery current drain can be achieved when compared to the power consumption of an FSK transmitter. The FSK transmitter must be switched on continuously during the time data is being transferred. Similarly, other amplitude shift keying (ASK) modulations could be used for transmission.

In the preferred embodiment, the received RF signal is a FSK signal which is down-converted at the subharmonic mixers 23-1 and 23-2 to zero intermediate frequency (IF). A gap in the middle of the FSK signal band (i.e. Orthogonal Frequency Shift Keying or OFSK) may reduce offset sensitivity of the receiver part. The channel filters 24-1 and 24-2 and the demodulator (i.e. frequency detector) 26 may be simple analog building blocks without large power consumption. If desired, the receiver part can be modified in a simple manner to receive ASK (Amplitude Shift Keying) modulation formats by generating the Euclidean distance between the I and Q down-converted signals. Gain control of gain control stages 25-1 and 25-2 is based on an amplitude information generated by an Euclidean distance computing circuit 15

(implemented by e.g. an analog processor circuit or the like) after squaring and low pass filtering (power information) or peak detection (amplitude information). The LNA 17 may as well have some gain control capability.

In Fig.l, the modulation circuit is simply depicted as a power amplifier 18 which is switched on and off based on the transmission data which controls a schematic switch connected to the power amplifier 18, to achieve OOK modulation.

Fig.2 shows a schematic functional diagram of the frequency detector 26 of Fig.l. Frequency detection can be based on a Differentiate And Multiply (DAM) algorithm, wherein both I and Q signals are supplied to a respective differentiating function or circuit 100-1 and 100-2, and the differentiated signals are cross-multiplied with the non- differentiated other signal in respective multipliers 110-1 and 110-2. The multiplier outputs are subtracted at a subtractor 120-1 to obtain the output signal of the frequency detector 26, which represents the frequency information taken from the transitions of the I and Q vectors. A small amplitude and phase mismatch will produce a small amplitude error at the detector output of the frequency detector 26. Amplitude and phase mismatch will thus generate a small gain error at the output. Therefore, the DAM algorithm is less sensitive to amplitude and phase mismatch. The NRZ (Non-Return to Zero) data at the output can be recovered after a limiting stage (not shown). The offset problem inherent to homodyne receivers can be handled as follows. Using OFSK data modulation, no information is generated around DC (Direct Current). A series capacitor together with some active stages may provide a high pass filter with a cut-off frequency of e.g. 10OkHz. The time constant may be an active time constant realized in the first stages of the filter.

Fig.3 shows a schematic circuit diagram of the LNA 17 with an automatic gain control (GAC) stage. The circuit of the LNA 17 in Fig.3 is an inductor degenerated, cascode LNA with resonant load (LC tank). The inductor L₂ provides input power matching and low noise operation together with weak selectivity. The output of the LC tank consisting of an inductor Ll and a capacitor Cl is applied to a super emitter-follower Q2 that conveys the RF signal in the collector current of a transistor Q3 without distortion of the RF signal. The need for a linear logarithmic gain control arises from the inevitable trade-offs between noise and distortion, as well as the more basic need to reduce an input signal of very large dynamic range to one of essentially constant amplitude. Linear IF amplifiers always have to provide variable gain, with a range of up to 80 dB. Using a detector cell at the output of the last IF stage, the control bias may be generated within a simple autonomous loop.

When the loop has settled in response to a change in signal amplitude, this bias provides a measure of the signal strength, the so-called Received Signal Strength Indication (RSSI). If the gain function is linear logarithmic, the RSSI voltage is a logarithmic measure of signal power. A possible gain-control function is one in which an increasing RSSI voltage reduces the gain. Thus, the RSSI output increases with signal power. Some part of this gain control can be done in the LNA 17 to reduce the DR (Dynamic Reconfiguration) requirements for the IF part. If the AGC control is present, a PTAT (proportional-to-absolute- temperature) voltage drop appears at the resistor R3 as a consequence of the difference between the currents Il and 12 of the respective current sources, which voltage drop can be expressed as follows:

VrTAT = VM L = V_T \nk (1)

wherein k = Ii/I₂, and V_T designates the threshold voltage of the transistors.

In general, the collector current I₀ of the transistor Q4 can be obtained based on the following equation:

By choosing the ratio k of the two currents such that:

A linear logarithmic gain control of the current Io can be achieved, as expressed in the following equation:

The voltage ΔV is a voltage drop on a resistor or a reference voltage. For generating the two currents Ii and I₂, the following circuit can be used. Fig.4 shows a schematic circuit diagram of a current generation circuit for the digital control. The current I₂ flows in the transistor Ql. The difference between the base- emitter voltages V_BE of the transistors Ql and Q2 is obtained at the resistor Rx. The DAC (Digital-to-Analog Conversion) current I_DAC generates a voltage drop I_DACRX which depends on an applied digital DAC code. A buffer transistor Q3 delivers the current I_DAC and controls the current I₂ in the transistor Ql. The two currents are related as follows:

Therefore, the linear logarithmic relationship can be expressed as follows:

Thus, a digital control based on a control word can be achieved by means of the circuit of Fig.4.

Fig.5 shows an example of an architecture of the BALUN 21. The current Io biases a transistor Ql at a current necessary for broadband operation and low distortion. The differential amplifier at the output senses the offset voltage between the two branches and adjusts the base voltage of a transistor Q3 so that the current of the transistor Q3 and a transistor Q2 will match the current Io. The resistive divider R2/R3 provides necessary bias for the base of the transistor Q3. Once the current Io is chosen, the total circuit is self-biased ensuring zero offset at the output. The input capacitor Cin decouples the bias of the base of the transistor Q3 from the input node.

The circuit operates at low voltages, e.g., 1.6 V. For this specific circuit, the maximum linearity condition, which can be defined by the 2nd-Order Input Intercept Point (IIP2), corresponds to the zero offset condition at the output. Due to the fact that no offset is provided, DC blocking capacitors between the BALUN 21 and the following stage can be avoided. The coupling capacitors introduce extra losses and extra phase errors in the signal path.

The output voltage of the LNA 17 is converted into a current by a resistor Rl and the incremental input resistance of the transistor Ql. Using a class AB operation of the circuit provides differential RF currents at the output. This is observed by taking the combination of the transistor Ql and the resistor Rl as a single transistor and the combination of the transistor Q3 and the other resistor Rl as well. The constant sum of based emitter voltages gives a constant product of the collector currents of the transistors Ql and Q3. The resistor Rl provides correct input resistance of the BALUN 21 for maximum power transfer from the LNA 17 to the BALUN 21. At the same time, it improves the linearity condition, e.g. 3rd-Order Input Intercept Point (IIP3), of the circuit and helps achieving matching between the transistors Ql and Q3.

Fig.6 shows the complete circuit combination with the LNA 17 and the BALUN 21. The output transistor of the LNA 17 provides the input current IO for the BALUN 21. The AGC control part controls the gain of the LNA 17 in a linear logarithmic ("linear in dB") fashion. This prevents overloading effects of the BALUN 21 and can help accommodating a large range of input voltages.

Fig.7 shows a schematic circuit diagram with signal waveforms of the balanced subharmonic mixers 23-1 and 23-2 of Fig.1, which form a basis of IQ demodulation. According to Fig.7, the subharmonic mixers 23-1 and 23-2 comprise a RF feed-through but no LO feed-through. For the mixing operation, the RF input needs to be multiplied with an LO signal. A possible option could be to switch-off the bias current I_BIAS in the tail of the differential stage of Ql and Q2. However, this option is not valid for speed reasons. A current source which generates the bias current I_BIAS cannot be switched off sufficiently fast for high-speed operation. A differential understage cannot be used for voltage room reasons. A better option is to route the bias current I_BIAS to the supply voltage current, thereby starving the RF transistor pair Ql and Q2. Transistors Q3 and Q4 are controlled by both LO phases LOo and L0i8o- Given the LO waveform, assumed sinusoidal, the momentary current I₀ can be deducted by inspection as shown in the center of the three waveform diagrams of Fig.7. The upper LO signal has a period of T = 1/CO_LO which changes with a period T' = 1/2G)_LO when referring to a momentary current I₀. The momentary current Io varies from 0 to I_B1AS and has the second harmonic of the LO signal in its spectrum, which can be expressed as follows:

Ut) = M¹ lM- Y ^l *_eJ2kω_{u)t (7)} _{π k}=-∞ \ - 4k

A simplifying assumption is a cosine like waveform of the momentary current io. Therefore, the momentary current Io can be written as a harmonic Fourier series, as follows:

2/ BIAS

J₀(O = ^-2 ∑-τ^—*cos(2kω_LOt) (8) π κ—1 ^tK 1

At maximum, the momentary current io has a value I_BIAS, thus starving the differential pair Ql and Q2. The RF signal will therefore not pass to the IF nodes IF- and IF+. At minimum, the momentary current io is zero and the RF signal will be sampled at the IF nodes IF- and IF+. The total tail current flowing in the differential pair is ii + i₂ which is the difference between I_BIAS and io. This difference can be expressed as follows:

I₁ (t) + i₂ (t) = ^{BIAs K }} + 4^^ ∑ -^- * cos(2kω_LOt) (9) π π k=i 4k -1 Given the tanh(x) relationship between the IF output voltage and the RF voltage, the following expression is obtained:

(π -2)I_B!ASR_{L |} ^ I^R^ 1 RF

IF(t) = ^ccos(2kω_LOt) tanh - (10) π π k=i 4k - l 2K

For simplicity, it can be approximated that tanh(x) « x - x³/3 + ... with only the linear term, neglecting the extra polynomial terms given the magnitude of RF signals. Therefore, the IF signal has RF feed-through in its spectrum, as follows:

_{IF{t) ≥} J^π -²^^BiAsRL _RF{t) _ 2 ^IBIASRL RF{I) ∑ — J — *cos(2kω_LOt) (11)

This is the reason why the subharmonic mixers 23-1 and 23-2 are considered singly balanced. Assuming a harmonic RF signal with phase and amplitude modulation, the RF signal can be expressed as follows:

RF(t) = A(t)cos(2ω_LOt + φ(t)) (12)

Assuming an RF frequency with double LO frequency or CO_RF = 2* CO_LO, the baseband signal after low -pass filtering of the IF signal can be expressed as follows:

(13)

Accordingly, the conversion gain/loss of the subharmonic mixers 23-1 and 23- 2 is as follows:

/"< _ ^ Bl ¹AS '"^■ L ( λ A\

°C ,sm gle-blanced ^~ _{τ 7} K ¹^J πV_τ and becomes the conversion gain when the DC voltage drop on the load resistor R_L is larger than πVχ/2. This condition may easily be achieved due to the large voltage room provided by this configuration. Voltage drops of few hundreds of millivolts are thus possible. In conclusion, the subharmonic mixers 23-1 and 23-2 operate in a subharmonic mode at half RF frequency and generating internally the double LO frequency. Due to this construction, the signal with double LO frequency is a common mode signal which cannot disturb the differential RF signal. Keeping the LO frequency at half RF frequency, the possibility of LO self-mixing and therefore LO offset generation at IF can be reduced in a homodyne configuration. A double balanced configuration can be derived from the singly balanced one.

Fig.8 shows a schematic circuit diagram of the frequency doubler 19 and the power amplifier 18 of Fig.1. The frequency doubler 19 can be based on the same configuration as the subharmonic mixers 23-1 and 23-2. The base of transistors Q4 and Q5 are connected to a DC voltage. They provide a path for the bias current I_BIAS for the time intervals when the current I₀ in the load Rl is zero. At the base of the transistor QI a signal at carrier frequency (e.g. 17GHz) is obtained. The transistor Ql amplifies a carrier and the selective circuit with the LC combination filters out the carrier frequency. For matching purposes with the antenna, a capacitive divider Cl, C2 can be used. The 0OK transmission data is applied to a MOS transistor Ml which provides the bias current to the transistor Ql. When the transistor Ml is switched off, no amplified carrier is provided at the output.

In the following, the question how to match transmission (Tx) and reception (Rx) frequency of the two ends of a link and how to accommodate for different modulation formats between Tx and Rx will be addressed.

The problem of absolute frequency accuracy and the different Tx/Rx modulation formats can be solved in a master- slave asymmetrical link system as shown in Fig.9.

According to Fig.9, slave radio transceiver devices ULPl to ULPn are provided which exchange signals with a master device MAS. The slave device ULPl transmits first an 0OK signal at frequency f . The master device MAS may be located in the same room and locks onto this signal and re-transmits on the ULP frequency f a FSK signal with the required data. Now, a link can be established. The sensitivity of the master device MAS (operated from mains) and the transmitted power of the master device MAS are higher in comparison to the slave devices ULPl to ULPn. In the same way, the master device MAS will sense the presence of the other devices ULPl to ULPn, which send on different frequencies f_Rjn. The master device MAS may allocate a time slot for the transmission, depending on the required data rate, and the time slot may also be allocated for the master device MAS to rejoin the link, if necessary, at a later point in time. Then, the master device MAS may drop out of the link or remain in a monitoring mode. In this way, the slave devices ULPl to ULPn contain minimum processing power, and may be simply timers with Rx/Tx functions. The protocol may be TDMA (Time Division Multiple Access) based and scheduling of the devices can avoid collisions and simultaneous operation of the slave devices ULPl to ULPn. Two slave devices can thus initiate a peer-to-peer virtual data transfer via the master device MAS.

The slave devices ULPl to ULPn should be initially configured in a simple way and ideally without wires. They may be based on a tag or an NFC using magnetic coupling, so as to provide a cheap solution for this specific purpose. They can operate autonomously at lower frequencies and provide a simple installation of the network. Studies of indoor propagation at 17GHz have shown that the received power level was found to decay with distance, at a rate equal to that of free space propagation or less, for line of sight (LOS) scenarios (damping coefficient α = 1.3 - 2). For non-LOS or obscured (OBS) scenarios, the received power levels were highly dependent on the obstructions in place. Path loss exponents for OBS scenarios tend to be α = 2 - 4 indicating that the received power levels generally decay at a rate greater than that of free space. In an OBS situation, the signal is further attenuated by about 2OdB when materials such as thick concrete and thermalite are present. Metal furniture attenuates the signal by about 7dB, and the effect of a person between the transmitter and the receiver has been shown to result in a similar attenuation. The delay spread is highly dependent on several factors including individual building and/or room construction, furnishing, and room size. The delay spread at a frequency of 17GHz amounts to σx = 10 - 20 ns for the majority of scenarios and can be improved by using directive antennas. To mitigate the effects of multipath reception, the bandwidth of the signal can be chosen as BW ≤ 1/σχ = 50MHz. Based on the above observations, the radio system according to the preferred embodiment can be implemented, for example, with a link radius r = 10m, OFSK modulation for reception (with BW = 20MHz) and OOK modulation for transmission, both at a data rate of 10Mb/s. The signal-to-noise (S/N) ratio at baseband can be 1 IdB (with 2dB implementation margin) for a block error rate (BER) of 10^~3 without forward error correction. The Tx/Rx antenna gain may be selected to a value of 5dBi in case of one patch antenna. Fig.10 shows a schematic diagram indicating receiver sensitivity Sin as a function of the noise figure (NF) of the receiver part of the radio transceiver device. For optimized implementation of the ULP radio transceiver, a trade-off needs to be made between the required NF of the receiver, the transmission power, and the fade margin. A 2OdB fade margin will include the LOS-OBS combination and the losses encountered in the matching filters and the antenna interfaces.

Fig.11 shows as diagram including these trade-offs. For a S/N at the baseband of 1 IdB, with a fade margin of 2OdB, the receiver noise figure NF should be better than 3.5dB at a transmit power of OdBm at the antenna. This can be achieved with conventional SiGe technologies. Apart from low power operation, an advantage of the low transmission powers in the order of only OdBm is that this helps to ensure signal confinement in a room, which will reduce interference between adjacent networks. In summary, a radio system, radio transceiver device, master transceiver device and method for transmitting data between a plurality of the radio transceiver devices has been described, wherein the radio transceiver devices are arranged as nodes of an asymmetrical network controlled by the functionality of the master transceiver device. Data is transmitted between the radio transceiver devices via a control master functionality by using a magnetic coupling. Thereby, an ultra low-power radio architecture can be achieved which satisfies simultaneous requirements of ultra low-power operation, small size and low cost. The proposed ULP radio system according to the preferred embodiment provides a simple implementation which is efficient in terms of energy-per-message. A power consumption in the order of 20 mW is required for an effective ULP implementation at a frequency of 17GHz. This power consumption target can be achieved with conventional SiGe processes. Thereby, a wireless infrastructure capable of supporting features and applications in an Ambient Intelligence Environment can be provided. However, of course the proposed ULP radio system may as well be implemented for other data communication system with similar requirements. The demands on such a wireless infrastructure may be extremely diverse and it is clear that ULP radio transceivers with small form factor and low cost will be essential for a plurality of applications, including wireless sensor networks. Higher data rates and short turn-on times are required to minimize the energy required to send a message. The robustness of the system, the availability of a free band and limitations on power, cost and size, point to the use of the 17GHz band. However, the system may as well be used in other frequency bands, without substantial loss of advantages.

The implementation of the frequency reference based on acoustic wave resonators leads to the advantages of small size, high quality factor Q, and absolute accuracy as high as 0.3% even in mass production. This is sufficient to ensure in-band transmissions, while the accuracy problem at system level is solved at the protocol level in a master slave asymmetrical link system. Using these approaches, an ULP system solution is obtained which satisfies simultaneous requirements of ultra low power operation (energy-per-message), small size and low cost. It is noted that the present invention is not restricted to the above preferred embodiment but can be implemented at different frequencies and different asymmetrical Tx and Rx modulation schemes without suffering any of the above advantages. The preferred embodiment may thus vary within the scope of the attached claims.

Finally but yet importantly, it is noted that the term "comprises" or "comprising" when used in the specification including the claims is intended to specify the presence of stated features, means, steps or components, but does not exclude the presence or addition of one or more other features, means, steps, components or group thereof. Further, the word "a" or "an" preceding an element in a claim does not exclude the presence of a plurality of such elements. Moreover, any reference sign does not limit the scope of the claims.

Claims

CLAIMS:

1. A radio system comprising: a) a plurality of radio transceiver devices (ULPl - ULPn) arranged as nodes of an asymmetrical network; and b) at least one master transceiver device (MAS) for controlling said asymmetrical network; c) wherein said radio system is arranged to perform radio transmission between said radio transceiver devices via said master transceiver device (MAS) by using magnetic coupling.

2. A radio system according to claim 1, wherein said radio transceiver devices

(ULPl - ULPn) comprise acoustic wave resonator means for providing a time reference.

3. A radio system according to claim 1 or 2, wherein a first type of modulation is used for transmission from said radio transceiver devices (ULPl - ULPn) to said master transceiver device (MAS), and a second type of modulation is used for transmission from said master transceiver device (MAS) to said radio transceiver devices (ULPl - ULPn).

4. A radio system according to claim 3, wherein said first type of modulation is an amplitude keying modulation and said second type of modulation is a frequency shift keying modulation.

5. A radio system according to any one of the preceding claims, wherein said master transceiver device (MAS) has a higher sensitivity of reception than said radio transceiver devices (ULPl - ULPn).

6. A radio system according to any one of the preceding claims, wherein said master transceiver device (MAS) is adapted to allocate time slots for said radio transmission between said radio transceiver devices (ULPl - ULPn).

7. A master transceiver device for use in a radio system according to any one of claims 1 to 6.

8) A radio transceiver device comprising: a) transceiving means for transmitting and receiving based on magnetic coupling; b) oscillator means (20) having acoustic wave resonator means for generating a fixed oscillator frequency, based on which a transmission frequency and a local oscillation frequency of said transceiving means are generated; c) modulation means (18) for modulating transmission data according to a first type of modulation; and d) demodulation means (22, 23-1, 23-2) for demodulating a received signal modulated based on a second type of modulation; e) wherein said fixed oscillator frequency is supplied to said modulation means (18) and said demodulation means (22, 23-1, 23-2).

9. A radio transceiver device according to claim 8, wherein said first type of modulation is an amplitude keying modulation and said second type of modulation is a frequency shift keying modulation.

10. A radio transceiver device according to claim 8 or 9, wherein said demodulation means comprise passive poly-phase shifter means (22) for generating two quadrature output signals from said received signal.

11. A radio transceiver device according to any one of claims 8 to 10, wherein said modulation means (18) are arranged to switch on and off a transmission frequency derived from said fixed oscillator frequency.

12. A radio transceiver device according to claim 11, further comprising frequency multiplier means (19) for deriving said transmission frequency as a multiple of said fixed oscillator frequency.

13. A radio transceiver device according to any one of claims 8 to 12, wherein said radio transceiver device (ULPl - ULPn) is implemented as a tag device.

14. A radio transceiver device according to any one of claims 8 to 13, wherein said demodulation means comprise mixer means (23-1, 23-2) arranged to be operated in a subharmonic mode.

15) A method of transmitting data between a plurality of radio transceiver devices

(ULPl - ULPn), said method comprising the steps of: a) arranging said radio transceiver devices (ULPl - ULPn) as nodes of an asymmetrical network; b) controlling said asymmetrical network by a central master functionality; and c) transmitting said data between said radio transceiver devices (ULPl - ULPn) via said central master functionality by using a magnetic coupling.