Transmitter with means for complementarily scaling the input and output signals of a D/A converter

[0001] 1. Field of the Invention

[0002] The present invention relates generally to transmission systems and more specifically to a technique for improving the carrier-to-noise ratio (CNR) of radio transmissions. The present invention is particularly suitable for improving the CNR value of a code division multiple access (CDMA) transmitters such as cell-site base transmitters.

[0003] 2. Description of the Related Art

[0004] In cell-site CDMA transmitters of a cellular mobile communication network, individual spread spectrum channel signals are synchronised and multiplexed (or digitally summed) into digital amplitude data. This amplitude data is converted by a digital-to-analog converter to an analog amplitude signal, which is up-converted, power-amplified and then transmitted. In order to prevent the digital-to-analog converter from being overloaded, the maximum amplitude of each transmit channel signal is determined in advance so that the total amplitude of the multiplexed signals corresponds to the upper limit of the digital-to-analog converter. However, the performance of the transmission system is determined by the CNR requirement of the largest signal level to be power amplified prior to modulation onto a radio frequency carrier. As a result, when not all individual channels are multiplexed during low traffic periods, the carrier-to-noise ratio of the system becomes lower than the required value.

[0005] One approach to this problem is to use a digital-to-analog converter for each channel signal to produce a multiplex of analog amplitude signals, instead of using the single digital-to-analog converter for producing a multiplex of digital amplitude signals. However, this approach is not an economical implementation and requires additional circuitry that ensures orthogonality between any pair of analog spread spectrum channel signals.

[0006] It is therefore an object of the present invention to provide a transmitter and a communication method for improving the carrier-to-noise ratio of a signal that has been subjected to a digital-to-analog conversion process prior to transmission.

[0007] According to a first aspect of the present invention, there is provided a transmitter comprising average power level calculation circuitry for determining the time-average power of a digital amplitude signal, and conversion circuitry for scaling the digital amplitude signal according to a first scale factor, converting the scaled digital amplitude signal to an analog amplitude signal, scaling the analog amplitude signal according to a second scale factor, and control circuitry for complementarily varying the first and second scale factors according to the time-average power.

[0008] According to a second aspect, the present invention provides a spread spectrum transmitter comprising a multiplexer for multiplexing a plurality of spread spectrum channel signals to produce a digital amplitude signal, average power level calculation circuitry for determining the time-average power of the digital amplitude signal, and conversion circuitry for scaling the digital amplitude signal according to a first scale factor, converting the scaled digital amplitude signal to an analog amplitude signal, scaling the analog amplitude signal according to a second scale factor, and control circuitry for complementarily varying the first and second scale factors according to the time-average power.

[0009] According to a third aspect of the present invention, there is provided a communication method comprising the steps of determining the time-average power of a digital amplitude signal, scaling the digital amplitude signal according to the time-average power, converting the scaled digital amplitude signal to an analog amplitude signal, scaling the analog amplitude signal according to the time-average power complementarily to the scaling of the digital amplitude signal and transmitting the scaled analog amplitude signal.

[0016] Referring now to FIG. 1, there is shown a CDMA (code division multiple access) transmission circuit according to a first embodiment of the present invention for a communication system such as a cell-site base station of a cellular mobile communication network. It is seen in FIG. 1 that a plurality of individual digital spread spectrum channel signals (either speech or data) are multiplexed (or “digitally summed”) in a multiplexer 1 into digital amplitude data. Each of the input spread spectrum signals is a signal that has been spread (scrambled) with short and long spreading codes and time-synchronized with other spread spectrum signals in a manner well known in the CDMA art.

[0017] According to this invention, the digital amplitude data from the multiplexer 1 is supplied to an interpolator 2 and an average power calculator 3 simultaneously. Average power calculator 3 calculates a time-average value of the total power level of the multiplexed amplitude data and feeds a control circuit 4.

[0018] Interpolator 2 provides interpolation on N bits of the amplitude data and produces M bits of multiplexed amplitude data, where M is greater than N. For example, if the N-bit multiplexed amplitude data has a sequence of bits {A₁, A₂, . . . , A_N} occurring at T-intervals, as indicated by circles in FIG. 2, the interpolator 2 calculates the amplitude difference between successive bits and determines the gradient of a slope as (A_i+A_i+1)/T (where i=1, 2, . . . N) during each bit interval T. If one bit is interpolated between successive input bits as indicated by dots in FIG. 2, the interpolator 2 produces an M-bit output sequence in which the individual bits occur at interval T′ equal to T/2.

[0019] The output of the interpolator 2 is sent to a bit shifter 5. Bit shifter selects a sequence of a predetermined number of “L” bits from bit positions of each input M-bit sequence which are specified by the control circuit 4, where the integer L is larger than N but smaller than M. As shown in FIG. 3, the bit shifter 5 may be implemented with an M-stage shift register in which default bit positions (i+1) to (i+L) are defined. According to the average power of the multiplexed amplitude data relative to a reference power level, the control circuit 4 determines the scale factor of the input signal to the D/A converter 6 and determines the bit positions of an output L-bit sequence within an input M-bit sequence. Control circuit 4 specifies the bit positions of the output L-bit sequence and instructs the bit shifter 5 to select an L-bit sequence from the specified bit positions of the M-bit sequence.

[0020] The L-bit sequence selected by the bit shifter 5 is converted to a corresponding analog amplitude signal by a digital-to-analog converter 6. A gain-controlled amplifier 7 is connected to the output of D/A converter 6 for amplifying the analog amplitude signal according to a scale factor specified by the control circuit 4. The output of the gain-controlled amplifier 7 is modulated onto a radio frequency carrier in an up-converter 8. The modulated carrier is power-amplified by a transmission power amplifier stage 9 and transmitted from an antenna 10.

[0021] In order to meet the carrier-to-noise ratio requirement of the transmitter at all levels of transmissions, the bit shifting of shifter 5 and the gain of amplifier 7 are complementarily controlled by the control circuit 4 in a compression-expansion mode or in an expansion-compression mode depending on the average power of the amplitude data relative to a reference power level as follows.

[0022] Control circuit 4 determines whether the calculated average power level is higher or lower than a reference (default) value which may be a maximum average power level, a minimum average power level or an intermediate value that lies anywhere between the maximum and minimum values. Assume that an intermediate value is used as the reference value. In this case, the gain-controlled amplifier 7 is initialized so that, when the average power level of the output of the multiplexer 1 is equal to the reference value, the carrier-to-noise ratio of the transmitter is controlled at the required value.

[0023] If the control circuit determines that the calculated average power is lower than the intermediate reference value, it produces a down-shift control signal that causes the bit shifter 5 to select from an interpolated M-bit sequence an L-bit sequence {(i+k+1) through (i+k+L)} that is shifted downwards by “k” bits with respect to the default bit positions as illustrated in FIG. 3, where the integer “k” represents the difference between the calculated average power and the intermediate reference power level. Then, the control circuit 4 proceeds to increment the gain of the amplifier 7 by an integral multiple of 6 dB, i.e., 6 dB times the integer “k”. Since the k-bit downward-shifted L-bit sequence is 2^k times smaller than an L-bit sequence which would be selected from the default bit positions when the calculated average power were equal to the intermediate reference value, the complementary increment of the amplifier gain by a factor 2^k controls the CNR value of the current lower-than-reference signal at the same value which would be obtained when the average power of the input signals is equal to the reference value.

[0024] If the control circuit determines that the calculated average power is higher than the intermediate reference value, it produces a down-shift control signal that causes the bit shifter 5 to select from an interpolated M-bit sequence an L-bit sequence {(i−j+1) through (i−j+L)} that is shifted upwards by “j” bits with respect to the default bit positions as illustrated in FIG. 3, where the integer “j” represents the difference between the calculated average power and the intermediate reference value. Control circuit 4 proceeds to control the amplifier 7 by decrementing its gain by an integral multiple of 6 dB, i.e., 6 dB times the integer “j”. Since the j-bit upward-shifted L-bit sequence is 2^j times greater then the L-bit sequence which would be selected from the default bit positions, the complementary decrement of the amplifier gain by a factor 2^j controls the CNR value of the current higher-than-reference signal at the same value which would be obtained when the average power of the input signals is equal to the reference value.

[0025] If the calculated average power of the multiplexed data is equal to the intermediate reference value corresponding to the default bit-positions and designated as S_R/n, where n is the background noise.

[0026] When the calculated average power of multiplexed amplitude data is ½ of the reference level and designated as S_L, the bit shifter 5 selects an L-bit sequence from positions one bit downward-shifted with respect to the default positions. This one-bit downward shift causes the input signal of the D/A converter 6 to be scaled up by a factor 2. Complementary to the operation of bit shifter 5, the output of the D/A converter 6 is then scaled down by a factor 2 by decrementing the gain of amplifier 7 by 6 dB. Thus, the following relation holds:
{(S_L)×(2)+n}×(½)=S_L+n/2
Thus, the CNR of the signal S_L is equal to 2×(S_L)/n, and hence to S_R/n.

[0027] When the calculated average power of multiplexed amplitude data is two times as higher than the reference level and designated as S_H, the input of the D/A converter 6 is scaled down by a factor 2 by shifting the L-bit sequence by one bit upward with respect to the default positions. The output of the D/A converter 6 is scaled up by a factor 2 by incrementing the gain of amplifier 7 by 6 dB. Thus, the following relation holds:
{(S_H)×(½)+n}×(2)=S_H+2n
Thus, the CNR of the signal S_H is equal to (S_H)/2n, and hence to S_R/n.

[0028] It is seen that transmissions from a CDMA cell-site base station can be maintained at a substantially constant CNR value regardless of its transmission power levels. Another feature of the present invention is that since the average power of the multiplexed channel signals is compared with a reference power level in the control circuit 4, abnormal state of the signals can be detected. In such instances, the control circuit 4 controls the transmission power so that no interference can occur with other communication systems.

[0029] FIG. 4 shows a modified embodiment of the present invention in which a down-converter 11 and a detector 12 are connected to the output of power amplifier 9 for detecting the power level of the power amplifier 9. The output of detector 12 is used by the control circuit 8 to detect the difference between the calculated average power of the multiplexed digital amplitude data and the actual power level of the signal transmitted from the antenna 10.

[0030] Control circuit controls the bit shifter 5 according to the difference between the calculated average power and the reference power level in a manner as described above. Control circuit controls the gain of the amplifier 7 according to the difference between the calculated average power and the reference power level and the difference between the calculated average power and the actual power level of transmission. Thus, the gain of amplifier 7 is varied by an integral multiple of 6 dB according to the difference between the calculated average power and the reference power level in a manner as described above, plus a scale factor “α” which varies with the difference between the calculated average power and the actual transmission power level of the power amplifier 9. This feedback arrangement serves to prevent the power amplifier 9 from varying its transmission power which would otherwise by caused by environmental conditions.

[0031] The average power of the multiplexed digital amplitude data can also be obtained by a channel management unit as shown in FIG. 5. Channel management unit 13 provides not only channel assignment and power control functions, but provides detection of the total average power of individual channel signals. Channel management unit 13 also determines spreading codes to be used by spreading circuitry 14 for scrambling the signals of assigned channels. The scrambled signals are supplied to the multiplexer 1 as the transmit spread spectrum signals mentioned above. The total average power of the transmit channel 11 signals is supplied from the channel management unit 13 to the control circuit 4, instead of the signal calculated by the average power calculator 3 of the previous embodiments.