Method and apparatus for determining signal strength in a variable data rate system

Method and device for determining signal strength in a variable data rate system

Background of the invention

The present invention relates to a variable data rate communication system, in particular to a method for detecting signal strength using common data frame information.

In a spread spectrum phone or personal communication system (PCS), a large number of subscriber units communicate with each other through "cell sites" or "base stations." The communication path used by a base station to transmit data frames to a user unit is called "forward link". Conversely, the communication path used by the subscriber units to send data frames back to a base station is called a "reverse link".

The system capacity in a spread spectrum communication system can be limited by the amount of interference present in the system transmission frequency. When transmitting at the same frequency, a signal desired by a user can be transmitted by other users, causing interference. When the signal is continuously transmitted, the interference reaches the highest. The resulting interference limits the overall system capabilities.

System interference can be reduced by reducing the probability of subscriber units transmitting data in parallel on a common frequency. This reduction can be achieved by transferring large amounts of data and randomly distributing data bursts in a data frame at a given time. A spread spectrum digital communication system like CDMA can use a variable data rate vocoder that can generate less data than full rate. Data that is less than the full rate can then be distributed in a pseudo-random manner in a data frame using a device like a data set randomizer. For further information on this data set randomizer, please refer to the US Patent Application No. 08/194,823 entitled "DATA BURST RANDOMIZER" approved on February 14, 1994. The pseudo-random distribution in the data frame reduces the probability of parallel data transmission through multiple users with simultaneous transmission.

In the design of a spread spectrum communication system, the encoding of a subscriber unit or other transmitter depends on the data frame of a quarter rate of the users voice activity, such as "full rate" and half of the full rate. One-half, or one-eighth of continuous data at a low rate. Each rate less than the full rate transmits a relatively small amount of time per data frame than the corresponding high rate. One-half data rate transmits 1/2 the time, one quarter data rate transmits 1/4 the time, and one eighth data rate transmits only 1/8 of the time.

In an exemplary variable data rate spread spectrum system, data is represented by a set of data symbols composed of power control groups. The position of the power control group in each data frame is randomly determined. The subscriber unit then transmits the data frame to a base station or other receiver. For further details on the example method of a pseudo-random data format, please refer to US Patent No. 08/194,823 and July 1993 TIA/EIA/IS-95, TIA/ETA Intermediate Standard "Used for Dual Mode Broadband Spread Spectrum" "Mobile Station-Base Station Compatibility Standard for Dual-Mode Wideband Spread Spectrum Cellular System" (Mobile Station-Base Station Compatibility Standard for Dual-Mode Wideband Spread Spectrum Cellular System), Association of Communications Industries, Washington, DC.

A base station that receives a data frame whose data rate is not known beforehand, has decoded the data frame in one of the subscriber units. In a template spread spectrum system, a base station receives a data frame from a subscriber unit every 20 ms (milliseconds). With variable rate decoding, the current data frame received by a base station can be decoded at a different data rate instead of the previously received data frame. In order to properly process an input signal, the base station must determine the total amount of energy contained in the signal. Given the transmission data rate, a base station receiving variable rate data readablely calculates the position of a power control group in a data frame, and from this information makes an accurate signal strength measurement. However, the base station does not know that the data rate has been selected by the subscriber unit at the time when the signal strength must be measured and cannot be calculated in this way. Therefore, any method for determining the strength of a relevant signal must be done in a rate-independent manner. The present invention provides a method for determining signal strength in a rate-independent method.

One of the objectives of the present invention is to use common data frame information in a variable data rate communication system to create a data rate independent determination of signal strength.

Another object of the present invention is to use the signal strength measurement of a variable rate communication system to indicate that a signal is sufficient for voice and data traffic communication. The present invention measures the signal strength and provides a lock indication when the signal can be used for communication.

Another object of the present invention is to use signal strength measurement in a rake receiver design to determine whether the signal combination is appropriate. The present invention provides a combination indication when the signal strength is sufficient for multi-directional (diversity) signal combination.

The present invention determines the signal strength in a rate-independent manner in a variable rate communication system. In the preferred embodiment of the present invention, the data grouped in the data composition symbol is called a power control group. The power control group is transmitted at different locations in a data frame using the pseudo random distribution method.

By using a power control group group designed as an active power control group, the accuracy of signal strength measurement is greatly improved. The active power control group defines a time period during which a power control group has been transmitted. A receiver that uses the signal energy for the power control group can produce an accurate intensity measurement.

In a preferred embodiment using CDMA, the active power control group includes all power control groups transmitted at the lowest data rate. The time period including the power control group shows the lowest data rate and is also included in all the power control groups with the highest data rate. The lowest data rate in the preferred embodiment is one-eighth of the full transfer rate.

The system operation starts with a startup reset, or when a demodulator element is assigned to a new input signal. After resetting, the signal energy level is collected in the time period according to the active power control group. The higher the energy level of a signal, the stronger the signal. The energy levels are stored in one or more memory elements and are summed over time to obtain an average signal strength measurement. Generally, the memory storage element containing the oldest measurement is cleared through an aging process and reused as a new energy measurement.

The energy measurements are summed and compared with the threshold levels of some configurations to produce hysteresis. A first predetermined threshold level represents the energy level that a signal should have before further signal processing occurs. Signal processing occurs when the total energy level meets or exceeds the first predetermined threshold. If the energy threshold subsequently falls below the first threshold level but still remains above a second threshold level, the processing continues. Only when the signal energy falls below the second predetermined threshold level, the signal strength is considered insufficient for a special signal processing method, and further signal processing is interrupted.

The threshold level is determined based on the signal processing technology used. If the signal processing requirements require a lock indicator, a pair of lock thresholds can be set. The lock indicator can be used to indicate whether the signal received by the receiver is strong enough for demodulation to generate reliable data. Threshold pairs can also be used to enable multi-directional combination technology. In particular, the receivers designed with a rake receiver can demodulate the multipath version of the same signal in parallel. Combining some versions of the same signal is a method of reconstructing the original signal transmitted by the subscriber unit. The correct threshold level for any signal processing technique is set to achieve a desired result based on the operating parameters of the system.

The present invention uses the above-mentioned method in a variable rate communication system to determine whether the strength of a given signal is acceptable in communication. When the signal energy detected in the lowest data rate meets the determined predetermined threshold, the signal lock and a data combination indicator can be set.

The features, objectives and advantages of the present invention can be more prominent from the following detailed description and accompanying drawings, in which: Figure 1 is a block diagram showing an example subscriber unit transmission part of a variable data rate spread spectrum communication transceiver; Figure 2 is a block diagram showing an example base station receiver including multiple independent demodulation components; Figure 3 is a detailed block diagram showing an example lock detection in an example base station receiver demodulation component Figure 4 is an illustration of an exemplary lock detection device operating according to the principles of the present invention; and Figure 5 shows a given data frame when transmitted at different data rates and received by the receiver of Figure 2 How to be formatted.

In a communication system, accurate signal strength measurement is always required to maintain a reliable transmission chain. When a strong and relatively weak signal encounters a similar path along a given transmission path, the stronger signal usually provides a higher signal-to-noise ratio (SNR) from a better communication quality. ). A communication system capable of measuring the slight energy level change of a signal can accurately distinguish between strong and weak signals.

In a variable data rate communication system, a signal transmitted by a subscriber unit may include spuriously transmitted data. Since data changes are represented by signal strength, this distribution method makes accurate signal strength measurement difficult. In fact, a system that can lock to a proper signal will have proper signal strength measurements. The present invention provides a lock detection device and method for a variable data rate system, and the system can make the signal strength measurement independent of the data transfer rate.

The active power control group includes signal information independent of the data transfer rate, and it is a subset of the power control group in a data frame. In a preferred embodiment, a data frame has 16 power control groups, at least two of which are classified as power control groups. Figure 5 shows an example data frame formatted to transmit at every possible data rate. As shown in Figure 5, the power control group for each rate is black. A pseudo-random processor (as described later) controls the "placement" of the power control group of each transmission data frame. Observe that the processor indicates at least two power control groups, and the active power control group is transmitted in each data frame. These active power control groups are displayed at the same position in a data frame regardless of the data rate of the data frame. In a preferred embodiment, the active power control group is transmitted at a rate of 1/8 of the lower data rate. In FIG. 5, there are two power control groups; control groups 2 and 9 are colored with two black segments of the same 1/8 rate data frame 506. The position of the control group in the subsequent data frame changes. The position of the active power control group can be determined in the receiving unit in the same way as the position of the 8th rate data frame determined by the transmitting unit. The position of the active power control group can be based on one or more parameters, such as the identification of the transmitting or receiving unit, a given number of calls, the time of the transmitting day, or the PN code used by the transmitting or receiving unit.

Accumulating the energy level indicated by each active power group provides a reference for a reliable signal strength measurement. This approach is progressive to the aforementioned lock detection method because it restricts the signal measurement of the active power control group that includes real data rather than all of the transmitted power control group in a given data frame. It is possible that some groups will make the measurement inaccurate due to the inclusion of noise.

Another important point is to use multiple threshold levels to measure accurate signal energy levels. Once the signal is received, the individual accumulated energy is compared to multiple threshold levels instead of a signal energy threshold level. Multiple threshold levels are used to improve signal detection and reduce false signal detection errors caused by changes in the energy level of the area where the signal is transmitted. An exemplary embodiment uses two threshold levels when it makes a signal detection, but a more complex technique using multiple threshold levels can be used in more demanding communication environments. For example, a higher threshold level can be used to indicate when the energy level is sufficient for communication. A lower threshold level can be used to indicate when the energy level is insufficient for communication. When a signal exceeds the upper threshold, it can still be considered as limited communication, even if the subsequent minor changes make it fall below the same upper threshold. Once the signal exceeds the upper threshold, it is continuously considered as valid communication, only if it remains above the lower threshold.

In order to better understand the present invention, a simple description of the data decoded and transmitted by the subscriber unit is provided. Fig. 1 shows an exemplary embodiment of the transmitting part 100 in a variable data transceiver. In an exemplary communication system, such as a CDMA cellular or personal communication network (PCN) variable rate communication system, "forward link" describes a transmission path from a base station to a subscriber unit. Conversely, "reverse link" describes a transmission path from a subscriber unit to a base station. Usually, signal transmission is from a subscriber unit through one of an access channel or a traffic channel. The access channel is used for short signal messages such as call source, response call, and registration. The traffic channel is used for communication (1) main traffic, usually user voice data, (2) second traffic, usually user non-voice data, (3) signal traffic, such as command and control signals, (4) main The combination of the flow rate and the second flow rate, or (5) the combination of the main flow rate and the signal flow rate.

When functionalized in the main traffic presentation mode, the transmission part 100 transmits audio signals, such as voice and/or background noise, through the transmission medium as digital signals. In order to facilitate digital communication using audio signals, these signals are sampled and digitized using conventional techniques. For example, in FIG. 1, sound is converted into an analog signal by a microphone 102, and then converted into a digital signal by a codec 104. The codec 104 usually uses a standard 8-bit/μlaw format technology to perform an analog-to-digital conversion process. In addition, the analog signal can be directly converted into a digital format, in a single pulse code modulation (PCM) format. In an exemplary embodiment, the codec 104 uses 8 kHz sampling and provides an output of an 8-bit sample at the sampling rate to confirm a data rate of 64 kbps.

The 8-bit samples are output from the codec 104 to the variable-rate speech encoder 106, where a μlaw/uniform code conversion process is performed. In the variable rate audio control encoder 106, input data samples form a data frame with a predetermined sample. In the variable rate audio encoder 106 of the preferred implementation, each data frame includes 160 sequential samples or approximately 20 ms of speech at a sampling rate of 8 Hz. Other sampling rates and data frame sizes can be used as specific communication system designs that those skilled in the art will understand. The data frame of each sampled speech is a variable rate encoded by the variable rate audio control encoder 106. The voice-controlled encoder data packet is then output to a microprocessor 108 and associated circuits for additional formatting.

The microprocessor 108 receives data packets along a voice coding rate indicator every 20 ms. If it exists, the microprocessor 108 also receives a second flow data input. The microprocessor 108 also internally generates communication signal data (ie, commands). The microprocessor 108 usually includes a program command in a program command memory, a data memory, an appropriate interface and related circuits known in the art. The data is output by the microprocessor 108 to the CRC and tail bit generator 112 for each data frame. The CRC and tail bit generator 112 calculates a check bit of a set of data at a specific data rate, and generates a set of tail bits for each data frame.

In FIG. 1, a data frame with check bits and tail bits is output to a convolutional encoder 114. In an exemplary embodiment, the convolutional encoder 114 preferably uses a 1/3 rate, suppression length k=9 convolution code to encode the input data. For example, the rotary encoder 114 is marked with g₀=557 (octal), g₁=711 (octal) and g₂= 711 (octal) generator function construction. As known in the art, the convolutional encoder includes a series of time-shifted delay data sequences with modulo-2 other than the selected branch. The data sequence delay length is equal to k-1, where k is the code suppression length. Because the preferred embodiment uses a rate 1/3 code, three code symbols: code symbols (c₀), (c₁) And (c₂) Is generated for each data bit output to the encoder. Code symbol (c₀), (c₁) And (c₂) Respectively by the generator function g₀, G₁And g₂produce. The code symbols are output by the convolution encoder 114 and an interleaver 116 (interleaver). These code symbols are represented by code symbols (c₀) First, the code symbol (c₁) Second, and the code symbol (c₂) The final sequence is provided to the inserter 116. The tail bit can be used at the end of each data frame to reset the cyclotron encoder 114 to an all-zero state as a preparation for the next data frame.

The symbol output from the convolutional encoder 114 is provided to the square inserter 116 for repeating and inserting code symbols under the control of the microprocessor 108. Generally, the code symbols stored in random access memory (RAM) are arranged in a way that the code symbols are repeated according to the data transfer rate. When the transmission data rate is full rate, the inserter operates at a 100% duty cycle, and the code symbols from the square inserter 116 are not repeated. At half rate, the interpolator operates with a 50% duty cycle, and each code symbol is repeated once (that is, each symbol occurs twice). At 1/4 rate, the interpolator operates with a 25% duty cycle, and each code symbol is repeated three times (ie, each symbol is generated four times). At 1/8 data rate, the interpolator operates with a 125% duty cycle, and each code symbol is repeated seven times (ie, each symbol is generated eight times). For all data rates, code repetition results in a constant number symbol of 28,800 code symbols per second, for the data output from the cube inserter 116. Finally, as described below, only one occurrence of each code symbol is truly transmitted in the reverse link. In this example, the inserter 116 is a block inserter, which is constructed by a method known in the art, and outputs a total of one code symbol during a period of 20 ms.

Referring again to Figure 1, the inserted code symbols are output from the block inserter 116 to the 64-ary orthogonal modulator 118 in the data frame, corresponding to a symbol rate of 28.8 thousand symbols per second (28.8 ksps). ). Generally, the modulation of the spread spectrum communication channel through the reverse link uses M (M-ary) vertical signals. For example, when 64 (64-ary) vertical signals are used, every six code symbols provided by the square inserter 116 results in one of 64 possible modulation symbols being selected and output by the modulator 118. Usually, each 64 (64-ary) modulation symbol corresponding to the Walsh function contains 64 code slices. For a more detailed description of the use of the Walsh function and M (M-ary) vertical signals, please refer to US Patent No. 5,103,459, published on April 7, 1992, entitled "System for Generating Signal Waveforms in a CDMA Cellular Telephone SystemAnd method" (SYSTEM AND HETHOD FOR GENERATING SIGNAL WAVEFORMS IN A CDMA CELLULAR TELEPHONE SYSTEM), which is assigned to the assignee of the present invention and is incorporated by reference in this quotation. Although 64 (64-ary) modulation techniques are used to show the higher or lower M (M-ary) modulation techniques that may be used.

Each symbol is output from the modulator 118 to an input, an exclusive-OR gate, of a modulo-2 adder in the pseudo random (PN) modulator 124. These symbols are output by the modulator 118 at a rate of 4.8 ksps, which corresponds to a Walsh code slice rate of 307.2 thousand code slices per second (kcps). The other input to the mutex or gate is provided by the long code generator 120 which generates a masked pseudo-noise (PN) code. The long code sequence provided by the long code generator 120 is a code slicing rate that quadruples the Walsh code slicing rate of the modulator 118, that is, a PN code slicing rate of 1.2288 million code slices per second (Mcps). The mutual exclusion or gate in the PN modulator 124 combines the two input signals to provide a data intermediate output with a 1.2288 Mcps code slice rate.

This long code sequence is a length of 2⁴²The time shift of -1 code slice sequence, and is produced by the linear generator using the following polynomials well known in the art: p(x)=x⁴²+x³⁵+x³³+x³¹+x²⁷+x²⁶+x²⁵+x²²+x²¹+x¹⁹+x¹⁸+x¹⁷+x¹⁶+x¹⁰+x⁷+x⁶+x⁵+x³+x²+x¹+1。(4)

The intermediate output in the PN modulator 124 is separately provided as an output of a pair of modulo-2 adders (for example, mutual exclusion or gate). The other output of each pair of gates is the second and third PN sequences used to cover the modulated signal. The second and third PN sequences (or "short codes", short codes) are provided by individual I and Q channel PN generators included in the short code generator 121. This data is the OQPSK carried out before the actual transmission using the second and third PN sequences. The OQPSK deployed on the reverse link traffic channel uses the same I and Q PN codes as the forward channel I and Q pilot PN codes. The I and Q PN codes generated by the short code generator 121 are of length 2¹⁵And preferably, the zero-time separation code is related to the forward link channel recognized by the subscriber unit. For further understanding, the forward link channel pilot signal is generated by each base station. The pilot channel signal of each base station is carried out by the aforementioned I and Q PN codes. The I and Q PN codes of all base stations are the same, but between base stations, these codes are shifted and separated by a code sequence to provide the difference between base station transmissions. The generation functions for I and Q short PN codes are as follows: P_I(x)=x¹⁵+x¹³+x⁹+x⁸+x⁷+x⁵+1 (5) and P_Q(x)=x¹⁵+x¹²+x¹¹+x¹⁰+x⁶+x⁵+x⁴+x³+1 (6) The short code generator 121 can be constructed with a technique well known in the art to provide an output sequence according to equations (5) and (6). An example of a code generator is disclosed in US Patent No. 5,228,054 issued on July 3, 1993, entitled "POWER-OF-TWO LENGTH PSEUDO" (POWER-OF-TWO LENGTH PSEUDO). -NOISE SEQUENCE GENERATOR WITH FAST OFFSET ADJUSTMENT).

The I and Q waveforms are respectively output by a mod-2 adder (for example, mutual exclusion or gate) pair, and are provided as inputs to a pair of finite impulse response (FIR) filters 122, respectively. The FIR filter 122 is a digital filter whose bandwidth is limited to the resulting I and Q waveforms. The FIR digital filter 122 shapes the I and Q waveforms so that the resulting spectrum is contained in a given spectrum mask. The FIR filter 122 can be constructed according to conventional digital filter technology, and preferably provides a desired frequency response.

The binary '0' and '1' inputs to the FIR filter 122 generated by the PN expansion function can be mapped to +1 and -1, respectively. The sampling frequency of the digital filter is 4.9152 MHz=4×1.2288 MHz. The binary '0' and '1' input sequences added in synchronization with one of the I and Q digital waveforms are provided to the FIR filter 122. The special sequence referred to by the same mask sequence is an output generated by a data burst randomizer. The mask sequence is multiplied by the I and Q binary waveforms to generate a ternary (-1, 0, and +1) FIR filter 122 input.

As mentioned earlier, the data rate used for reverse link traffic channel transmission is variable (9.6, 4.8, 2.4, or 1.2 kbps) and is frame-by-data frame-by-frame. frame) based on changes. Since these data frames are a fixed 20 ms length for the reverse link traffic channel, the number of information bits of each data frame at 9.6, 4.8, 2.4, or 1.2 kbps is 192, respectively. 96, 48 or 24. As mentioned earlier, the information is encoded using a 1/3 rate cyclotron encoder, and the code symbols are respectively corresponding to a data rate of 9.6, 4.8, 2.4, or 1.2 kbps with a coefficient of 1, 2, 4, or 8 Repeat. Assume that the resulting repetition code symbol is therefore fixed at 28,800 symbols per second. The insertion of this 28.8 ksps data stream is as described above.

In the previous transmission, the reverse link channel signal was gated, so the selected symbol was deleted and the remaining symbols were transmitted. As the data rate changes, the transmission gate duty cycle also changes. When the transmission data rate is full rate (9.6 kbps), the transmission gate allows all symbols to be transmitted. However, when the transmission data rate is half the rate (4.8 kbps), the transmission gate allows half of the symbols to be transmitted. At 1/4 rate (2.4 kbps), the transmission gate allows 1/4 symbols to be transmitted. Similarly, at 1/8 rate (1.2 kbps), the gate only allows 1/8 of the symbols to be transmitted through the same gate processing.

The gate processing operation consists of dividing each data frame into a predetermined number of equal length periods, called power control groups. Each power control group contains 12-bit or 36-bit data symbols in the form of 6 Walsh symbols, and each Walsh symbol is composed of 64 Walsh code slices. The particular power control group is gated on (that is, to be transmitted), while the other groups are gated off (that is, not to be transmitted). The decision to block or unblock the group is controlled by the data set randomizer logic 110. The gated power control group is pseudo-randomized at these positions in a data frame, so that the real traffic loaded on the reverse link channel is equally divided among different user units. The randomized power control group is different from Spread the energy of the system in a pseudo-random time range. In addition, the gated power control group is therefore only transmitted once for each code symbol input to the repetitive processing. During the unlocking period, the subscriber unit does not transmit energy, and the interference of other subscriber units operating on the same reverse link channel frequency is reduced, and the subscriber unit energy is preserved.

Transmission gated processing can be disabled by specific transmissions, such as when a subscriber unit attempts to access the system or transmit other non-traffic messages. In such an example, the subscriber unit can transmit a message at 4.8 ksps with a code symbol repeated once (each symbol occurs twice) in the transmission.

When the data set randomizer function is completed, the data set randomizer logic 110 generates a mask stream of '0' and '1', which pseudo-randomly masks redundant data generated by code repetition. The mask flow pattern is determined by the data rate and a 14-bit block, and the 14-bit block is taken from the long code sequence generated by the long code generator 120. The masking bits are synchronized with the data stream and the data is selectively masked by these bits through the operation of the FIR filter 122. In the data set randomizer logic 110, the 1.2288 MHz long code sequence output by the long code generator 120 is input to a 14-bit shift register, which is shifted at a 1.2288 MHz rate. The contents of this shift register are loaded into a 14-bit latch just before a power control group (1.25 ms) before the data frame range of each reverse link flow channel. The logic 110 uses the data accompanying the input rate from the microprocessor 108 to determine the specific power control group of the FIR filter 122 that the data can be allowed to pass through according to a predetermined rendering method. The logic 110 therefore outputs a '1' or a '0' for all power control groups for each power control group according to the data to filter out ('0') or pass ('1'). For further description of the function of the data set randomizer, please refer to the aforementioned TIA/EIA/IS-95 standard and the jointly filed U.S. Patent No. 08/194,823, titled "DATA BURST RAMDOMIZER), which is assigned to the assignee of the present invention, and is hereby for reference.

In FIG. 1, the I channel data output from the FIR filter 122 is directly provided to a digital-to-analog (D/A) converter and an anti-alias filter circuit 126. However, the Q channel data is output by the FIR filter 122 to a delay element (not shown) that adds 1/2 PN code slice time delay (406.9 nsec) to the Q channel data. The Q channel data is then output from the delay element to the digital-to-analog (D/A) converter and the anti-alias filter circuit 126. The circuit 126 converts the digital data into an analog form, and filters the analog signal. The signal output from the circuit 126 is provided to an Offset Quadrature Phase Shift Key (OQPSK, Offset Quadrature Phase Shift Key) modulator 128 that is adjustable and output to the RF transmitter 132. The RF transmitter 132 amplifies the filter and up-converts the transmitted signal. The RF transmitter 132 can also receive the signal 130 from the logic 110, so that the transmitter is on when transmitting the data frame part and off at other times. The RF signal is transmitted from the transmitter 132 to the antenna 134 to communicate with the base station.

The data transmitted on the reverse link channel is encoded, inserted and modulated by M vertical modulations and the direct sequence PN before transmission. Table I further defines the relationship between data values, symbols, and transfer rates on the reverse link traffic channel. Since the data frame has a fixed length of 20 ms for both the access channel and the reverse link traffic channel, the number of information per data frame at the data rate of 9.6, 4.8, 2.4 or 1.2 kbps during transmission is 192 respectively , 96, 48 or 24.

<tables><img file="TW413985B_D0001.tif" /></tables>

In order to further understand the present invention, the aforementioned signal reception and processing will be discussed. Figure 2 shows how a signal transmitted by a subscriber unit is received and processed by a base station. FIG. 3 shows how the received signal is processed by the demodulation unit 204 and finally reaches the energy accumulator and lock detector 344.

A typical base station receiver has at least one demodulation element, which utilizes at least one lock detector device in the demodulation unit to assist signal detection. Similarly, a base station containing multiple independent demodulation elements can utilize multiple such detectors. Figure 2 shows an embodiment of a base station with multiple demodulation elements. Figure 2 shows a three-sectored base station, where each antenna 222A-222C is a sector antenna. Each antenna 222A'-222C' corresponds to one of the antennas 222A-222C, and is a branch antenna of the segment corresponding to the same number of antennas. Each antenna 222A-222C includes the same range area as the corresponding antenna 222A'-222C'. In a general base station, the antennas 222A-222C have an overlapping range area, which subdivides the base station into three segments, and each antenna in the three segments contains more than 1/3 of the total aggregate range of the base station. Therefore, a signal from a signal subscriber unit can behave more than one antenna at a time. The number of segments and the number of antennas given to each segment are variable. This type of change will not affect the general principles of the present invention.

The antennas 222A, 222B, 222C, 222A', 222B', and 222C' provide the received signals to the RF processing and provide them to the digital conversion circuits 224A, 224B, 224C, 224A', 224B', and 224C', respectively. RF processing and digital conversion 224A, 224B, 224C, 224A', 224B' and 224C' process RF signals and convert the signals into digital data. The RF processing and digital conversion circuits 224A, 224B, 224C, 224A', 224B', and 224C' filter the digital data and provide the resultant digital data to the interface port 226. The interface port 226 can connect any one of the six input signal paths to any one of the searcher element or the demodulation element through the internal connection 212 under the control of the controller 200.

The searcher and demodulator components are also controlled by the controller 200 through the internal connection 212. The searcher components 202A-202N, when controlled by the system controller 200, continuously scan a time range window to search for the data signal of a particular user unit. The searcher components 202A-202N also scan a set of time intervals near the normal arrival of the signal to search for multiple signals generated.

The searcher components 202A-202N transfer the received data to the controller 200 for storage in the memory 218. The searcher components 202A-202N can transmit data through a standard bus or transmit data to the memory 218 (not shown) through direct memory access. The controller 200 uses the data stored in the memory 218 to give the demodulation elements 204A-204N to one of a plurality of information signals from a signal subscriber unit.

In order to provide a fast search process, more than one searcher element is used to manage a full search. Each search element 202A-202N is given by the controller 200 to search for a set of time gaps. Each of the searcher components 202A-202N provides the search results performed back to the controller 200. The controller 200 uses a given method to tabulate these results.

The demodulation elements 204A-204N demodulate the received signals to generate data symbols combined with a symbol combiner 208. The output of the symbol combiner 208 (not shown) is a collection of software decision data suitable for Viterbi decoding. The symbol combiner 208 can only combine the symbols of one segment or can combine the symbols of multiple segments selected by the interface port 226. When the symbol combiner 208 combines signals from a subscriber unit communicating through multiple segments, this state is considered a soft handoff. The base station can transmit the output of the symbol combiner 208 to a cellular system controller, where the symbol from a common subscriber unit is also combined with signals from other base stations to generate a signal output. This process is also regarded as soft handover. The demodulation elements 204A-204N also provide multiple output control signals to the controller 200 through the internal connection 212 for a given process such as a lock or unlock instruction.

Each demodulation element 204A-204N is very similar in structure to the others. FIG. 3 shows the demodulation element 204 in FIG. 2 in more detail. In Figure 3, the digitized input signal is assumed to be an offset quadrature phase shift key (OQPSK) signal with in-phase (I) (in-phase) and quadrature-phase (Q) (quadrature-phase) signal samples. The I and Q signal samples each have a multi-bit value and are input to a decimator and a despreader 330. Usually I and Q signals are sampled in full at the input received at a data rate higher than the code slice rate. In the decimator and de-expander 330, the data is decimated from the data rate of the entire sampling to the PN code slice rate. The data is then decoded and developed by the tenter and decoder 330, using the same three PN sequences in the subscriber unit to modulate the signal.

The tenter and solution developing device 330 outputs solution developing I and Q signal elements to the accumulator 338. The accumulator 338 accumulates the I and Q signal elements in a Walsh code slice period to generate accumulated I and Q code slice data. The accumulated I and Q code slice data are then processed by the Fast Hardamard Transformer (FHT, Fast Hardamard Transformer) and the selector 340. The FHT part of the FHT and the selector 340 are related to the accumulated I and Q code slice data with all possible Walsh sequences. The results of each I and Q correlation are then used to estimate the size of the relevant Walsh symbol. From the magnitude of each I and Q correlation result estimates can be compared with each other. The Walsh symbols corresponding to the I and Q correlation results with the largest energy are selected by the selection part of the FHT and the selector 340 to be used as the demodulated Walsh symbols. The demodulated Walsh symbol is output with the estimated size of the corresponding Walsh symbol.

Since different arrival times of the signal path are set to different demodulation elements, the demodulation element 204 performs a symbol de-skewing. Time dewarping 342 delays the output so that each demodulation element provides synchronized symbol data related to the other demodulation elements.

The energy accumulator and lock detector 344 sum a sequence of continuous Walsh symbol energy. The result sum is output when the signal strength 364 is sent to the controller 200 (as shown in FIG. 2) for the setting of the demodulation element 204. The result sum is also compared with the threshold to indicate a locked or unlocked state and a signal combined state.

The arrival time of the signal path set to the demodulation element 204 can be changed due to the change of the subscriber unit through time, or the environment of the subscriber unit can be changed. Therefore, the base station demodulation component 204 incorporates time tracking circuitry. The tenter and de-developer 330 outputs an earlier and later version of I and Q signal elements for time tracing processing. The accumulator 332 accumulates the earlier and later solutions to develop the I and Q signal elements through a Walsh code slicing cycle to generate the accumulated earlier and later I and Q code slice data. The earlier and later metric generator 334 adds and multiplies the earlier and later I and Q code slice data accumulated by the Walsh sequence of the corresponding demodulated Walsh symbol, and accumulates the result to generate the earlier and later I and Q Walsh symbols. The earlier Walsh symbol size is generated based on the earlier I and Q Walsh symbols, and the later Walsh symbol size is generated based on the later I and Q symbols. The size of the earlier symbol is subtracted from the size of the later symbol to generate an error measurement. This error measurement is output to the time tracking circuit 336. The time tracking circuit 336 uses the error measurement to determine whether the development operation in the tenter and the solver 330 is performed ahead, behind, or on time. The time tracking circuit 336 also keeps track of the absolute demodulation path time of the demodulation element to output to the controller 200 (as shown in FIG. 2).

The searcher components 202A-202N in FIG. 2 are similar to the demodulation components, except that the searcher components do not have time tracking and lock detection. Time tracking is not needed in the searcher components 202A-202N because the search process is completed quickly with respect to the channel parallel system, and the time drift is ignored during the time used to perform a single search. For similar reasons, a lock detector device of the present invention is not required in the searcher components 202A-202N.

The object of the present invention is to provide a reliable method for determining the strength of a signal, and the signal includes variable rate data transmitted by the subscriber unit and received by the base station for lock detection purposes. Depending on the data rate selected by the subscriber unit, the symbols included in the power control group are transmitted in a pseudo random position determined by a reverse link transmission data frame. In the present invention, the lock detector included in the demodulation element 204 in FIG. 2 and FIG. 3 indicates the signal strength of a received signal.

The magnitude of the signal strength generated by the energy accumulator and lock detector 344 is important for many reasons. Signal strength The main advantage of this method is that it has a fast and reliable signal strength prediction when demodulating. The present invention, when generated by the fast attenuation feature of the earth channel, generates an immediate response to a rapidly changing signal strength. It should be noted that a more accurate signal strength measurement can be generated after the data rate of the input signal is generated. However, the rate decision process usually includes a Viterbi decoding step that induces a delay. In the present invention, the signal strength indicator can be used to determine whether the input signal has sufficient strength to be worth combining the output from other demodulation elements with subsequent Viterbi decoding. The signal strength can be used to temporarily deactivate time tracking during periods of low signal strength. The demodulation element should be resettable to a new time interval which can be used as an indication.

The implementation of an example of the present invention is illustrated in FIG. 4. The demodulator energy 432 is used to determine whether the signal strength is sufficient for the receiver in a rate-independent manner. If the signal strength is equal to or exceeds the predetermined lock threshold level, the received signal is considered to be sufficient for communication and the lock indicator signal 422 can be set appropriately. The demodulation element is considered to be locked to the input signal.

Using a similar method as described above, if the signal strength is sufficient for different combining techniques, a combined indicator signal can be provided. Once the detected signal energy level reaches a predetermined combination threshold level, the received signal strength is deemed to be sufficient for combination and the combination indicator signal 424 is appropriately set.

First, the power control group clock 428 provides a signal indicating that a power control group can appear at a certain time in a fixed-length data frame. As mentioned above, an example data frame occupies a 20 ms time period and is further subdivided into 16 power control groups of 1.25 ms. The power group clock signal indicates each of the 16 power control groups in each data frame.

Because the signal gate processing takes place in the subscriber unit, when the data transmission is less than the full rate, some power control groups do not contain data. The active power control group is one of the smallest subsets of the power control groups included in the data of all data transfer rates. The active power control group clock signal provides a position indication and a gated-on power control group frequency independent of the data rate.

The base station receiver uses the same long PN code to decode the signal that is used in the subscriber unit to develop the signal. Long PN codes are also used in subscriber units to locate power control groups in data frames of different rates. Similarly, the long PN code is used to identify the active power control group in the base station. Until the complete synchronization is determined, the active power group can be used as an indication of the completion of the signal synchronization.

An exemplary embodiment uses the lowest rate in the communication system to determine which time interval constitutes the active power control group. The active power control group at the lowest data rate is a subset of the power control group included in all higher rate data. Therefore, the basic lock indication only works on the power control group to ensure correct signal strength measurement. An example 1/8 rate data frame 506 in FIG. 5 shows that two power control groups containing data are regarded as active power control groups 2 and 9. These two power control groups are regarded as active power control groups, because they used 1/4-rate data frame 504, 1/2-rate data frame 502, or full-rate data frame 500. These same power controls The group will still contain data. In this example, the signal strength of all data frames is sampled as a 1/8 rate data frame 506 to ensure correct strength measurement regardless of the correct data rate. The teaching of the present invention also intends to select other active power control groups based on other criteria of lower data rate in the system.

In FIG. 4, the Walsh symbol energy level value output from a demodulator element 204 (as shown in FIGS. 2 and 3) is provided as a demodulator energy signal 432 to an accumulator 401 input. The other input of the accumulator 401 is the symbol clock signal 430. In response to the symbol clock signal 430, the total of the accumulator 401 corresponds to the energy level values of the six input Walsh code slices of a power control group. The sum value is output to a filter included in the shift register 446 and the digital adder or totalizer 438. The accumulator 401 is reset or cleared due to the next set sum of the Walsh code slice energy level value corresponding to the next power control group.

The active power control group clock 426 and the power control group clock 428 are connected to the input of the "and" gate 400. When an active power control group is received, both active power control group clock 426 and power control group clock 428 will each provide a signal indication. The power control group clock 428 can be implemented with a two-phase clock to provide a first clock signal to the gate 400 and a delayed clock signal to the reset input of an accumulator 401.

When the two inputs to the gate 400 are instructed, the energy value of the accumulator 401 is output to the shift register 446. The example shift register 446 that connects and responds to the gate 400 includes a plurality of serially connected register stages 402-416 that store the sum energy levels of multiple active power control groups. The shift register 446 accepts a new energy measurement in a register stage 402, and shifts the pre-energy information from the register stage 402 to the subsequent register stage 404 in response to the gate 400. The subsequent register stages 404-414 also shift their current energy measurements to their respective subsequent register stages 406-416 in response to the gate 400. This method is a first-in-first-out (FIFO, first-in-first-out) method of collecting energy samples loaded into the shift register 446. In this example, the shift register 446 includes 8 registers. class. Many variations in the type of shift register and shift registers of more or less stages can be used. Furthermore, many methods for shifting the values between such register stages can be used depending on the specific application and are included in the teachings of the present invention.

The register preset signal 448 is connected to the register elements 402-416 in the shifter 446 to load a register with a set of preset values. As will be seen, the value loaded into this register can be used to initialize the enable or disable combination or lock indication.

The energy level of each register stage 402-416 in the shift register 446 is each connected to the input of the totalizer 438. The totalizer 438 obtains the individual energy level value provided by each register stage 402-416, and provides a total energy level value 440 of the sum of the energy levels shown in the energy value shift register 446 . Obviously, the shift register 446 and the totalizer 438 perform a FIR (Finite Impulse Response) type filter function. Furthermore, a selected group of energy specific gravity elements 402'-416' can be used to modify the individual energy level contribution generated by each register element to a total energy level value 440. The signal strength received by the demodulation element is related to the total energy indicated by the output of the summing unit 401. The higher the level of energy collected in the register stage 402-416, the more the demodulated signal can be used in communication.

The total energy level 440 is provided as an input to each of the one to one pair of comparators 418 and 420, respectively. The other inputs to the comparator 418 are the lock threshold signal 434 and the non-lock threshold signal 436. Depending on the value of the total energy level 440, the lock threshold signal 434, and the non-lock threshold signal 436, the lock indication signal 422 may be provided. The second comparator also operates in the same way. The additional input to the comparator 420 also combines the threshold signal 442 and the non-combined threshold signal 444. Depending on the value of the total energy level 440, the combined signal 442, and the non-combined signal 444, the combined indicator signal 424 may be provided.

The operation of the lock detection method starts from the reset of a system, usually when the system is started or when the corresponding demodulation element is set to a new signal. In one of these cases, each new signal occupies a single time gap. When a new signal is given, the existing data is removed from the register stages 402-416, and the shift register 446 is pre-loaded with preset initial data. The shift register stage is then ready to start the evaluation of the demodulated signal energy of the received signal. In an exemplary embodiment, the initial data setting is loaded so that the lock detector only provides a lock indication after the actual energy level is shifted to the shift register 446. The value used to initialize the filter is selected by the system designer based on the known system requirements. The system operator can determine that the detector should initialize the lock on the signal and therefore preload the shift register 446 with the same value of the initial indication signal when locked.

As mentioned above, the shift register 446 stores eight energy values, and shifts each active power group lock period using a FIFO data storage stage with a new energy value. In an exemplary embodiment, the energy totalizer 438 totals the energy values stored in the eight most recently applied power supply groups, and provides a total energy measurement. In an implementation, the energy totalizer 438 weights the contribution of each register stage from the lock detector shift register to make it equal, and provides the resulting accumulated energy signal 440 to the lock comparator 418 and the combined comparator 420.

The execution of the other totalizer 438 can change the energy contribution from the power group according to the relative length of the time shift register 446 that maintains the power of the power group. Changing the energy contribution as a function of time is called aging. Generally, a filter element receiving a new power group indicates the current signal strength state more accurately than any filter element. For example, the energy weighting element 402' combined with the register stage 402 can be set so that the register stage 402 establishes a larger energy sum ratio than the subsequent register stages 404-416. The energy weighting elements 404'-416' can be set so that the register stages 404-416 establish a smaller energy value according to the ratio to the register stage 402 to the totalizer 438. The different variations of the energy weighting elements 402'-416' can be used as a good modulation-filter. If the number of filter elements increases or if more signal control is usually required. Furthermore, the filter function can utilize a conventional IIR (infinite impulse response) filter.

In one of the above implementations, the hysteresis of a threshold level is compared with the total energy provided by the totalizer 438 to determine whether a lock indicator signal or a combination indicator should be generated. The hysteresis of an energy level is received by the demodulation element due to the transient change of energy to reduce the change of the lock indication. For example, when the accumulator energy signal 440 meets or exceeds the value of the threshold signal 434, a lock indication 422 is provided. Once the lock indication signal 422 is provided, a subsequent fall value from the value of the accumulated energy signal 440 must fall below the non-lock threshold level 436 to unlock the lock indication signal 422. When the accumulated energy signal 440 falls below the non-lock threshold signal 436, the signal strength is not considered sufficient for communication and the lock indication signal 422 will no longer be provided. The real lock indicator level can be calibrated according to the experimental values collected in a time interval.

A similar threshold quasi-hysteresis phenomenon can be compared with the accumulated energy signal 440 to determine when a diversity combine indication signal 424 should be generated. For example, when the accumulated energy signal 440 meets or exceeds the threshold signal 442, a combination indication 424 is provided. The combination indicator signal 424 is obvious. The current signal from a specific demodulation element can be combined with other multipath versions of the same signal provided by other demodulation elements to improve all signal-to-noise ratios (SNR, signal). -to-noise ratios). The accumulated energy signal 440 must subsequently fall below a non-combined threshold level 444 before the combined indication signal 424 is no longer provided. In an exemplary embodiment, the binding threshold signal 442 is greater than the non-binding threshold. Signal 444 When the accumulated energy signal 440 falls below the second binding threshold, the signal strength is no longer considered sufficient for the multi-path and multi-directional combination technology, and the combination indication signal 424 is no longer provided. The true binding indicator level can be calibrated based on experimental values collected in a time period.

The above-mentioned preferred embodiments are provided so that anyone skilled in the art can make or use the present invention. The different modifications of these embodiments are obvious to those who are familiar with the art, and the general rules defined here can be applied to other embodiments without using the inventive art. Therefore, the present invention is not limited to the embodiments shown here, but can be based on the broad scope of the rules and prominent features disclosed in the present invention.