Method and apparatus for variable rate data communication

IS 2 273 136 T3

DESCRIPTION

Procedure and apparatus for variable speed data communication.

Background of the invention

I. Field of the invention

The present invention relates generally to communication systems and, in particular, to power control in a code division multiple access communication system.

II. Description of Related Art

The Federal Communications Commission (CFC) governs the use of the radio frequency (RF) spectrum, when deciding which industry gets which frequencies. Since the RF spectrum is limited, only a small portion of the spectrum can be allocated to each industry. The allocated spectrum must therefore be used efficiently in order to allow as many frequency users as possible to have access to the spectrum.

Multiple access modulation techniques are some of the most efficient techniques for using the RF spectrum. Examples of such modulation techniques include time division multiple access (TDMA), frequency division multiple access (FDMA), and code division multiple access (CDMA).

CDMA modulation uses a spread spectrum technique for the transmission of information. A spread spectrum system uses a modulation technique that spreads the transmitted signal over a wide frequency band. Typically, this frequency band is significantly wider than the minimum bandwidth required to transmit the signal. The spread spectrum technique is performed by modulating each basic band data signal to be transmitted with a unique wide band spreading code. Using this technique, a signal having a bandwidth of only a few kilohertz can be spread over a bandwidth of more than one megahertz. Typical examples of spread spectrum techniques can be found in Spread Spectrum Communications, Volume I, MK Simon, Ch. 5, pp. 262-358.

A variety of frequency diversity is obtained by spreading the transmitted signal over a wide range of frequencies. Since only 200 to 300 kHz of a signal are typically affected by selective frequency fading, the remaining spectrum of the transmitted signal is unaffected. A receiver that receives the spread spectrum signal, therefore, will be less affected by the fading condition.

In a CDMA type radiotelephony system, multiple signals are transmitted simultaneously at the same frequency. A specific receiver then determines which signal is destined for that receiver by the unique spreading code in the signal. Signals at that frequency, without the specific extension code intended for that specific receiver, appear to be noise for that receiver, and are ignored.

Fig. 1 shows a typical prior art CDMA transmitter for use on the reverse channel of a radiotelephony system, the reverse channel being the link from the mobile to the base station. First, a basic band digital signal is generated by a vocoder (speech encoder / decoder). The vocoder 100 digitizes an analog voice or data signal using a coding method such as the Code Excited Linear Prediction (CELP) method, which is well known in the art.

The basic band digital signal enters a convolutional encoder (101) at a specific rate, such as 9600 bps (bits per second). Encoder 101 convolutionally encodes the input data bits into data symbols at a fixed encoding rate. For example, the encoder (101) could encode the data bits at a fixed encoding rate of one data bit per three data symbols, such that the encoder (101) outputs data symbols at a rate of 28, 8 kilosymbols / second, with an input speed of 9600 bps.

The data symbols from the encoder enter an interleaver (102). The interleaver (102) shuffles the symbols such that the lost symbols are not contiguous. Therefore, if more than one symbol is lost in the communication channel, the error-correcting code is able to recover the information. Data symbols enter interleaver (102) in a column-by-column matrix, and exit the matrix row by row. Collation occurs at the same rate of 28.8 kilosymbols / second that the data symbols were entered.

The interleaved data symbols enter a modulator (104). Modulator 104 derives a sequence of fixed length Walsh codes from the interleaved data symbols. In 64-element orthogonal code signaling, the interleaved data symbols are grouped into sets of six in order to select one of the 64 orthogonal codes to represent the set of six data symbols. These 64 orthogonal codes correspond to Walsh codes of a Hadamard matrix of 64 rows by 64 columns, in which a Walsh code is a single row or column of the matrix. The modulator outputs a sequence of Walsh codes, corresponding to input data symbols at a fixed symbol rate, towards an input of an XOR combiner (107).

IS 2 273 136 T3

A pseudo-random (SR) noise generator (103) uses a long sequence of SRs to generate a user-specific sequence of symbols. In a mobile radiotelephone having an electronic serial number (ESN), the ESN can undergo a logical exclusive-OR operation with the long SR sequence in order to generate the sequence, making the sequence specific to that radiotelephony user . The long SR generator (103) inputs and outputs data at the extension rate of the system. The output of the SR generator (103) is coupled to the XOR combiner (107).

The extended Walsh code symbols of the combiner 107 are quadraturely extended below. The symbols enter two XOR combiners (108 and 109) that generate a pair of short SR sequences. The first XOR combiner (108) performs a logical exclusive-OR operation between the extended Walsh code symbols and the sequence (105) in phase (I), while the second combiner (109) performs a logical exclusive-OR operation between the Walsh code extended symbols and the quadrature phase (Q) sequence (106).

The resulting extended I and Q channel code sequences are used to biphasically modulate a quadrature pair of sinusoids, regulating the power level of the pair of sinusoids. The sinusoidal output signals are then summed, band-pass filtered, translated to an RF frequency, amplified, filtered, and radiated by an antenna.

The typical prior art CDMA transmitter used on the forward channel of a radiotelephony system (the link from the base station to the mobile) is similar to the reverse channel. This transmitter is illustrated in Fig. 4. The difference between the transmitters of the forward and reverse channels is the addition of a Walsh code generator (401) and a power control bit multiplexer (420) between the SR generator combiner (107) and the combiners (108 and 109) of quadrature extension, for the forward channel transmitter.

The power control bit multiplexer (420) multiplexes one power control bit instead of another bit in the frame. The mobile knows the location of this bit and looks for this power control bit at that location. As an example, a "0" bit instructs the mobile to increase its average output power level by a predetermined amount, and a "1" bit instructs the mobile to decrease its average output level by a predetermined amount.

The code division channel selection generator (401) couples to a combiner (402) and provides a specific Walsh code to the combiner (402). Generator 401 provides one of 64 orthogonal codes corresponding to 64 Walsh codes, from a Hadamard matrix of 64 rows and 64 columns, in which a Walsh code is a single row or column of the matrix. Combiner 402 uses the specific Walsh code input by code division channel generator 401 in order to extend the input mixed data symbols to Walsh code extended data symbols. The Walsh code spread data symbols are output from the XOR combiner 402 and input to the quadrature extender combiners at a fixed chip rate of 1.2288 Mchp / second.

The mobile can assist the base station in power control on the forward channel by transmitting a power control message to the base station on the reverse link. The mobile collects statistics on its error distribution and informs the base station by means of the power control message. The base station can then adjust its power in correspondence with the specific user.

The problem with the type of power control described above is that, for forward link control, the power control message replaces the voice or data bits, thereby reducing the quality of the voice or data throughput. This fundamentally limits the speed at which mobile stations can send power control messages to the base station and, in turn, the speed at which the base station can adjust the output power to this specific mobile. A high update rate transmit power setting would allow the base station to fine tune the transmit power for each individual mobile station to a minimum level necessary to maintain a link of a specified quality. By minimizing each individual transmit power, the total interference generated is also minimized, thereby enhancing the capacity of the system. There is a resulting need to upgrade a transmitter's power output to a higher rate, without significantly degrading the quality of the data in transmission.

According to the present invention, there are provided a method for providing a variable data rate communication link, as set forth in claim 1, and an apparatus for providing a variable data rate communication link, as set forth in claim 4. Further embodiments of the invention are claimed in the dependent claims.

Additional aspects of the invention

According to a first aspect, the described method of the present invention allows a transmitter to update the power transmitted to each mobile station with which it is communicating, frame by frame. The procedure is carried out through a response mechanism from the mobile station to the base station. Through the response mechanism, the mobile station informs the base station whether it is receiving frames correctly or incorrectly, including such information in each data frame transmitted to the base station.

IS 2 273 136 T3

The method first determines whether the emitted power of the transmitter, with which communication is established, is to be increased or decreased. The procedure then informs that transmitter to change its power accordingly, including power control bits in each transmitted data frame.

According to a further aspect, the method of the present invention allows a communication link to have a higher data rate input signal, while maintaining a constant data rate output signal. The method first convolutionally encodes the input data signal to produce a plurality of convolutionally encoded signals. Each of the convolutionally coded signals is composed of a plurality of data symbols. Each data symbol is repeated a predetermined number of times, in order to produce a code repetition data sequence at a predetermined and fixed rate. The data stream is then punched in such a way that symbols are erased at predetermined locations in the data stream, thereby generating a data stream at a predetermined and fixed rate, which is lower than that of the original data stream. The signals encoded with the repeated data symbols are multiplexed to produce a data sequence.

Brief description of the drawings

Fig. 1 shows a typical prior art reverse link CDMA transmitter for use in a radiotelephony system.

Fig. 2 shows the direct communication link method of the present invention, as used in a CDMA radiotelephony system.

Fig. 3 shows the mobile radio method of the present invention, as used in a CDMA radiotelephony system.

Fig. 4 shows a typical prior art forward link CDMA transmitter for use in a radiotelephony system.

Fig. 5 shows the forward link power control procedure of the present invention.

Detailed description of the preferred embodiment

The variable data rate communication link method of the present invention allows the data rate of a signal input to a convolutional encoder to be variable, without changing the data rate of the encoded signal. This allows a higher quality voice channel to be used, or a faster fax or data channel, without increasing the fixed output speed of 19.2 kbps (kilobits per second). The variable data rate is obtained by punching a convolutional code of rate ½ in order to obtain a convolutional code of rate ¾. For example, a fixed input data rate of 9600 bps, encoded by a ½ rate convolutional code, produces a fixed output data rate of 9600 x 2 = 19.2 kbps. Equivalently, a fixed input data rate of 14,400 bps, encoded by a convolutional code of rate ¾ produces a fixed output data rate of 14,400 x 4/3 = 19.2 kbps.

The direct communication link procedure of the present invention is illustrated in Fig. 2. The procedure begins with a data signal, I (D), input to the convolutional encoder (201). The procedure allows the data rate of this signal to be variable, and up to 14.4 kbps. The convolutional encoder (201), in the preferred embodiment, is a ½ rate encoder.

The convolutional code has the generator polynomials Gi = 753 and G₂ = 561. In polynomial notation, generator polynomials appear as:

G₁(D) = 1 + D + D² + D³ + D⁵ + D⁷ + D⁸

G₂(D) = 1 + D² + D³ + D⁴ + D⁸

Since this is a / rate encoder 201, for each bit input to encoder 201, two symbols will be output. As an example, if the input signal consists of bits b₀, b₁ and b₂, the output symbol sequences are: C₁₁, C₁₂, C₁₃, C₁₄, C₁₅, C₁₆... for G₁ and C₂₁, C₂₂, C₂₃, C₂₄, C₂₅, C₂₆ ... for G₂. Therefore, without the method of the present invention, the input must be 9.6 kbps in order to maintain the standard 19.2 kbps output of the rate / encoder.

The next step of the procedure inserts a repetition (202 and 203) of each of the emitted symbols in the sequence of symbols. The data rate is set by the speech coder or the data services controller, so it knows how many symbol repetitions need to be inserted to get the proper data rate. In the preferred embodiment, the symbols repeat once, so the output symbol sequences are:

IS 2 273 136 T3

C11, C11, C₁₂, C₁₂, C₁₃, C₁₃, C₁₄, C₁₄, C₁₅, C₁₅, C₁₆, C

.. for Gi and

.. for G₂.

C₂₁, C21, C₂₂, C₂₂, C₂3, C₂3, C₂4, C₂4, C₂₅, C₂₅, C₂₆, C₂₆.

A parallel to serial conversion is performed by a multiplexer (204) on these symbol sequences. The two sequences of symbols enter multiplexer 204 at a rate of 14.4 kbps, and are output from the multiplexer as a single sequence having a data rate of 28.8 kbps. This multiplexing stage generates the sequence of symbols:

C11, C21, C11, C21, C12, C22, C12, C22, C13, C23, C13, C23, C14, C24, C14, C24, C15, C25, C15, C25, C16, C26, C16,

C26 ...

This sequence is punched (205) then using 110101 as the puncture pattern, with every 0 being the punched bit. This pattern is implemented by erasing from the symbol sequence all the bits that are at locations 6n + 3 and 6n + 5, where n is an integer in the range from 0 to ro. Alternative embodiments may puncture the sequence of symbols at different locations and at different speeds. The result of this operation is the following sequence of symbols:

C11, C21, C21, C22, C12, C22, C23, C23, C14, C24, C24, C25, Cu, C25, C26, C26 ...

The symbols then enter a block interleaver (207). Those skilled in the art will appreciate that other types of interleaving can be used in alternative embodiments, without departing from the scope of the present invention. The interleaved data symbols are broadcast by interleaver 207 at the same data symbol rate that they entered: 19.2 kbps. The interleaved symbol sequence enters an input of the XOR combiner (226).

Collation is necessary to reduce the probability that a fading or interference will cause a large gap in the data stream. In the case where the symbols are also repeated, the loss of a symbol will not necessarily cause a total loss of data, thereby providing improved performance.

A long pseudo-noise (SR) generator (220) is coupled to the other input of the XOR combiner (226) to provide an extended sequence to the XOR combiner (226). The long pseudo-noise generator 220 uses a long SR sequence to generate a user-specific sequence of symbols or a unique user code, at a fixed rate of 19.2 kbps in the preferred embodiment. . In addition to providing identification as to which user sent the traffic channel data bits over the communication channel, the unique user code improves communication security on the communication channel by mixing the data bits from the communication channel. traffic. The XOR combiner (226) uses the unique user code entered by the long SR generator (220) to extend the input symbols from Walsh code data to extended user code data symbols. This extension by the XOR combiner (226) provides an increase factor in the overall extension of the traffic channel data bits to data symbols. Extended user code symbols are output from the XOR combiner (226) at a fixed chip rate: 1.228 Mchp / second in the preferred embodiment.

The extended code symbols enter a combiner (260) which is also coupled with a code division channel selection generator (250), which provides a Walsh code of specific length to the combiner (260). Generator 250 provides one of 64 orthogonal codes corresponding to 64 Walsh codes of a 64-row, 64-column Hadamard matrix, in which a Walsh code is a single row or column of the matrix. Combiner 260 uses the specific Walsh code input by code division channel generator 250 in order to extend mixed input data symbols to Walsh code coverage data symbols. Walsh code coverage data symbols are output from the XOR combiner 260 and input to quadrature combiners 227 and 229 at a fixed chip rate of 1.2288 Mchp / second.

A pair of short SR sequences (that is, short compared to the long SR sequence used by the long SR generator (220)) are generated by an I channel SR generator (225) and a generator (228) of Q channel SRs. These SR generators (225 and 228) can generate identical or different short SR sequences. The XOR combiners (227 and 229) further extend the input Walsh code spread data with the short SR sequences generated by the I channel SR generator (225) and by the Q channel SR generator (228), respectively. . The resulting I channel extended code sequence and the resulting Q channel extended code sequence are used to modulate a quadrature sinusoid pair biphasically, regulating the power level controls of the sinusoid pair. The sinusoids are summed, band-pass filtered, translated to an RF frequency, amplified, filtered, and radiated through an antenna to complete the transmission of the symbol sequence over the direct communication link.

In a CDMA cellular radiotelephony system, a method is required in the mobile radio unit to interpret the sequence of symbols transmitted over the direct communication link. This method of the mobile unit of the present invention is illustrated in Fig. 3.

The mobile unit procedure first demodulates the sequence (301) of received symbols. The demodulated signal then enters a deinterleaving procedure (302), to reverse the interleaving of the procedure of the

ES 2 273 136 T3 direct link. The result of this operation is the original sequence of symbols, including the repeating symbols, as entered into the interleaver of the direct link procedure.

The output symbol sequence is then processed to insert the symbols that were erased in the forward link puncture procedure (303). Since the receiving mobile has the same puncture pattern as the base, it knows which symbols were erased and can therefore replace these erased symbols with empty spaces, also known as erasures. The output of this operation is the following, where E is the erasure:

C „, C21^{, E}> C21 » ^AND> ^C22^{, C}^ ^C22^{, E}> ^C23^{, E}> ^C23^{, C}14> C24> ^C2^ ^AND> C24> ^AND> ^C25^{, C}15^{, C}25^{, E}> C26> ^AND. ^C2i

This sequence then enters a warehouse (304) for temporary storage. The buffer allows the Viterbi decoder to process the symbol sequence multiple times in order to determine the data rate.

The Viterbi decoder (305) also assigns a null metric to the erasure bits, as is well known in the art. The output of the Viterbi decoder is digital data that is converted to an analog signal by a digital-to-analog (CDA) converter (306). The analog signal can then be used to drive a speaker (307) in the mobile unit.

The symbols transmitted on the forward and reverse channels are matched to frame formats, each frame being 20 milliseconds long. The co-pending US patent application Serial No. 07 / 822,164 to Padovani et al., And assigned to the awardee of the present invention, sets forth a more detailed explanation of these frames. The amount of data transmitted in each frame depends on the data rate. The composition of frames for each data rate, for the forward and reverse channels, is illustrated in the following table:

<td>No. of raw bits</td><td>Control of Redundancy Cyclical</td><td>Epilogue</td><td>Reserved</td><td>Information bits</td><td>Speed</td>

<td> 288</td><td> 12</td><td> 8</td><td> 3</td><td> 265</td><td> 13250</td>

<td> 144</td><td> 10</td><td> 8</td><td> 2</td><td> 124</td><td> 6200</td>

<td> 72</td><td> 8</td><td> 8</td><td> 2</td><td> 54</td><td> 2700</td>

<td> 36</td><td> 6</td><td> 8</td><td> 2</td><td> 20</td><td> 1000</td>

The rate listed in the table is the information bit rate. The bits reserved for the forward and reverse channels, in the preferred embodiment, are for signaling, power control, and future use.

The transmitting power of the forward channel transmitters can be controlled in the reverse channel by the power control method of the present invention, illustrated in Fig. 5. The power control method will be described as used in the control system. CDMA cellular radiotelephony, but the procedure can be used in other communication systems.

The terrestrial network selector determines the rate at which a frame is sent to a mobile (501) and sends the frame to all base stations communicating with that specific mobile. The selector is part of the base station and is responsible for the call processing requirements of the base station.

During soft handover, more than one base station is communicating with one mobile. The base stations transmit the frame to the mobile (505). After combining the data from various possible base stations, the mobile determines if the last frame (510) has been received and decoded correctly. If the mobile correctly decoded the last frame, the mobile sets the power control bit in the next frame (520) which is transmitted to the base stations.

Since the selector knows the speed at which it transmitted the last frame to the mobile, and now has the response from the mobile as to whether that frame was correctly decoded, the selector compiles a table of statistics (525) on the error rates in those incurred by the mobile station at each speed. The "correctly received" entries in the table are incremented only if the reverse link frame from the mobile, which contains the response bit, was correctly received and decoded (515).

IS 2 273 136 T3

<td></td><td>Transmission to Maximum speed</td><td>Transmission to 1/2 speed</td><td>Transmission to 3/4 speed</td><td>Transmission to 1/8 speed</td>

<td>Correctly received</td><td>I1</td><td>J1</td><td>K1</td><td>L1</td>

<td>Erased</td><td>I2</td><td>J2</td><td>K2</td><td>L2</td>

<td>Total</td><td>I = I1 + I2</td><td>J = J1 + J2</td><td>K = K1 + K2</td><td>L = L1 + L2</td>

<td>Error rate</td><td>I2 / I</td><td>J2 / J</td><td>K2 / K</td><td>^L2^{/ L}</td>

The selector also maintains a table of predetermined target error rates T1, T2, T3, and T4, one for each speed. If the present invention is used in a cellular radiotelephony system, these error rates can be set by the cellular service carrier in order to provide a specific grade of service.

The selector then calculates the following differences:

E1 = I2 / I-T1

E2 = J2 / J - T2

E3 = K2 / K - T3

E4 = L2 / L-T4.

The selector determines the power level with which the next frame is to be transmitted, comparing the respective difference just calculated with zero. For example, if the frame is to be transmitted at a maximum rate and E1&gt; 0 (530), the power level will be P_llolllillal + P (535), where P is a function of the value of E1 and P_nominal is the power level set by the carrier for that geographic area. If E1 = 0 (540), the power level will be P_payrolll (545). If E1 &lt;0, the power level is P_nominal - P (550). The other data rates follow the same procedure. The selector forwards the next frame to be transmitted to the mobile to the base stations that are communicating with the mobile. An indication of the power level with which the frame is to be transmitted is included in this frame.

Alternative embodiments of the present invention insert more than one repetition of each symbol in the sequence of symbols, depending on the data rate input to the encoder. For example, if you input a data rate of 2.4 kbps to the encoder, the symbols should be repeated three more times, with a total of 4 of the same symbols, in the output sequence, in order to maintain the data rate of 19.2 kbps output. By adding more or fewer repetitions, the input data rate can be varied while keeping the output at 19.2 kbps, as required by the Electronic Industries Association / Telephone Industries Association CDMA temporal specification, IS-95.

Alternative embodiments can be punctured first and repeated after the puncture procedure. However, the preferred embodiment does not destroy the symbol, as would be the case if the symbols were punctured prior to the repeat procedure. By repeating first, the repetition of the symbol still exists after the puncture and therefore this information can still be transmitted.

Alternative embodiments may also require an output rate other than the 19.2 kbps required by the CDMA specification for the link between the base station and the mobile. An example of such an embodiment is the mobile station to base station link in which the specification calls for a rate of 28800 bps. In this case, an information rate of 14400 bps, coupled with a convolutional code of rate /, achieves the desired rate of 14400 x 2 = 28800 bps.

By punching a rate code 1/2 to obtain a rate code 3/4, the method of the present invention enables an encoder to support a higher data rate, while the output remains constant. The puncture procedure and code symbol repeat procedure also allow the encoder to support variable data rates such as 14.4, 7.2, 3.6, and 1.8 kbps, while keeping the data rate stable. encoder output at 19.2 kbps, increasing the number of symbol repetitions. By using the lancing procedure on a radiotelephone that has the capability of operating on the CDMA radiotelephony system, higher voice quality and faster data and fax transmissions are achieved.

IS 2 273 136 T3

The rapidly increasing power control method of the present invention allows a mobile to instruct a base station to change its power output at a higher rate. This procedure allows the mobile to send a power change command with each data frame, without degrading the quality of the voice or data.

The degradation of performance associated with the method of puncturing a rate 1/2 code on the base station to mobile station link is more than compensated for by the rapidly increasing power control method of the present invention. The rapidly increasing power control method of the present invention allows a mobile to instruct base stations to adjust their power output at a rate of 50 Hz (for each frame), compared to rates of 0.2 Hz that can be achieved through other signaling procedures that replace entire frames with power control information. This procedure allows the mobile to send a request for a power change with each data frame, using a single bit of information per frame and, therefore, without degrading the quality of the voice and without significantly reducing the data throughput.