Method and apparatus for achieving data rate variability in orthogonal spread spectrum communications systems

<u>Field of the Invention</u>

The present invention relates to cellular and other communication systems, and more particularly, to methods and apparatus for achieving additional data rates in communication systems utilizing code division multiple access (CDMA).

<u>Background of the Invention</u>

When multiple users share a transmission medium, some form of multiplexing is required to provide separable user sub-channels. There are many multiplexing techniques available which simultaneously transmit information signals within the available bandwidth, while still maintaining the quality and intelligibility that are required for a given application. Code division multiple access (CDMA) techniques, for example, transmit multiple information signals on the same channel, and differentiate each user sub-channel by encoding it with a unique spreading code.

Code division multiple access (CDMA) techniques have been employed in many digital wireless communication systems to permit a large number of system users to communicate with one another. Current CDMA networks were designed to carry only voice traffic and exhibit limited data rate variability. CDMA networks, however, must evolve to encompass a variety of multimedia applications, each having potentially different data rates. Thus, CDMA networks will be required to carry information associated with the various multimedia applications at various data rates, corresponding to the requirements of diverse wireless services demanded by customers.

Many communication systems utilizing code division multiple access (CDMA) comply with the IS-95 standard, adopted by the Telecommunication Industry Association (TIA). Under the IS-95 standard, a communication system substantially eliminates co-channel interference and improves the bit energy-to-noise density ratio, E_b/N_o, on the forward link from a base station or cell site to a mobile receiver unit by modulating the information signals with Walsh orthogonal function sequences. To produce corresponding orthogonal information signals, these CDMA systems require that the forward link information signals be transmitted in a synchronized manner. A more detailed discussion of the IS-95 standard is provided in "Mobile Station-Base Station Compatibility Standard for Dual-Mode Wideband Spread Spectrum Cellular System," Telecommunication Industry Association Doc. No. TIA/EIA/IS-95 (1993), incorporated by reference herein.

Current implementations of the IS-95 standard, however, allow only a limited number of data rates. Specifically, the Telecommunication Industry Association has recently adopted a new standard, IS-95B, for increasing the data rates that may be achieved with CDMA networks. The IS-95B standard, however, only permits data rates that are integer or power-of-two multiples of the basic CDMA rate. While such techniques satisfy the data rate needs of many applications, the utility of CDMA networks could be significantly extended if further data rate variability could be achieved.

<u>Summary of the Invention</u>

Generally, a CDMA communication system offering a near continuum of data rates up to a maximum rate is disclosed. The disclosed CDMA communication system increases the data rate variability, without disturbing the orthogonality between users. An increased number of data rates is obtained by time multiplexing the data rates achievable with conventional CDMA systems to provide additional data rates for various multimedia applications.

In accordance with one feature of the present invention, if a user is assigned a specific orthogonal code sequence, such as a Walsh sequence, which permits a nominal data rate, R, and higher data rates, <i>n</i>R, the user can obtain additional desired data rates by time multiplexing the specific orthogonal code set. In this manner, desired bit rates are provided for applications requiring bit rates that are not an integer multiple of the rates provided by conventional "fat pipe" techniques.

In an illustrative sixteen dimension implementation, if a specific code quadruple, such as the code quadruple <i>(wk,w</i>_<i>k</i> + <i>4, w</i>_<i>k</i>₊_<i>8</i><i>, w</i>_<i>k</i>₊_<i>12</i>) (k=0, 1, 2, 3), is assigned to an information source, any rate can be achieved, up to a rate of 4R. Rates R, 2R, 4R and 16R can be achieved by assigning other appropriate code sets to the information source, in a conventional manner. According to a feature of the invention, any desired rate up to 4R is achieved by time multiplexing the codes <i>w</i>_<i>k</i><i>, w</i>_<i>k</i>^<i>2</i><i> and w</i>_<i>k</i>^<i>4</i><i>.</i> The symbol, <i>w</i>_<i>k</i>^<i>n</i>, denotes the first <i>n</i>th of the spread spectrum code <i>w</i>_<i>k</i> (hence, n=2 denotes the first half of the sequence). Thus, the user can achieve any data rate that is a rational scale factor combination of the rates R and 4R.

The transmitter of the present invention achieves a rate conversion to match the appropriate codeword to the appropriate data rate, by employing buffering and reclocking of the data, with time multiplexing of the orthogonal Walsh codes corresponding to the available lower and upper rates. The rate conversion must be synchronized with the selection of the orthogonal Walsh code (corresponding to the lower and upper rates). For example, an application requiring a data rate between 2R and 4R, such as 2.3R, would require multiplexing of the 2R and 4R Walsh codes (<i>w</i>_<i>k</i>^<i>2</i><i> and w</i>_<i>k</i>^<i>4</i>) to achieve the desired 2.3R rate. If the 2.3R application was assigned the specific code quadruple (<i>w</i>_<i>1</i><i>, w</i>_<i>5</i><i>, w</i>_<i>9</i>, <i>w</i>_<i>13</i>), the sequence <i>w</i>_<i>1</i>^<i>2</i> is utilized to achieve a rate of 2R and the sequence <i>w</i>_<i>1</i>^<i>4</i> is utilized to achieve a rate of 4R, with an appropriate balance to achieve an overall 2.3R rate, and the codes <i>w</i>_<i>5</i><i>, w</i>_<i>9</i><i>, w</i>_<i>13</i> are excluded from use.

<u>Brief Description of the Drawings</u>

<ul id="ul0001" list-style="none"><li>FIG. 1 is a block diagram of a conventional CDMA transmitter that provides user streams (or sub-channels) at a single data rate;</li><li>FIG. 2 is a table of an illustrative set of orthogonal spread spectrum codes, <i>w</i>_<i>0</i> through <i>w</i>_<i>15</i>,</li><li>FIG. 3 is a block diagram of a conventional CDMA transmitter that provides at least one information signal at an integer multiple of the data rate of FIG. 1;</li><li>FIG. 4 is a block diagram of a conventional CDMA transmitter that provides at least one information signal at two times the data rate of FIG. 1;</li><li>FIG. 5 is a block diagram of a conventional CDMA transmitter that transmits at least one information signal at four times the data rate of FIG. 1;</li><li>FIG. 6 is a block diagram of one sub-channel of a CDMA transmitter offering data rate variability in accordance with one embodiment of the present invention;</li><li>FIG. 7 is a block diagram of one embodiment of the rate converter of FIG. 6; and</li><li>FIG. 8 is a block diagram of a CDMA receiver offering data rate variability in accordance with one embodiment of the present invention.</li></ul>

<u>Detailed Description</u>

The present invention relates to a method and apparatus for modulating a communication signal in a code division multiple access (CDMA) environment using orthogonal spread spectrum codes. As discussed further below in conjunction with FIGS. 1 through 3, conventional CDMA spread spectrum modulation techniques with orthogonal codes only permit data rates that are integer or power-of-two multiples of the chip rate. The present invention improves on conventional CDMA spread spectrum modulation techniques by time multiplexing the data rates achievable with conventional CDMA systems to provide additional data rates for various multimedia applications. In one preferred embodiment, a continuum of data rates are provided.

As previously indicated, it is often desirable to differentiate a number of information signals transmitted on the same carrier frequency with unique orthogonal spread spectrum codes. This is typically accomplished by encoding each bit to be transmitted using a predefined codeword, or pattern, consisting of <i>n</i> "chips" or "signal elements." In standard antipodal encoding, the predefined codeword may be utilized to represent a value of binary "0" and the inverse of the predefined codeword may be utilized to represent a value of binary "1."

A number of orthogonal spread spectrum codes, consisting of a number of consecutive positive and negative signal elements, such as Walsh codes, have been discovered which have unique properties that optimize the detection of the transmitted information. Under the IS-95 standard for CDMA networks, for example, sixty four different Walsh codewords, <i>w</i>_<i>0</i> through <i>w</i>_<i>63</i>, each consisting of sixty four chips, permit 64 different information signals to be transmitted on the same carrier frequency. Since a number of channels are reserved for administration, such as the pilot, synch and paging channels, less than the available sixty four channels typically transmit user information.

In the illustrative embodiment discussed herein, sixteen different Walsh codewords, <i>w</i>_<i>o</i> through <i>w</i>_<i>15</i>, each consisting of sixteen chips, permit up to sixteen different information signals, <i>d</i>_<i>0</i> through <i>d</i>_<i>15</i><i>,</i> to be transmitted on the same carrier frequency. Thus, as shown in FIG. 1, in order to transmit data on the downlink portion of a conventional CDMA system from one or more information sources, such as sources 101-116, to a plurality of end users, such as end users 181-196, a transmitter 100 encodes each of sixteen data streams, <i>d</i>_<i>0</i> through <i>d</i>_<i>15</i><i>,</i> to be transmitted using one of sixteen different Walsh codewords, <i>w</i>_<i>0</i> through <i>w</i>_<i>15</i><i>.</i> The illustrative Walsh codewords, <i>w</i>_<i>0</i> through <i>w</i>_<i>15</i><i>,</i> are shown in FIG. 2. The encoded signals will then be combined and modulated in a conventional manner, prior to transmission over a transmission medium 130. The transmission medium 130 may be embodied as a conventional or wireless communications network. The modulator may employ a modulation technique, for example, which multiplies the codeword by a sinusoidal carrier wave in order to shift the signal frequency upward to the carrier frequency (not shown). In this manner, the original signal spectrum may be translated into a particular frequency band allocated by the Federal Communications Commission (FCC) or another regulatory body.

Upon receipt of the transmitted signal by the receiver 150, the frequency of the received signal is typically first shifted down (not shown) to the base band by a demodulator, thus returning the signal to its original form prior to modulation. Thereafter, the received signal is passed through a series of filters, such as filters 161-176, that are each matched to the characteristics of the appropriate codeword, <i>w</i>_<i>0</i> through <i>w</i>_<i>15</i>. It is noted that the receiver 150 may be associated with all sixteen end users 181 through 196, as shown in FIG. 1. More typically, however, each end user, such as end user 181, will have its own receiver 150. For a more detailed discussion of a prior art CDMA communication system, see U.S. Patent No. 4,901,307, incorporated by reference herein.

For the transmitter 100 shown in FIG. 1, each of the information sources 101-116 transmits at a uniform rate, R, and the symbol duration is equal to the Walsh code duration. Additional data rate variability can be achieved within the implementation shown in FIG. 1, by assigning multiple orthogonal spread spectrum codes, such as Walsh codes, to the same high rate information source, such as the source 101. For example, in order to achieve a transmission rate, such as 3R, for a single source 101, three Walsh codewords (sub-channels) can be assigned to the information source 101. Thus, the set of achievable data rates are integer multiples of the symbol rate, R. As shown in FIG. 3, multicode source 101 will require additional buffering, typically in the form of a serial-to-parallel converter 310. It is noted that the transmitter 100' shown in FIG. 3 can combine any of the sixteen sub-channels associated with the codewords, <i>w</i>_<i>0</i> through <i>w</i>_<i>15</i>, to increase the data rate for a single information source 101 to a maximum of 16R.

A second well-known method for increasing the transmission rate in a code division multiple access (CDMA) network, commonly referred to as the "fat pipe" method, "punctures" the set of illustrative Walsh codes shown in FIG. 2. As discussed below, the "fat pipe" method achieves data rates that are power-of-two multiples of the chip rate, without requiring buffering of the multiple rate user. In other words, the achievable data rates are R, 2R, 4R, 8R and 16R (for the illustrative 16 dimensional Walsh codes shown in FIG. 2). Specific code pairs, such as the code pairs (<i>w</i>_<i>0</i>, <i>w</i>_<i>8</i>) or (<i>w</i>_<i>1</i><i>, w</i>_<i>9</i>), are assigned to the double rate users, transmitting at rate 2R Each double rate user, however, will encode data with only the first half of one of the assigned orthogonal spread spectrum codes. The symbol, <i>w</i>_<i>k</i>^<i>2</i>, denotes the first half of the spread spectrum code <i>w</i>_<i>k</i> shown in FIG. 2. Thus, as shown in FIG. 4, for the spread spectrum code pair, (<i>w</i>_<i>1</i><i>, w</i>_<i>9</i>) assigned to the information source 410, <i>w</i>_<i>1</i>^<i>2</i> is utilized to encode data at a rate of 2R, and <i>w</i>_<i>9</i> is excluded from use by all information sources. It is noted that for each pair (<i>w</i>_<i>k</i>, <i>w</i>_<i>k+8</i>), <i>w</i>_<i>k+8</i>^<i>2</i> = <i>w</i>_<i>k</i>^<i>2</i>, and the second half of the spread spectrum code <i>w</i>_<i>k+8</i> is the opposite polarity of the spread spectrum code <i>w</i>_<i>k</i>.

Likewise, specific code quadruples, such as the code quadruples (<i>w</i>_<i>0</i><i>,w</i>_<i>4</i><i>, w</i>_<i>8</i><i>, w</i>_<i>12</i>) or (<i>w</i>_<i>1</i>, <i>w</i>_<i>5</i><i>, w</i>_<i>9,</i><i> w</i>_<i>13</i>), are assigned to each quadruple rate user, transmitting at a rate 4R. Thus, as shown in FIG. 5, for the orthogonal spread spectrum code quadruple (<i>w</i>_<i>0</i><i>,w</i>_<i>4</i><i>, w</i>_<i>8</i><i>,w</i>_<i>12</i>) assigned to an information source 510, <i>w</i>_<i>0</i> is utilized to encode data and <i>w</i>_<i>4</i>, <i>w</i>_<i>8</i>, and <i>w</i>_<i>12</i> are excluded from use by all information sources. Similarly, for the spread spectrum code quadruple (<i>w</i>_<i>1</i><i>, w</i>_<i>5</i><i>, w</i>_<i>9,</i><i> w</i>_<i>13</i>) assigned to an information source 520, <i>w</i>_<i>1</i> is utilized to encode data and <i>w</i>_<i>5</i>_,<i>w</i>_<i>9</i> and <i>w</i>_<i>13</i> are excluded from use by all information sources. The symbol, <i>w</i>_<i>k</i>^<i>4</i><i>,</i> denotes the first quarter of the spread spectrum code <i>w</i>_<i>k</i> shown in FIG. 2. It is noted that for each code quadruple <i>(w</i>_<i>k</i><i>, w</i>_<i>k+4</i><i>, w</i>_<i>k+8</i><i>, w</i>_<i>k+12</i><i>), w</i>_<i>k+12</i>^<i>4</i><i> = w</i>_<i>k+8</i>^<i>4</i><i> = w</i>_<i>k+4</i>^<i>4</i><i> = wk</i>^<i>4</i>, and the last three-quarters of each spread spectrum code <i>w</i>_<i>k+</i>4, <i>w</i>_<i>k+</i>8, <i>w</i>_<i>k+</i>12 is the same or opposite polarity of the spread spectrum code <i>w</i>_<i>k</i>. It is further noted that the "fat pipe" method does not require buffering of the higher rate users, the codes are not summed, and the signal's peak-to-average ratio is not increased.

According to a feature of the present invention, additional data rates are achieved while maintaining the orthogonal signal structure discussed above for conventional code division multiple access (CDMA) transmitters. In accordance with the present invention, if a user is assigned a specific code set in the manner described above, which permits a maximum data rate <i>n</i>R, the user can obtain additional desired data rates up to the maximum rate <i>n</i>R, by time multiplexing the specific code set. In order to achieve a non-integer rate with conventional systems, however, such as a rate of 2.3R, conventional techniques would require "dummy" data, such as all zeros, to be added to the transmitted stream to bring the presented rate up to the next greatest available "fat pipe" rate, such as 4R. Thus, the present invention provides desired bit rates for applications that have a bit rate that is not an integer multiple of the rates provided by the "fat pipe" technique discussed above.

For example, if a specific code quadruple, (<i>w</i>_<i>k</i><i>,w</i>_<i>k+4</i><i>, w</i>_<i>k+8</i><i>, w</i>_<i>k+12</i>) (k=0, 1, 2, 3), is assigned to an information source 610, any rate can be achieved, up to a rate of 4R. While the user can achieve a fundamental rate of R in the manner described above in conjunction with FIG. 1 by employing the spread spectrum code w₀, a "fat pipe" rate of 2R in the manner described above in conjunction with FIG. 4 by employing the spread spectrum code w₀², or a "fat pipe" rate of 4R in the manner described above in conjunction with FIG. 5 by employing the spread spectrum code w₀⁴ in the present invention, the user can achieve desired additional lower rates by time multiplexing the codes w₀, w₀² and w₀⁴, to achieve rates between R and 4R. In other words, the user can achieve any data rate that is a rational scale factor combination of the rates R and 4R.

As shown in FIG. 6, the transmitter 600 of the present invention achieves a rate conversion by employing buffering and reclocking of the data, with time multiplexing of the orthogonal Walsh codes corresponding to the available lower and upper rates. As discussed further below, the rate conversion must be synchronized with the selection of the orthogonal Walsh code (corresponding to the lower and upper rates). For example, the illustrative 2.3R application would require multiplexing of the 2R and 4R Walsh codes to achieve the desired 2.3R rate. If the 2.3R application was assigned the specific code quadruple (<i>w</i>_<i>1</i><i>, w</i>_<i>5</i>, <i>w</i>_<i>9,</i><i> w</i>_<i>13</i><i>),</i> the sequence w₁⁴ is utilized to achieve a rate of 4R, and the codes <i>w</i>_<i>5</i><i>, w</i>_<i>9</i><i>, w</i>_<i>13</i> are excluded from use.

Thus, when the high rate service is requested, by either the mobile terminal (end user) or the base station (information source), the sync channel (or another administration channel) is preferably utilized to coordinate the wireless service. The originating entity will send a message either requesting or confirming the rate, as well as the particular codes desired. For example, if the mobile terminal requests a channel having a 2.3R rate, the base station will grant or deny the request, and supply the assigned sequences, if available. In one preferred embodiment, the base station will supply the mobile terminal with the assigned sequences, as well as the amount of data that should be transmitted at each of the lower and upper rates. Thus, the sync channel protocol may consist of a message containing a user identifier (indicating, for example, the mobile terminal), a sequence number (indicating the particular Walsh code, <i>w</i>_<i>0</i> though <i>w</i>_<i>15</i>, to utilize), the number of repetitions, N_L and N_U, and the rate exponents, R_L and R_U, for the lower and upper rates, respectively (such as 2R, 4R, or 8R). Thus, from the sequence number and the upper rate, R_U, the mobile terminal knows which other sequences must be excluded.

The number of repetitions, N_L and N_U, is the number of symbols to transmit at each of the lower and upper rates, respectively. The number of repetitions is determined by the following equation:<maths id="math0001" num=""><img file="EP0939508A2_D0001.tif" /></maths> where R_final is the desired rate to be achieved and R is the fundamental rate. Thus, to achieve a desired rate of 2.3R, R_L = 2 and R_H = 4, one solution is N_L = 17 and N_U = 6. One method for determining this solution recognizes that 2.3 can bo written as a ratio of the integers 23 and 10. Writing R_final as this ratio, and using the above equation, leads to a system of linear equations:<maths id="math0002" num=""><math display="block"><mrow><msub><mrow><mtext>4N</mtext></mrow><mrow><mtext>L</mtext></mrow></msub><msub><mrow><mtext> + 4N</mtext></mrow><mrow><mtext>U</mtext></mrow></msub><mtext> = 23</mtext><mtext mathvariant="italic">n</mtext></mrow></math><img file="EP0939508A2_D0002.tif" /></maths><maths id="math0003" num=""><math display="block"><mrow><msub><mrow><mtext>2N</mtext></mrow><mrow><mtext>L</mtext></mrow></msub><msub><mrow><mtext> + N</mtext></mrow><mrow><mtext>U</mtext></mrow></msub><mtext> = 10</mtext><mtext mathvariant="italic">n</mtext><mtext>,</mtext></mrow></math><img file="EP0939508A2_D0003.tif" /></maths> where <i>n</i> is an arbitrary integer. These equations can be solved repeatedly for <i>n</i> = 1, 2, 3, ..., until a solution is found for which N_L and N_U are both integers.

Thus, as shown in FIG. 6, the transmitter 600 of the present invention includes a rate converter 620 and a clock 630 to buffer and reclock the data, respectively, received from a high rate source 610. The rate converter 620 must be synchronized with the selection of the orthogonal Walsh code (corresponding to the lower and upper rates), <i>w</i>_<i>k</i>^<i>L</i> and <i>w</i>_<i>k</i>^<i>U</i> by a switch 640. The output of the rate converter 620 exhibits two different bit durations, and the Walsh codes, <i>w</i>_<i>k</i>^<i>L</i> and <i>w</i>_<i>k</i><i>U</i>, are used accordingly to multiply the data by a multiplier 650. Thus, the information source has been matched to the existing fat pipe rates.

FIG. 7 illustrates one implementation of the rate converter 620. Thus, the rate converter 620 includes two buffers 710 and 720. The first buffer 710 fills with data at a constant rate, and the second buffer 720 is read by an asymmetric multiplexer 730 for every (N_L + N_U) bits. Thus, for the 2.3R example discussed above, 17 bits (N_L) are read out of the buffer 720 at a rate of 2R (R_L) and the last 6 bits (N_U) are read out of the buffer 720 at a rate of 4R (R_U). Thus, the clock at the input of the buffer 710 is symmetric (evenly spaced), while the clock at the output of the asymmetric buffer 720 is asymmetric. The clock 630 controls the buffer 720, the asymmetric multiplexer 730 and the selection of the Walsh code by the switch 640 for the multiplier 650.

FIG. 8 illustrates one implementation of a receiver 800 in accordance with the present invention. The receiver 800 includes a multiplier 810, which selectively multiplies the received data by the appropriate Walsh code, as selected by a switch 820, an integrate, dump, detect block 830 for determining the received bit polarities, a rate converter 840 and a clock 850. The rate converter 840 operates in a similar manner to the rate converter 620, discussed above in conjunction with FIGS. 6 and 7. The selection of the low rate or upper rate Walsh waveform at the multiplier 810 is tied to the integrator 830 and the input clock of the rate converter 840. In the receiver 800, the input clock of the rate converter 840 is asymmetric while the output clock is symmetric.

It is to be understood that the embodiments and variations shown and described herein are merely illustrative of the principles of this invention and that various modifications may be implemented by those skilled in the art without departing from the scope and spirit of the invention.