FI115177B - Apparatus and Method for Determining Speed in Variable Speed Communication Systems Using Interrupt Keying - Google Patents

Description

115177

The present invention relates to the determination of speed in variable rate communication systems, and in particular to the determination of speed in communication systems using interrupt keying or amplitude modulation to variable rate transmissions.

Efficient use of limited resources has long been an important goal for communication system designers. For example, in point-to-point communication systems - such as telephone modems - this usually involves maximizing the data rate (information bits per second) with a given limited transmitter power and channel bandwidth. In multi-channel systems, such as cellular network sizes, the available resources are shared among multiple users, and the goal of the system designer is to maximize system capacity. This • ·. ·. is often expressed as the number of users that the system can simultaneously serve at a specific user-specific data rate.

i) 4) 1 * »* '* In multichannel systems serving speech transmission, the nature of the speech activity of each user can be used to further improve the efficiency of the communication system. For example, the well-known Digital Speech Interpolation method, summarized by K. Feher in Advanced Digital 'Communications, has long been used to increase the capacity of * * (Trunked) telephone systems.

35 si encoding and digitally encoding each N

A · »115177 2 user speech and multiplexing the resulting packet flows to a single communication channel. Although there is a limited likelihood that some packets will be erased, the effective capacity of the connection will increase to 5.

Related examples are found in cellular systems, such as the GSM (Groupe Special Mobile) system specifications published by the European Telecommunications Stand-10 Ards Institute (ETSI). In this case, the speech activity detector in the digital speech encoder used to encode each user's speech disables the user's transmitter when the user is not speaking. This technique, known as "discontinuous transmission" or "DTX" ("Discontinuous Transmission"), provides an important advantage in the form of a reduction in cellular battery consumption and nominally allows smaller radio frequency (RF) channel re-use distances (because the average burst power of the parallel channel decreases), thus increasing the capacity of the system.

it> • 4 i - *: 'Similar technology is used by the North American Telecom- • munications Industry Association (TIA) Standard IS- • 25 95-A Mobile Station-Base Station Compatibility Standard, * · *' for Dual- Mode Wideband Spread in a cellular system described by the Spectrum Cellular System. This system defines t /: for Code Division Multiple (CDMA)

Access) radio interface (air interface). Since interference between 30 users limits the capacity of such systems, the system attempts to minimize the power transmitted by each user by utilizing voice activity.

V v <I • K K «« k * 115 3 115177 TIA standard IS-95-A describes various techniques for accomplishing this on each downlink (base station to mobile) and reverse (mobile to base) but only the counter-link method is relevant to this review. As shown in Figure 1 (an expanded summary of TIA standard IS-95-A in Figure 6.1.3.1.2), a mobile station (MS) distributes a user's 8 kHz pulse code modulated (100) pulse code modulated speech signal. 20 ms into 10 segments, i.e. frames, and thereafter encode these frames into information packets using a digital speech coder (101). The exact specification of the digital speech encoder (101) is apparent from the TIA standard IS-96-A Speech Service Option Standard for wideband Spread Spectrum Digi-15 or Cellular System.

During encoding, the speech activity detector (107) associated with the digital speech encoder (101) classifies each frame as belonging to one of four different transmission rates.

These are referred to herein as "1/1-speed", "1/2-speed", "1/4-speed" and "1/8-speed". Voice Activity · *. the exact description of the detector (107) is still apparent from the TIA standard · ·] '·, IS-96-A. Suffice it to note that the speech encoder • · · (101) uses more information bits to encode frames that occur during active speech bursts, and uses fewer bits during silent periods * · I “. 1/1-speed uses the most information bits, while ·. · 'Uses 1/8-speed the least. The use of bits with 1/2 and 1/4 rate coded frames - which usually occur during changes between active speech bursts and silent periods - is somewhere in between these limits.

The speed indication generated by the speech activity detector (107): 35 (108) indicates the result of the frame classification process for the digital speech encoder (101). The 1/1-rate and 1/2-rate packets are then subjected to block coding or cyclic coding (102) as defined in TIA Standard IS-95-A, Section 6.1.3.3.2.1 Reverse Traf-5 f ie Channel Frame Quality Indicator. This is followed by channel coding (103) defined as 1/3 rate convolutional coding in TIA Standard IS-95-A, Section 6.1.3.1.3 Convolutional Encoding. At this point, the number of channel-coded bits 10 forming the encoded packet (104) is shown in Table 1.

Selected Number of encoded speed bits 15 in packet 1/1 speed 576 20 1/2 speed 288 1/4 speed 144 1/8 speed 72 25 -

Table 1. Number of encoded bits * 1 per packet without speed.

• · · • 1 ♦ · · *. ' 1 30 After interleaving (105) (defined by TIA standard IS-95-A paragraph «· ♦ 6.1.3.1.5 Block Interleaving), each • 1 · · encoded packet is still being prepared for transmission using 64 (64) -ary) orthogonal modulation followed by direct sequence spreading ^ 35 1.2288 Mc / s using pseudor noise (PN) code (105) (see TIA; IS-95 -A sections 6.1.3.1.6 Orthogonal Modula-ί .1 tion and 6.1.3.1.9 Quadrature Spreading). For the purposes of this presentation, only the following are required for this process. 40 details. Based on Table 1, since the modulation • · · 5 11517? The method is 64-bit orthogonal, 96 symbols (commonly referred to as Walsh symbols) are required to transmit 1/1-rate packets. Correspondingly, 48 symbols are required for 1/2-speed transmission, 24 symbols are required for 1/4-speed transmission and 12 symbols are required for 1/8-speed transmission. The TIA standard IS-95-A specifies that symbols are transmitted in bursts of 6 consecutive Walsh symbols. To this end, the TIA standard IS-95-A divides each 20 ms traffic channel frame into 16 groups, 10 groups called "Power Control Groups" (PCGs), each capable of transmitting one group of 6 Walsh symbols. (See TIA Standard IS-95-A, Section 6.1.3.1.7, Variable Data Rate Transmission). Depending on the selected rate, the number of PCGs during which the MS is actively transmitting may therefore be 16, 8, 4, or 2, respectively, at 1/1 to 1/8. Note that the MS transmitter is deactivated not. during active PCGs.

This process is illustrated in more detail in Fig. 2, which shows the transmitter activity during a 20 ms traffic, · · ·, channel frame (200) at each possible packet size or rate. In Fig. 2, the v · · slash in the PCG time slot (201) indicates that a burst of 6 consecutive straight · · · 25 sequence-scattered Walsh symbols was transmitted; ·· * during this PCG. Thus, if the selected packet rate is 1/1: ", then all 16 PCGs in the 20 ms traffic channel V 'frame are active as shown in the case (202) of Figure 2. If the selected packet rate is 1/2 the rate, then • · The 30 MS transmitter is active only for 8 PCGs as shown in case (203) Similarly, the 1/4 rate packet generates the active PCGs shown in case (204), whereas the 1/8 the velocity case is presented as the case (205).

i · «· · 6 115177

Note that at 1/2, 1/4 and 1/8 speeds (sometimes collectively referred to as "partial speeds"), PCGs that are active in any 20 ms frame are determined by a pseudorandom PCG selection pro-5 major which is controlled by monitoring the user-specific PN code in the traffic channel frame prior to the frame being analyzed, as described in TIA Standard IS-95-A, Section 6.1.3.1.7.2 Data Burst Randomizing Algorithm. More specifically, the last 14 bits of the user-specific PN code generated during the last PCG over the previous frame are stored and used at each transmission rate to select the PCGs active in the next frame. Because the user-specific code has a multi-day repeat period and is transmitted for each user, the location of active PCGs in any given traffic channel frame will therefore depend on the time at which the frame is transmitted, the user identifier and the packet rate transmitted during this frame. Of course, the number of active 20 PCGs remains constant at each speed - only the position of the active PCGs changes • V. pseudorandom by time and user ID.

• · <* t * • · *

It is important to note that - in order to perform despreading - the base station (BS) receiver must re-develop the user-specific PN code used in the MS to perform direct sequence spreading. Thus, the BS receiver may also uniquely identify the PCGs used to transmit the packet at * * ·· 30 at any of four possible rates. Further, since the PCG selection procedure depends only on the user-specific PN code sequence detected during the previous frame, the BS receiver is able to recognize the active: · * / set PCGs at the beginning of the current frame.

:: / 35 * · * 7 115177 Using this variable rate transmission method, TIA standard IS-95-A mobile stations are capable of reducing battery consumption and reducing the average interference power of other IS-95-A mobile communication devices using the same carrier frequency. However, it should be noted that the TIA IS-96-A digital speech encoder (101) and the TIA IS-95-A traffic channel physical layer structure do not provide page information that would indicate to a base station receiver which of the four speeds to move the packet. The base station receiver is thus required to estimate the transmission rate - a process sometimes referred to as "rate determination".

15

Prior methods for performing rate determination include a technique implemented in a Base Station Modem (BSM) integrated circuit implemented by Qualcomm, Inc., San Diego, CA for use with 20 TIA standard IS-95-A transmitters. . A simplified block diagram of this solution is shown in Fig. 3.

* * »·

The procedure begins by converting (300, 301) a received * 4 \ t * 25 radio frequency (RF) waveform comprising each direct sequence spread Walsh symbol, initially from an RF signal at a chip rate displayed - · * '/ J a baseband signal. This requires various known RF, intermediate frequency (VT) and baseband functions such as frequency conversion, auto gain control, symbol sampling, etc., but these need not be defined in detail here. Each of the transferred Walsh symbol waveforms ii is restored t, * * 'after being distorted by noise and the communication channel by despreading (302), «4 en * * 8 115177, which requires correlation on a per-user PN basis. sequence (303) used to spread the transmitted Walsh symbols. Thereafter, detection of Walsh symbols (304) is performed using a known method for correlating a received Walsh symbol waveform to a plurality of 64 waveforms of symbol alphabets in a 64 correlator (312). The exact details of this technique and its performance are known and are described in conventional texts on digital communications, including J.G. Proakis

Digital Communications. The maximum likelihood method for identifying a transmitted Walsh symbol when no received signal phase reference is available (i.e., in "non-coherent" conditions) is to select the correlator and corresponding symbol index that maximizes the complex value of the 64 possible symbol waveforms and received waveforms. This selection process is performed in the block designated 'MAX' in Figure 3 (313). The absolute value of the square of the absolute value of this cross-correlation result having the highest it-20 standing value is referred to herein as "maximum Walsh symbol energy", or "MWSE" ('maximum Walsh symbol energy'), '11. and is shown as an output (315) developed.

The index (314) corresponding to the correlator output with the highest absolute value * 31 · 1 is then transferred to the block deinterleaving block (305), where the channel symbols containing each '[[' Walsh symbol] interleaving is removed and convolutionally decoded (306) using a known Viter-1 30 bi algorithm. The deinterleaving and Viterbi-de-1 ·, '·· encoding processes (305) and (306) are performed four times -: once for each supposed packet rate. During Viterbi decoding (306), »· *.! 'Calculates an estimate of the number of channel errors in the received packet for each of the four possible fphs, by a known method, where the received channel coded symbols are compared with the symbols obtained by performing convolutional coding of the Viterbi decoder output at each rate. The 4-digit vector containing the number of symbol errors by Es-5 by rate is then passed to the rate determining function (309). Finally, the block associated with the 1/1-rate and 1/2-rate packet, i.e., the cyclical codes (referred to as TIA standard IS-95-A called Frame Quality Indicators), are decoded (307) and their syndromes, such as checksums In S. Lin and DJ Costello, Error Control Coding is defined) sets the rate determination function (309) to be used as a 2-digit vector (310).

15

The rate determination is then performed by partitioning the 6-digit solution space formed by the 4-digit symbol error number vector and the 2-digit checksum vector (309) and (310), and then estimating the transmission rate by identifying the region of the solution space corresponding to the received packet. vector located. If the rate of the transmitted packet cannot be reliably determined (i.e., if it is outside the range of resolution of all four rates), then the packet may be reported as "wiping". Such packets are not subjected to any further processing such as speech coding, for example.

This solution is plagued by the disadvantage that it requires 30 separate Viterbi decoding operations (one per each bit rate assumption). This is computationally • V · expensive and inefficient in terms of power consumption. In addition, the audio delay due to the need to perform multiple Vi • · ·· terbi decodings before the specified rate can be transferred to a speech decoder may reduce the perceived audio quality. Therefore, there is a need for a device and method for determining speed in variable rate communication systems that is efficient in terms of computation and power consumption and does not degrade the perceived sound quality.

5

The present invention provides a method for determining the rate in a variable rate communication system. According to a first embodiment of the invention, a benchmark is calculated based on the symbol packet energy 10 of the data packet for each of a plurality of possible packet rates transmitted, and an optimal rate is selected based on the calculated benchmarks. According to another embodiment, a plurality of metrics is calculated based on the sum of symbol energy of a plurality of symbols each comprising a plurality of transmitted symbol rates, and the rate associated with the largest calculated index is selected.

According to a third embodiment, the present invention 20 provides a device for determining a rate in a variable rate communication system. Such a j ': device comprises a plurality of battery registers of the largest Walsh:; ; selectively accumulating symbol energy; and a selector coupled to the battery outputs for selecting the optimum speed based on these outputs. According to a fourth embodiment, the invention provides a method for determining a rate in a variable rate communication system that accumulates a plurality of maximum Walsh symbol energy values over a traffic channel frame and; selecting the optimum speed based on said values.

11, 115177

FIG. 1 illustrates a generally known method for performing speech and channel coding and modulation of a reverse link traffic channel frame and direct sequence spreading in a communication system 5 comprising a variable rate voice service option.

Figure 2 illustrates an example of the structure of a prior art power control group (PCG) structure of a specific reverse link traffic channel when a variable rate voice service option is used.

Figure 3 illustrates a prior art method for determining the rate of reverse link traffic channel packets.

FIG. 4 is a block diagram illustrating a preferred embodiment of the rate determining device of the present invention.

FIG. 5 is a block diagram illustrating an alternative preferred embodiment of the rate determining device of the present invention.

; ·: Y of the preferred embodiment of the speed determining device; a block diagram is illustrated in Figure 4. Figure 4 illustrates an RF conversion (300, 301) and assembly process (302) used to restore a noise-containing orthogonal waveform corresponding to each transmitted Walsh symbol comprising a traffic channel frame. It also discloses a user-specific PN code generator (303) used to generate the spreading sequence, and a Walsh symbol detector (304) comprising a 64-correlator (312) and a select function (313). The Walsh symbol detector is shown to generate the maximum Walsh symbol energy (MWSE, maximum Walsh symbol energy) (315); * 35.

,, 115177 12

The MWSE of each received Walsh symbol is then transferred to four accumulator (409-412) batteries, abbreviated as "1/1-battery" 5 (409) - "1/8-battery" (412 ), which means "1/1-speed battery" (409), etc. The contents of these batteries are set to zero uniformly at the beginning of each traffic channel frame. Adapters (405-408) associated with each battery are keyed to the control signals la from the PCG selector function (400), and they accumulate MWSE only if the corresponding PCG flag has a logical value of '1'. The PCG selector function (400) monitors the output of the user-specific PN code generator during the previous traffic channel frame and identifies the PCGs that are specified by the TIA standard IS-95-A paragraph 6.1.3.1.7.2 Data Burst 15 Randomizing Algorithm. would be active in the current traffic channel frame at each possible transmission rate. The PCG selector (400) prints 4 binary flags (401-404) labeled "1/1-speed PCG-flag" to "1/8-speed 20-PCG flag". These flags are logically "l" if and only if the Walsh symbol demodulated by the detector (304) is a • member of the active PCG at the rate associated with this flag.

;; In other words, and with reference to the example of Figure 2, 1/2-no-; . the PEG flag is logically "1" only during the dashed portions of the traffic channel corresponding to the 25 frame cases (203). Correspondingly, the value of the 1/4 rate PCG flag is "1" only during the diagonal lines of case (204), and so on. This process continues until all 96 Walsh symbols in the frame have been received.

• · · <'· 30 At the end of each frame, "1/1-Battery" has accumulated 96 MWSE, "1/2-Battery" has accumulated 48 MWSE, "1/4-Battery" has accumulated 24 MWSE values and the "1/8 n battery" has accumulated 12 MWSE values.

• I

• I I

• 1 · I »« »• · • · · 13 115177

After the entire frame is received, the contents of the batteries (409-412) are fed to the reducers (417-420). The scalar value, which is generally different for each battery and denoted M-1/1 to M-1/8 (413-416) in Figure 4, is then subtracted from the output of each battery (409-412). The resulting output of each subtractor (417-420) is then fed to a selector (421) which indicates the highest of the outputs of the four subtractors (417-420). This unambiguously indicates which of the batteries (409-412) led to the 10 highest values with the dial (421). The estimated transfer rate (422) is then identified as the transfer rate corresponding to the battery (409-412) that produced this maximum value with the selector (421). The estimated rate (422) is used to control the operation of the deinterleaver (423) and the Viterbi decoder 15 (424). These devices only perform a one-time operation to decode the received channel symbols according to the predicted rate (422).

Note that the scalar values M-1/1 to M-1/8 (413-416) are generated in advance by computer simulation or bench • testing. In practice, they are usually constant. . the total duration of the traffic channel frame to be analyzed. No. , however, may change in value at the limit of 25 of each Walsh symbol depending on other parameters associated with the base station receiver. One particular example of this arises from the use of a "rake receiver" to utilize multipath signal components on a communication channel. [Rake receivers (30 ceivers) are known in the art and need not be described herein.] This is illustrated in Fig. 5 in the case of a 4 element rake receiver, where each element (500) comprises at least a despreader (302) and 14,115,177 to a 64-digit correlator (312). Note that more or less than 4 elements (500) may be used depending on the application. In Figure 5, each element is assigned to a different multipath signal component, and the delay element 5 is At-A4. (502-505) compensate for the difference in observed delays for each multipath signal component. Each non-lm non-multipath signal component is then transmitted through a 64-digit correlator (312) of the same type as described above. A correlation to the output of each market (312) is then formed into a 64-digit vector, wherein the ith element of the 64-digit vector is the square of the absolute value of the uncorrelated correlator output. This results in the classic non-coherent method of combining called "square-15 law" combining in conventional literature such as J.G. Proakis in Digital Communications. Figure 5 combines (522) 4 such vectors by a simple addition of vectors. Subsequently, (523) the largest Walsh symbol energy (MWSE) (315) is identified as the highest value element of the resulting 20 64-bit vector, and subsequently this MWSE (315) is used to determine the velocity as shown in Figure 4.

»« T \\\ However, the relative intensity of each multipath component may vary over time. Thus, the number of elements (500) operating on multipath components that significantly affect the vector combining process (522). 'Amount may change. Provided that the RF converter (301) includes an auto gain adjustment degree, the relative partial effect of each element * 500 can be estimated: using a simple signal-to-noise ratio (SNR) estimator; (501), such as that shown in Figure 5. This SNR Estimator works by comparing the MWSE (506-509) of individual 64-i 35 vectors for each element to correlate.

* I

15 115177 r (312) to the threshold T (510). If the MWSE (506-509) of a given element has exceeded T (510), then the corresponding 64 vector at the output of the 64 correlator of this element is included in the merge process 5 (522), otherwise it is excluded. If an element is declared to be included in the merge process, it is said to be "in-lock". The lock indicator for each element is shown in Fig. 5 as binary lock flags L1-L4 (515-518). Figure 5 also shows a counter (519) that cumulates the number of elements in each lock by monitoring L1 to L4 (515-518). This number (520) is used to obtain the address of a row in the 4x4 lookup table (521). The horizontal rows in the lookup table contain the values of M-1/1 (413) to M-1/8 (416), which are used according to the number of elements in the lock. The contents of the lookup table are generated in advance through computer simulation or bench testing.

Obviously, the estimated rate (430) does not need to be determined solely by the reference numbers available at the output of the subtractors (417-420) in Figure 4. Instead, these metrics could be used as additional information to determine the rate, for example, based on the symbol error rate, or the metric derived from the Viterbi decoder. It is also clear that this technique can easily be extended to determine the rate of variable rate transmission:: from a source of information other than speech. This could include variable-speed data transmission.

Thus, it is immediately apparent that the present invention provides a power and computationally efficient device and method for determining a speed that does not adversely affect the perceived speech quality. In addition, 35 other advantages and modifications will be apparent to those skilled in the art. The invention, in its broader aspects, is not limited to the specific details, representative device and illustrative examples described herein. Various changes and modifications to the above definition are possible without departing from the scope or spirit of the present invention and it is intended that the present invention encompass all such changes and modifications insofar as they are within the scope of the following claims and their equivalent instances.

* ♦

Claims (10)

1. A method of rate determination in a variable speed communication system, characterized in that the method comprises the following steps: a comparison number is calculated (520) on the basis of the symbol energy of a data packet for each speed of a plurality of possible transmitted packet speeds; and an optimum velocity is selected (521) on the basis of the calculated comparisons. A method according to claim 1, characterized in that the step of calculating a comparison number comprises calculating a comparison number on the basis of symbol energy depending on a number of elements of a RAKE receiver which are in use. 20 Method according to claim 1, characterized in that it further comprises a step for calculating at least one additional comparison based on one of the following: a convolutional encoder's distance comparison number, a Viterbi decoder's comparison number, a block coding comparison number; and that: the step of vai further comprises vai of an optimum v speed on the basis of the selected comparison number and at least one additional comparison number. ·; · 30 «» »I Method for determining velocity in a variable speed data system, characterized in that the method comprises the following steps: i i m t *; A plurality of comparative numbers is calculated (417) on the basis of summing the symbol energy of a plurality of symbols, each comprising a plurality of transmitted symbol velocities; and a velocity associated with the largest calculated comparison number is selected (422). Method according to claim 4, characterized in that it further comprises calculating at least one additional comparison number on the basis of one of the following: a convolution encoder's distance comparison number, a decoder's comparison number, a block coding comparison number; and that the step of wai further comprises wai of an optimum speed on the basis of the selected comparison number and at least one additional comparison number. Device for speed determination in a variable speed communication system, characterized in that the device comprises: | ] a plurality of accumulator registers (409-412) for selectively cumulating a greatest Walsh symbol energy over a traffic channel frame, with each accumulator exhibiting an output; and V 'a selector (421) which is coupled to the output of the accumulators for optimum velocity on the basis of these outputs. * «! »» T Device according to Claim 6, characterized in that it further comprises a flag generator (400) of a% power control group, which is:; 35 connected to each accumulator register, and that in the accumulator registers I selectively accumulate a greatest Walsh symbol energy by the influence of the power control group's flag. Device according to claim 6, characterized in that the selector comprises a plurality of subtractors (417-420), which are correspondingly coupled to the outputs of said plurality of accumulator registers to subtract a scalar value from said outputs. Method for determining velocity in a variable speed communication system, characterized in that the method comprises the following steps: a plurality of largest Walsh symbol energy values are cumulated (409 - 412) over a traffic channel frame; and an optimum speed is selected (422) on the basis of said values. Method according to claim 9, characterized in that the step for cumulation of a plurality of largest Walsh symbol energy values comprises cumulation of a plurality of largest Walsh symbol energy values based on the power control groups containing the traffic channel frame. > I i »» f; * f t i t »-