Features of WCDMA

CHAPTER 1 Introduction

Introduction to WCDMA:

UMTS W-CDMA FDD is a direct-sequence CDMA system with a nominal bandwidth of 5 MHz. The second system, UMTS W-CDMA TDD, also uses CDMA with a bandwidth of 5 MHz, but now the frequency band is time shared in both directionsone half of the time, it is used for transmission in the forward direction and the other half of the time in the reverse direction. cdma2000 is a multi-carrier, direct-sequence CDMA FDD system. Like cdmaOne, its first phase is expected to use a single carrier with a bandwidth of 1.25 MHz. In the second phase, it may have as many as three carriers. In this system, even though each carrier has a nominal bandwidth of 1.25 MHz, the total bandwidth required is 5 MHz.

Features and Benefits of WCDMA:

Most "no insider" followers of the wireless industry most likely think of the wonderful multimedia capabilities that 3G services hope to offer. Downloading the latest photo of their children while driving down the highway away on business, having a video conference while on the train, or playing a multiplayer video game during a school breakthese are the images most would have of what 3G could offer. The reality is many engineers and executives are looking forward to 3G technology because of the added voice capacity and spectrum efficiency it can bring, with the high-speed data serving as another revenue booster along the way. It is expected that the 5-MHz channel in WCDMA will be able to support more than 100 simultaneous voice calls. And because WCDMA uses CDMA, there are no frequency reuse issues to deal with. Earlier it was mentioned that WCDMA was developed without any constraints from needing to be backward compatible, as it would deployed from scratch. This is not altogether correct. While the air interface is entirely new, the format was designed to operate on the same core network as GSM, GSM-MAP. This means WCDMA radio network subsystems (RNSs), the UMTS name for the BSS, can be deployed right alongside GSM BSSs and use the same core network, which includes the MSC, equipment identity register (EIR), HLR, VLR, and AC. In addition, as most networks will use GPRS systems, support for GPRS data is compatible with the WCDMA system via the use of GPRS support nodes. Thus, even though different RF subsystems will be needed to support 2G, 2.5G, and 3G formats in the GSM-to-WCDMA transition, the core network can remain the same and thus offer multimode services to a common area .WCDMA, when compared to the GSM (TDMA) system it will complement and eventually replace, will utilize many new techniques to allow for the added voice capacity and high-speed data rates. Most of these new (in reference to GSM) techniques are already used in the CDMA2000 system. They include soft handoffs (in WCDMA, as in GSM, these will be referred to as handovers), Rake receiving, fast power control, advanced FEC such as convolutional encoding and turbo coding, and new modulation and spreading of the physical channels. 1.3 TDD and FDD Modes:

The system is designed to operate in two modes, FDD or TDD. You will remember from earlier chapters that this pertains to whether the uplink or downlink share the same frequency and share transmit time or they are given separate frequency bands.TDD is very convenient for locations where spectrum is not available because in FDD mode, 5 MHz would be required for both the uplink and downlink for each WCDMA channel, separated by a set amount of spectrum. In TDD mode, the 5-MHz channel is divided into 10-ms frames; these frames are then divided into 15 time slots of 666 s each. These time slots can then be assigned for transmit or receive at the base station in order to have the uplink and downlink share the same frequency. This allows for a second benefit, the ability to make the data transmissions asynchronous.

CHAPTER 2 Rake Receiver

Introduction to Rake receiver:

The basic idea of A RAKE receiver was first proposed by Price and Green and patented in 1956. They proposed a method of resolving multipath using wideband pseurandom (PN) sequences modulated onto a transmitter using other modulation methods (AM or FM).The PN sequences has the property that time-shifted versions of itself are almost uncorrelated. In conventional RAKE receivers, multiple correlators are used to despread the multipath signals and then to align and combine those signals in a later stage before making a bit decision.

Conventional RAKE receiver

A conventional RAKE receiver uses multipath diversity principle. It collects the energy in the multipath signals instead of suppressing them. RAKE receiver attempts to collect the time shifted versions of the original signal by providing a separate correlation receiver known as finger for each of the multipath signals. Each finger of the RAKE receiver has a separate code generator to generate codes with different phases, where each phase corresponds to a specific multipath.

The receiver is called RAKE receiver because the block diagram looks like a garden rake.

Importance of RAKE receiver in cellular systems:

When the CDMA systems were designed for cellular systems, the inherent wide-bandwidth signals with their orthogonal Walsh functions were natural for implementing a RAKE receiver. In addition, the RAKE receiver mitigates the effects of fading and in part is responsible for the claimed 10:1 spectral efficiency improvement of CDMA over analog cellular. HIGH throughput, flexibility of design and dedicated network with integration of service, which constitutes the basic quality, reliability and service respectively, are becoming more challenging task in recent mobile communication systems. Code divisional multiple access (CDMA) is one such spread spectrum technique that offers the potential of high spectrum efficiency, together with other features such as soft capacity, multipath resistance, and inherent frequency diversity.

Functions of Rake Receiver:

RAKE receiver in WCDMA has the following main functions:

  1. Channel delay estimation or Impulse Response (IR) Measurement for multipath components.
  2. RAKE receiver finger allocation based on the channel delay estimation
  3. RAKE receiver fingers to perform the descrambling and despreading operations
  4. Channel Estimation
  5. Maximal-Ratio Combining

CHAPTER 3 Impulse Response Measurement in Multipath


One of the main advantages of CDMA systems is the capability of using signals that arrive in the receivers with different time delays. This phenomenon is called multipath. FDMA and TDMA, which are narrow band systems, cannot discriminate between the multipath arrivals, and resort to equalization to mitigate the negative effects of multipath. Due to its wide bandwidth and rake receivers, WCDMA uses the multipath signals and combines them to make an even stronger signal at the receivers. CDMA subscriber units use rake receivers. This is essentially a set of several receivers. One of the receivers (fingers) constantly searches for different multipaths and feeds the information to the other fingers. Each finger then demodulates the signal corresponding to a strong multipath. The results are then combined together to make the signal stronger.

Rake receiver collects the time shifted versions of the original signal by providing a separate correlation receiver for each of the multipath signal.Multipath components are practically uncorrelated from one another when their propagation delays exceeds one chip period.

The performance of a CDMA receiver is improved if the signal energy carried by many multipath components is utilized. This is achieved by using a RAKE receiver, where each multipath component is assigned a despreader whose reference copy of the spreading code is delayed equally to the path delay of the corresponding multipath component. The outputs of the despreaders, i.e. the fingers of the RAKE receiver, are then coherently combined to produce a symbol estimate.

CHAPTER 4 Finger Allocation

Matched filter:

Here matched filter provides the impulse response measurements of the multipath channel profile, from which multipath delays are calculated and provided to different receiver blocks.

The channel Impulse Response Measurement (IRM) is performed by using Matched Filter (MF) type of correlators that correlate the received signal with known reference code sequence such as pilot channel code. The MF resources contain shorter filters (length of 64 chips time period for RACH and 32 chips time period for DPCCH),which can be concatenated in time domain to enable the proper delay estimation also in large cells with large delay spreads (e.g. hilly terrain environments). To improve the delay estimation performance and to increase signal to noise ratio the results of MFs are further processed by coherent and non-coherent averaging. The length of the coherent IR averaging is typically one time slot while the non-coherent averaging is typically done over radio frames. The length of the averaging operations can be selected by parametrization. The accuracy of the IR measurement is chip (65,1 ns).

Finger Allocation:

The purpose of the RAKE finger allocation procedure is to define the optimal finger delay positions that maximize the receiver performance. The allocation procedure defines the correct delay positions for despreading (in RAKE fingers) the received wideband signal to symbol level information. In the case of receiver antenna diversity the finger allocation procedure combines information from separate receiver antennas. In softer handover the allocation procedure defines the optimal finger delay positions by taking into account the information from all the sectors involved in the handover situation. The finger allocation procedure contains algorithms, which eliminate the unnecessary changes in the finger time positions between successive allocations. Thus the despreading of a certain multipath component is kept on the same RAKE finger as long as possible to maximize the performance of channel estimation and maximal-ratio combining. In the finger allocation procedure also the shape of the channel impulse response is taken into account when defining the optimum finger delay positions. It has been confirmed that the allocation must be done differently for the channels where the taps are very close to each others (so called "fat finger") than for channels with clearly separate taps. Typically the allocation frequency in normal operation mode is one allocation for a code channel in every 25 ms (accuracy of chip), which is enough for all the practical situations. Code tracking with accuracy of 1/8 chip is further used in RAKE fingers to track and compensate small delay deviations in multipath component timing. The change in the timing can be caused by the movement of the UE or by the transmission timing adjustment of the UE.

CHAPTER 5 Despreading and DeScrambling

Spread Spectrum Multiple Access

As we have just seen, each user in FDMA or TDMA is allocated a fraction of the available bandwidth. In the spread spectrum multiple access scheme, on the other hand, all users can transmit simultaneously

on the entire available bandwidth using a pseudorandom code that is unique for each user. These codes, which are also known as pseudonoise (PN) codes, are random sequences generated by means of a multistage shift register, where some selected outputs are added modulo 2 and fed back to the input of the shift register. As will be shown later in this chapter, the code sequences repeat themselves after a finite, although usually quite long, period and behave as random functions for all practical purposes. The receiver separates the different users by correlating the received signal with these codes. Many spread spectrum techniques are currently available. For example, there are direct sequence spread spectrum (DSSS), frequency-hopping spread spectrum (FHSS), time-hopping spread spectrum(THSS), and hybrid techniques, which are combinations of the first three. Because UMTS W-CDMA and cdma2000 use direct sequence spread spectrum, only this access scheme will be explained in this section. In a direct-sequence spread spectrum system [6], [7], also known as the CDMA, each user is assigned a unique PN code. Its data stream is first spread out by that PN code, and then modulates the carrier frequency. The clock rate of the spreading code is known as the chip rate. This principle is illustrated in above figure. The chip rate of the PN code is usually much higher than the user data rate. The ratio of the chip rate to the data rate is called the spreading factor. In UMTS, the spreading factor varies from 4 to 256.

Direct-Spread CDMA Principles

As will be seen later, PN codes have some unique properties. One of them is that any physical channel or user application, when spread by a PN code at the transmitter, can be uniquely identified at the receiver by multiplying the received baseband signal with a phase coherent copy of that PN code. To illustrate how a CDMA receiver can detect the signal from a desired user in the presence of signals received from other users in a CDMA system, consider Figure 5.1(a), which shows the block diagram of an overly simplified CDMA receiver. Suppose that the receiver wants to detect the data stream

from user 1. The received signal from multiple users is first demodulated. The output of the demodulator, which is a base band signal, is multiplied by the PN code assigned to user 1. The resulting output is applied to the input of an integrator where it is integrated over each symbol period. The decoder reads the output of the integrator and decodes it into binary data, following certain rules. The result is the recovered data from user 1.To see that this indeed is the case, assume that the data stream from any user is represented by si(t) and its associated PN code by Ci(t). The output at the transmitter after spreading is V1(t)=si(t)*Ci(t). Notice that in si(t) or Ci(t), the signal level is either or , with representing a binary 0 and a binary 1. If the noise introduced by the channel is negligible, the demodulated signal at the baseband is given by where N is the number of users in the system. If r(t) is now multiplied by a copy of the PN code C1(t) of user 1, the resulting output is given by Because the cross-correlation between C1(t) and C2(t) is very small, the second term appears as noise so that when it is integrated over a symbol period, the output of the integrator due to this term is virtually zero. The same is true of the third and following terms. However, the output of the integrator due to the first term, when averaged over a symbol period, is s1(t) because These ideas are illustrated in the time diagrams of Figure 5.2(b) and (c).


In UMTS and cdma2000, signaling and user data is spread twice in successionfirst with the channelization codes and later with the scrambling codes. The channelization codes are orthogonal Walsh codes, which are inherently more tolerant of interference caused by multiple users. The scrambling codes, on the other hand, are not necessarily orthogonal and are built from the so-called PN codes.

Walsh Codes

Various physical channels may exist at any time on a radio interface of a 3G system. For example, at a mobile station, there may be one or more dedicated physical data channels, a dedicated physical common control channel, a physical random access channel, and a physical common packet channel. To separate these channels at the receiver, they are spread with Walsh codes at the transmitter. These codes are formed by the rows of an square matrix, whose entries are either 0 or 1. Usually, where n is an integer. They are orthogonal because if a 0 is mapped to and a 1 to , then the sum of the term-by-term products of any two rows of this matrix is In other words, if the matrix is assumed to be

Thus, for example,

Channelization codes used in UMTS W-CDMA and cdma2000 are variable-length Walsh codes, also known as orthogonal variable spreading factor (OVSF) codes. The spreading factors in UMTS may vary from 4 to 256 chips on uplink channels and from 4 to 512 chips on downlink channels. In cdma2000, Walsh codes used on traffic channels may vary from 4 to 128 chips. IS-95 uses a set of 64 fixed-length Walsh codes to spread forward physical channels. For example, Walsh code 0 is assigned to the pilot channel, code 32 to the sync channel, codes 17 to paging channels, and the rest to the forward traffic channels. In the reverse direction, they are used for orthogonal modulation where every six symbols from the block inter-leaver output are modulated as one of 64 Walsh codes.

Scrambling Codes

PN codes form the basic building blocks of scrambling codes. These codes are generated by a multibit shift register, where some selected outputs are added modulo 2 and fed back to the input. The underlying theory is well documented in the literature. See, for example, reference [14] for a thorough description of shift register sequences. Reference [15] provides some relevant mathematical background.

Theory of PN Codes To illustrate the principle of PN codes, consider Figure 5.3. The shift register array consists of four single-bit shift registers and is clocked at the chip rate. The outputs of registers3 and 0 are added in a modulo-2 adder (that is, an exclusive-OR circuit), and then applied to the input of the array. Assume that the initial states of all stages of the shift register are 1, 1, 1, and 1. Then at instant, the output of the adder is 0. So when the first clock pulse appears, the states of the 4-bit shift register change to 0, 1, 1, and 1, respectively. Thus, the states with successive clock pulses are

1 1 1 1
0 1 1 1
1 0 1 1
0 1 0 1
1 0 1 0
1 1 0 1
0 1 1 0
0 0 1 1
1 0 0 1
0 1 0 0
0 0 1 0
0 0 0 1
1 0 0 0
1 1 0 0
1 1 1 0
1 1 1 1

There is no need to write the output sequence any more because the pattern repeats after every 15 clock pulses. The output may be taken from any one of the four registers. When taken from register 0, the output sequence is 1 1 1 1 0 1 0 1 1 0 0 1 0 0 0 . . . . The same sequence, with some delays, will be generated with any other initial states as long as they are not all 0's.A few things are to be noticed here. First, the output sequence is periodic, but its bit pattern is random. Thus, it may be termed a pseudo-random pattern. In this example, the sequence repeats itself after 15 clock pulses. In other words, its period is 15 or 24 _ 1, the maximum that can be achieved with a 4-bit shift register. A sequence with the longest possible period that can be generated with a given shift register is called a maximal sequence or a maximum-length sequence. The second point to observe is that in each period, there are eight 1's and seven 0's. In other words, the numbers of 1's and 0's are equal within one clock period. Third, this particular output sequence has one run of length 4, one run of length 3, two runs of length 2, and four runs of length 1.Some shift register sequences, finite as they are, may have all of the following randomness properties.

  1. In any given period, the number of 1's is equal to the number of 0's within one bit.
  2. Each period of the output sequence contains runs of different lengths. In a random sequence, one half of these runs are of length 1, one fourth of length 2, one eighth of length 4, and so on.
  3. The auto-correlation function C(j) of a periodic sequence {bn} of period N is defined by

In other words, it is the sum of the products of the elements of the sequence and its delayed copy. Assuming that each bi is either -1 or +1, the auto-correlation function C(j) of the shift register sequence has only two values:

To see the auto-correlation property of the shift register sequence of Figure 5.3, note, first of all, that the auto-correlation function with zero phase shift is C(0) = +15. Next, if the output sequence is 1 1 1 1 0 1 0 1 1 0 0 1 0 0 0, then the sequence shifted one bit to the right is 0 1 1 1 1 0 1 0 1 1 0 0 1 0 0. Thus, the auto-correlation function C (+1) = -1. Similarly, the auto-correlation function C (-1) with one bit shifted to the left is also -1. In fact, it can be shown that C (j) = -11, j 0 as shown in Figure 5.4. This is an important property of a shift register sequence for the following reason: If each user's data is modulated with a unique PN code and transmitted to the remote end, then it can be demodulated at the receiver by correlating the received signal with a local copy of that code. If the period of this sequence x(t) is very large, each user i may be assigned the sequence xi(t) = x(t +Ti), which is actually the same code but with an offset Ti. In this case, in order to be able to decode each user sequence as unambiguously as possible in the presence of noise, it is necessary that for j0, the value of C(j) be as small as possible compared to N.Similarly, the cross-correlation between two sequence xn and yn is defined in the following way:

Given a shift register of length, say, k, not all sequences will have the maximum period of length 2k _ 1. Consider again the shift register of figure

However, this time, outputs of registers 2 and 0 are added modulo 2, and then applied to the input of the first stage as shown Figure. Assuming that the initial states of the register array are 1111, the subsequent states are

1 1 1 1
0 1 1 1
0 0 1 1
1 0 0 1
1 1 0 0
1 1 1 0
1 1 1 1

So in this case, the period n =6.Because the period p depends on the way the feedback signal is constructed, it is necessary to see which registers should be connected to the adder so that the sequence has the previous desirable properties. To this end, assume that the binary sequence {bn} = b0, b1,b2, b3, . . . defines the states of register 3 as a function of time. Clearly, the same sequence delayed by one bit also gives the states of register 2, and so on. Thus, the sequence can be used to define the state of the whole shift register array. Using this sequence as coefficients, a polynomial f(x) can be defined:

In this case, we can determine a function such that where the division by g(x) is modulo 2. In other words, the shift register sequence f(x) can be generated by constructing a feedback signal by adding (modulo 2) the outputs of the k shift registers according to the function g(x). Function g(x) is called the characteristic polynomial for sequence bn. Clearly, in the previous definition of g(x), gi may be 0 for some values of i.We will now provide some important results without giving any proof:

  1. Given a function g(x) as defined previously, the period of a k-stage shift register sequence is the smallest positive integer n for which g(x) divides polynomial 1 + xexpn. The maximum possible period of a k-stage shift register is 2k -1.
  2. If the feedback in a k-stage shift register is such that the output sequence has indeed this maximum period 2k -1, then the function g(x) that generates this sequence is irreducible. In other words, g(x) is divisible only by itself (and of course by 1).Notice that this statement merely gives a necessary condition for the sequence to be maximal length. This condition is not sufficient because there are polynomials that are irreducible, but do not yield a maximal length sequence. For example, consider g(x)=1+x+xexp2 + xexp3 + xexp4. This polynomial is irreducible, but because it divides 1 _ x5, the period of the sequence is 5 and not 2exp4 -1.
  3. The period of a k-stage shift register sequence is a factor of 2k -1. Thus, if 2k -1. Is a prime, then every irreducible polynomial of degree k will generate a maximal length sequence. To illustrate these ideas with examples, consider the shift register of Figure. Here, the shift register consists of four stages. Thus k = 4. Because the outputs from registers 3 and 0 are being added and then applied to the input, the characteristic polynomial is

The period is 15, the maximum achievable with four registers. It can be shown that the smallest positive integer n for which this g(x) divides 1+ xexpn is indeed 15, thus satisfying statement 1 previously. Furthermore, as required by statement 2, it can be easily shown that g(x) is in fact irreducible because it cannot be factored into polynomials of lower degrees. But because 15 is not a prime, following statement 3, not all irreducible polynomials of degree 4 will generate a maximal length sequence. Referring to Figure, the characteristic polynomial for this shift register sequence is

It can be shown that this polynomial divides 1+ xexp6. Thus, the period of this sequence is n=6, as shown previously. Also, this characteristic polynomial is not irreducible because 1 +xexp2 +xexp4 =(1 +x+ xexp2)exp2, and hence, in accordance with statement 2 previously, cannot generate a maximal length sequence.

Scrambling Codes in UMTS:

The scrambling codes in UMTS are complex valued and may be either long or short. A long code has a length of 38,400 chips (that is, 10 ms) and a short code only 256 chips. As an example, Figure 5.5 shows how a long code for a UMTS uplink channel is generated. This code is constructed with two PN codes, whose characteristic polynomials are g1(x) = xexp25 + xexp3 + 1 and g2(x) = xexp25 + xexp3 +xexp2+ x + 1. They are implemented as sequences PN1 and PN2 using two 25-bit shift registers. PN1 and PN2 are added modulo 2, and the output is mapped to a real-valued function, say, I. Another function Q is derived by simply delaying I by 2exp24 +16 chips. Q is multiplied by j, where the sign changes every chip period, and then added to I to yield the long code.


The most important purpose of the spreading codes is to help preserve orthogonality among different physical channels of the uplink user. Walsh-Hadamard codes, also known as OVSF (Orthogonal Variable Spreading Factor) codes, are employed as uplink spreading codes. OVSF codes can be explained using the code tree shown in Fig 5.6 The subscript here gives the spreading factor and the argument within the parenthesis provides code number for that particular spreading factor.

Each level in the code tree defines spreading codes of length SF, corresponding to a particular spreading factor of SF. The spreading factor in uplink is defined as

The parameter k determines the number of bits in each slot .The spreading factor may thus range from 256 down to 4.

The number of codes for a particular spreading factor is equal to the spreading factor itself. All the codes of the same level constitute a set and they are orthogonal to each other. When the signal is transmitted by the transmitter in the free air after spreading the bandwidth, interference, channel noise etc are added with the transmitted signal. To separate the original signal at the receiver end we multiply the same PN Code (multiplied at the receiver end) twice to the received signal and then passed through a filter to remove the noise.

To apply an SS technique, simply inject the corresponding SS code somewhere in the transmitting chain before the antenna. (That injection is called the spreading operation.) The effect is to diffuse the information in a larger Bandwidth. Conversely, you can remove the SS code (dispreading operation) at a point in the receive chain before data retrieval. The effect of a despreading operation is to reconstitute the information in its original bandwidth. Obviously, the same code must be known in advance at both ends of the transmission channel. (In some circumstances, it should be known only by those two parties.)

Bandwidth Effects of the Spreading Operation:

The simple drawings below illustrate the evaluation of signal bandwidths in a communication link.

SS modulation is applied on top of a conventional modulation such as BPSK or direct conversion. One can demonstrate that all other signals not receiving the SS code will stay as they are, unspread.

Bandwidth Effects of the Despreading Operation:

An SS demodulation has been made on top of the normal demodulation operations above. One can also demonstrate that signals added during the transmission (such as an interferer or jammer) will be spread during the despreading operation.


In WCDMA, a PN sequence is used to scramble the signal in addition to the spreading signal. The purpose of the scrambling signal is to separate terminals or base station from each other. Scrambling is used on top of spreading. It does not change the signal bandwidth but only makes the signals from different source separable from each other. With the scrambling, it would not matter if the actual spreading were done with identical code for several transmitters. The relationship between spreading and scrambling is shown in Fig

In the uplink communication of WCDMA, Gold codes are selected to scramble the signals for different mobile terminal. Uplink scrambling codes help to maintain separation among different mobile stations. Each mobile terminal has a distinct Gold code. Transmitter and receiver should know the scrambling code for each terminal in advance.

CHAPTER 6 Channel Estimation

Types Channel Estimation:

The rake receiver combines these different paths into a composite signal with substantially better characteristics for demodulation than a single path. To combine the different paths meaningfully, the rake receiver needs such channel parameters as the number of paths, their location (in the delay domain), and their (complex-valued) attenuation. In a WCDMA system, the necessary channel parameters must be estimated and tracked throughout the transmission. Typically channel estimation is performed using one of three methods or a combination of these methods. The trade-offs between complexity and performance dictates the choice of algorithm

Data Aided Channel Estimation:

Known pilot symbols are transmitted. At the receiver end, the channel estimation algorithm operates on the received signal along with its stored symbols to generate an estimate of the transmission channel.

Decision-Directed Channel Estimation:

A rough estimate of the channel is obtained using a suitable estimation method. Then this estimate is used to make symbol decisions. The channel estimate is further improved using the resulting symbols as "pilot symbols." This type of estimation contains some inherent delay because the symbol decisions occur before the final channel estimate can be made. Also, there may be error propagations because any errors in the symbol decisions affect the final estimate.

Blind Channel Estimation

This estimation process relies not on pilot symbols or symbol decisions but rather on certain characteristics of the modulated signal. For example, the constant modulo algorithm (CMA) uses the amplitude of the signal as the criterion for estimating the channel. In constant energy modulation schemes such as Quadrature Phase Shift Keying (QPSK), the knowledge that all signals are transmitted with equal energy is used as the basis for obtaining the channel estimate. This type of algorithm typically requires a longer convergence time and usually has a higher mean square error (MSE) than the other two schemes. In a typical WCDMA rake receiver, channel estimates are used to combine the multipaths. Figure 6.1 shows how different paths are combined using Maximum Ratio Combining (MRC). In Figure 6.1, r(t) is the received signal, which is split into r(t-?i). Note that, g(t,?i) is the corresponding channel estimate for each path is r(t-?i). The objective is to estimate the channel phase and amplitude (denoted in Figure 6.1 as g(t,?i)) for each of the identified paths. This information is then used for combining each path of the received signal.

As Figure 6.1 show, the following steps occur in a WCDMA receiver (excluding the error correction coding):

  1. Descrambling. Received signals are multiplied by the scrambling code and delayed versions of the scrambling code. A path searcher determines the delays prior to descrambling. Each delay corresponds to a separate multipath that is to be combined by the rake receiver.
  2. Despreading. The descrambled data of each path is despread by simply multiplying the descrambled data by the spreading code.
  3. Integration and dump. The despread data is integrated over one symbol period, giving one complex sample output per QPSK symbol. This process is carried out for all paths to be combined by the rake receiver.
  4. Symbol combining. The same symbols obtained via different paths are then combined using the corresponding channel information and a combining scheme such as MRC.
  5. The combined output is sent to a simple decision device to decide on the transmitted bits. The objective of the channel estimation block is to estimate the channel phase and amplitude (denoted in Figure 6.1 as g(t,ti) ) for each of the identified paths. Once this information is known, it can be used for combining each path of the received signal.

The objective of the channel estimation block is to estimate the channel phase and amplitude (denoted in Figure 6.1 as g(t,?i)) for each of the identified paths. Once this information is known, it can be used for combining each path of the received signal.

Our goal is to estimate the characteristics of the time-variant channel. In WCDMA this is done by using Pilot Symbol Aided and adaptive filtering. Channel estimation is used to remove distortion caused by radio channel and it is based on the known pilot symbols on DPCCH. Filter adapts to the Doppler power spectrum i.e. both frequency and the shape of the spectrum. The estimation is done for each finger separately. The use of adaptive filter ensures good performance in all kind of propagation conditions. In the case of multiple receiver antennas the performance of channel estimation is further improved by combining the power spectrum information available from different receiver antennas. The combining process is based on maximal-ratio combining, which decreases the effect of additive noise, which can further be decreased by channel decoding. In WCDMA, channel estimation can be performed using a Common Pilot Channel (CPICH) or the time-multiplexed pilot bits in a "dedicated traffic channel." Here we describes both of these processes and considers their advantages/disadvantages. In the downlink traffic channel (DPDCH/DPCCH), pilot symbols (2 to 8 symbols) and control symbols are transmitted in every slot. There are 15 slots per WCDMA frame. Each frame is 10 ms long and has 38400 chips (3.84 Mcps). Channel estimates can be made using these pilot symbols. If these time-multiplexed pilot data bits are used, then the estimate for the data bits in between two consecutive sets of pilot bits (two slots) can be obtained by interpolation. A decision-directed approach can be used to improve performance. Figure 6.2 diagrams the process of channel estimation using time-multiplexed symbols.

Also, in the downlink of the WCDMA system, a Common Control Channel (CPICH) is transmitted with a higher power than the dedicated traffic channels. All mobiles in the cell receive this channel. The CPICH is transmitted with a constant spreading factor (SF) of 256 and a spreading code of all ones. Thus, there are 10 symbols per slot and 150 symbols per frame of a CPICH. All the symbols of the CPICH are 1-j. On the receiver end, the symbols work as pilot symbols and can be used for channel estimation.

The advantage of using CPICH for channel estimation is that all the data in the frame can be used as opposed to only a few symbols in the DPCCH/DPDCH. Also, since this data is transmitted with a higher power than in the traffic channel, reception at the handset is improved. One situation in which the time-multiplexed pilot bits become useful is when the mobile is at the cell edge. This is because the dedicated channels are power controlled, whereas the CPICH is not power controlled. Figure 5.2 shows a typical downlink WCDMA DPDCH/DPCCH frame.

For each independent path, the channel estimate is obtained as follows:

  1. Remove the modulation form CPICH by simply multiplying the CPICH data by its conjugate. In this case, it is simply 1+j because all the CPICH symbols are 1+j. The resulting channel estimate is noisy because of AWGN and multiple-user interference.
  2. Pass the noisy channel estimate through LMS.
  3. Decimate or interpolate the filtered channel estimate obtained in step 2 to match the data rate of the CPICH to the data rate of the DPCCH/DPDCH. The process of interpolation is performed by a simple zero-order hold (that is, a simple repeater). This works well in most cases because the channel is assumed to be stable for the symbol duration.

CHAPTER 7 Maximal Ratio Combining

Maximum Ratio Combining for a WCDMA Rake Receiver:

Wideband CDMA (WCDMA), a widely accepted third-generation interface, is based on direct-sequence (DS) CDMA technology. To minimize distortion of the signals in a DS-CDMA system, a rake receiver is used. A signal transmitted through the wireless channel may be severely distorted due to co-channel interference, adjacent channel interference (or multiple access interference), thermal noise, and multipath fading. The most severe distortion comes from fading, which changes the bit error rate (BER) curve from an exponential to a linear curve. One technique the rake receiver employs to combat these distortions is diversity. There are different diversity techniques, including frequency, time, and space. This application note explains how the DS-CDMA rake receiver employs multipath diversity techniques to minimize distortion.

Combining Schemes

In the diversity scenario are different combining schemes, including the following:

Selection diversity (SD):

When L multipath signals are received, the combining scheme selects the paths with the higher signal-to-noise ratio (SNR) and discards the L-1 paths. Selection diversity is used in reverse-link of IS-95, where multiple base stations receive the signal from the mobile, but only the base station that receives the strongest signal is chosen to serve the mobile.

Equal gain combining (EGC):

The receiver corrects the phase rotation of the received signals caused by the fading channel and combines the received signals of different paths with equal weight.

Maximum ratio combining (MRC):

The optimum way (in the sense of the least BER) to use information from different paths to achieve decoding in an additive white Gaussian channel (AWGN), The receiver corrects the phase rotation caused by a fading channel and then combines the received signals of different paths proportionally to the strength of each path. Since each path undergoes different attenuations, combining them with different weights yield an optimum solution under an AWGN channel. To achieve better diversity, the signals from different paths should be totally uncorrelated. If the signals from different paths are correlated, performance deteriorates because duplicate information is extracted from each path. In a real environment, the signals from different paths are almost correlated.

Ways of Combining

There are two ways to achieve combining.

  1. Symbol level combining.
  2. Chip level combining

Symbol level combining:

Symbol-level combining, as its name implies, combines the signals of different paths at the symbol level. The descrambling and despreading are performed before combining in order to convert chip-level signals into symbol-level signals. Figure 7.1 shows the block diagram for combining at the symbol level.

Chip-level combining:

Chip-level combining performs combining followed by descrambling and despreading. Figure 7.2 shows the block diagram for combining at the chip level

The performance of both combining schemes is the same under perfect channel estimation and path searching, assuming that the fading channel is constant over a symbol period. Table 7.1 shows the estimation of the computational loads of both combining schemes for one channel. The descrambling and despreading are combined and occur in one step. It is assumed that the scrambling code and the spreading code do not change during the transmission.

In the rake receiver, symbol-rate combining rather than chip-rate combining is used in the simulation because symbol-rate combining requires fewer computations, especially when the spreading factor goes up. Channel estimation is hard to achieve in chip-level combining.

In the rake receiver, symbol-rate combining rather than chip-rate combining is used in the simulation because symbol-rate combining requires fewer computations, especially when the spreading factor goes up. Channel estimation is hard to achieve in chip-level combining. Usually, it is estimated at the symbol level and interpolated to the chip level. Therefore, symbol-level channel estimation requires fewer cycles than chip-level estimation since the interpolation (even the simplest repetition or first-order hold) requires some cycles. The symbol and chip rate combining are compared in term of memory usage. In chip rate combining, one path is stored, but the stored data is in chip-rate, which is as many as 38400 samples. In symbol rate combining, the data of multiple

paths is stored. However, each path is in symbol level. The total number of samples is (38400/SF) * L. The discrepancy of memory usage between chip and symbol combining as follows:

Where L is the number of fingers (ranging from 2 to 8) and SF is the spreading factor (ranging from 4 to 512). If Equation is positive, chip-rate combining requires more memory than symbol-rate combining. Since SF is larger than L most of the time, and Equation is positive, symbol-rate combining uses less memory than chip-rate combining.

Please be aware that the free essay that you were just reading was not written by us. This essay, and all of the others available to view on the website, were provided to us by students in exchange for services that we offer. This relationship helps our students to get an even better deal while also contributing to the biggest free essay resource in the UK!