Multipath signal Detection in CDMA System

Transcription

1 Chapter 4 Multipath signal Detection in CDMA System Chapter 3 presented the implementation of CDMA test bed for wireless communication link. This test bed simulates a Code Division Multiple Access (CDMA) link. This can be used to study the performance of a CDMA wireless link with variation in system parameters and channel conditions [47]. This chapter discusses the effects of multipath signal, Basic CDMA signal transmission and reception concepts, proposed improved multipath signal detection method, experimental setup for performance evaluation of signal detection in five paths over AWGN and Rayleigh channel, statistically ordered algorithm, finally with results and conclusion. 4.1 Introduction Signal detection in multipath CDMA systems is particularly attractive when the desired signal is predicted and some interference is known, and multipath signals are not completely orthogonal at the receiver. To get orthogonality of multipath signal, Orthogonal Frequency Division Multiplexing (OFDM) is used. OFDM can be seen as either a modulation technique or a multiplexing technique. One of the main reasons to use OFDM is to increase the robustness against frequency selective fading or narrowband interference. As many multipath systems must deal with the near-far problem, a weakness common to several known blind multipath CDMA detectors is that the detection performance is very sensitive to the expected signal which occurs as a result of channel distortion [48-50]. We consider the design of a partially multipath CDMA detector that is robust to Signal to Interference Noise Ratio. Computer simulations indicate that the performance of the proposed robust multipath detector is superior to the one that does not utilize the interference information. With CDMA systems, all users transmit in the same frequency band using specialized codes as a basis of channelization. The transmitted information is spread in bandwidth by multiplying it by a wide bandwidth pseudo random sequence. Both the base station and the mobile station know these random codes that are used to modulate the data sent, allowing it to de-scramble the received signal. Page 59

2 4.2 Multipath Effects In a radio link, the RF signal from the transmitter may be reflected from objects such as hills, buildings, or vehicles. This gives rise to multiple transmission paths in the channels and multi signals are received at the receiver. Figure 4.1 shows some of the possible ways by which multipath signals are formed. As the name itself indicates, more than one signal is received by the receiver at different intervals of time [51-54]. Figure 4.1 Multipath Signals received by the receiver Rayleigh fading Multipath signal transmission gives rise to fading of the signals. The mobile antenna receives a large number of reflected and scattered waves. Because of wave cancellation effects, the instantaneous received power seen by a moving antenna becomes a random variable, dependent on the location of the antenna. The relative phase of multiple reflected signals can cause constructive or destructive interference at the receiver. This is experienced over very short distances (typically at half wavelength distances). Thus is given the term fast fading. These variations can vary from 10-30dB, over a short distance. Figure 4.2 shows the level of attenuation that can occur due to the fading [55]. Page 60

3 Figure 4.2 Level of attenuation of the signal due to Typical Rayleigh fading. The Rayleigh distribution is commonly used to describe the statistical time varying nature of the received signal power. It describes the probability of the signal level being received by the mobile receiver due to the effect of fading. Table 4.1 shows the probability of the signal level for the Rayleigh distribution. Table 4.1 Cumulative distribution for Rayleigh distribution Frequency Selective Fading In any radio transmission, the channel spectral response is not flat. It has dips or fades in the response due to reflections causing cancellation of certain frequencies at the receiver. Reflections of near-by objects (e.g. ground, buildings, trees, etc) can lead to multipath signals of similar signal power as the direct signal. This can result in deep nulls in the received signal power due to destructive interference [56-58]. For narrow bandwidth transmissions if the null in the frequency response occurs at the transmission frequency then the entire signal can be lost. This can be partly overcome in two ways. By transmitting a wide bandwidth signal or spread spectrum as CDMA, any dips in the spectrum only results in a small loss of signal power, rather than a complete loss. Another Page 61

4 method is to split up the transmission into many small bandwidth carriers, as is done in a COFDM/OFDM transmission [59-60]. The original signal is spread over a wide bandwidth and so nulls in the spectrum are likely to affect only a small number of carriers rather than the entire signal. The information in the lost carriers can be recovered by using forward error correction techniques Delay Spread The received radio signal from a transmitter consists of typically a direct signal, plus reflections of objects such as buildings, nearby hills, and other structures. The reflected signals arrive at a later time than the direct signal because of the extra path length. This gives rise to a slightly different arrival times of reflected signals, spreading the received energy in time. Delay spread is the time spread between the arrival of the first and last significant multipath signal seen by the receiver. In a digital system, the delay spread can lead to inter-symbol interference. This is due to the delayed multipath signal overlapping with the following symbols [61-62]. This can cause significant errors in high bit rate systems, especially when using time division multiplexing (TDMA). As the transmitted bit rate is increased, the amount of inter-symbol interference also increases. The effect starts to become very significant when the delay spread is greater than ~50% of the bit time. One of the causes of inter symbol interference is what is known as multipath propagation in which a wireless signal from a transmitter reaches the receiver via many different paths. The causes of this include reflection (for instance, the signal may bounce off buildings), refraction (such as through the foliage of a tree) and atmospheric effects such as atmospheric ducting and ionosphere reflection. Since the various paths are of different lengths, this results in the different versions of the signal arriving at the receiver at different times [63-65]. These delays mean that part or all of a given symbol will be spread into the subsequent symbols. The Multipath propagation causes inter symbol interference when a wireless signal being transmitted reaches a receiver through different paths. These paths will travel different lengths before reaching the receiver, thus creating different versions that reach at different time intervals. The delay in symbol transmission will interfere with correct symbol detection. The amplitude and/or phase of the signal can be distorted when the different Page 62

5 paths are received for additional interference. There by interfering with the correct detection of those symbols. [66]. Figure 4.3 Effect of ISI due to Delay spread on the signal. Figure 4.3 shows the effect of inter-symbol interference due to delay spread on the received signal. As the transmitted bit rate is increased the amount of inter symbol interference also increases. The effect starts to become very significant when the delay spread is greater than ~50% of the bit time. Table 4.2 shows the typical delay spread under various environments. The maximum delay spread in an outdoor environment is approximately 20 s, thus significant inter symbol interference can occur at bit rates as low as 25 kbps. Inter-symbol interference can be minimized in several ways. One method is to reduce the symbol rate by reducing the data rate for each channel (i.e. split the bandwidth into more channels using frequency division multiplexing, or OFDM). Another is to use a coding scheme that is tolerant to inter-symbol interference such as CDMA. Table 4.2 Typical Delay Spread Environment or cause Delay Spread Maximum Path Length Difference Indoor (room) 40ns 200 ns 12 m 60m Outdoor 1 s 20 s 300 m 6 km Page 63

6 4.2.4 Doppler Shift When a signal source and a receiver are moving relative to one another the frequency of the received signal will not be the same as the source. When they are moving toward each other the frequency of the received signal is higher than the source, and when they are approaching each other the frequency decreases. This is due to the Doppler Effect. An effect becomes important when developing mobile radio systems [67-70]. The amount of the frequency change due to the Doppler Effect depends on the relative motion between the source and receiver and on the speed of propagation of the wave. The Doppler shift in frequency can be written: (4.1) Where f is the change in frequency of the source seen at the receiver, fo is the frequency of the source, v is the speed difference between the source and transmitter, and c is the speed of light. 4.3 Basic CDMA transmission and reception concepts CDMA is achieved by modulating the data signal by a pseudo random noise sequence (PN code), which has a chip rate higher than the bit rate of the data. The PN code sequence is a sequence of ones and zeros (called chips), which alternate in a random fashion. Modulating the data with this PN sequence generates the CDMA signal. The modulation is performed by multiplying the data (XOR operator for binary signals) with the PN sequence [71] CDMA generation CDMA signal transmission process includes spreading, modulation, and dispreading. The principle behind spreading and de-spreading is that when a symbol is XOR-ed with a known pattern, and the result is again XOR-ed with the same pattern, the original symbol is recovered. In other words, the effect of an XOR operation if performed twice using the same code is null. In orthogonal spreading, each encoded symbol is XOR-ed with all 64 Page 64

7 chips of the Walsh code. Fig 4.4 gives the details of transmitted data with orthogonal sequence[72]. Figure 4.4 Transmitted Data in a CDMA system Information recovery The receiver de-spreads the chips by using the same Walsh code used in the transmitter. Notice that under no-noise conditions, the symbols or digits are completely recovered without any error. In reality, the channel is not noise-free, but CDMA system employs Forward Error Correction techniques to combat the effects of noise and enhance the performance of the system. When the wrong Walsh sequence is used for de-spreading, the resulting correlation yields an average of zero [73-75]. This is a clear demonstration of the advantage of the orthogonality property of the Walsh codes. Whether the wrong code is mistakenly used by the target user or other users attempting to decode the received signal, the resulting correlation is always zero because of the orthogonality property of Walsh sequences. Fig 4.5 gives the details of received data with correlation function. Page 65

8 Figure 4.5 Received Data in a CDMA system Spreading and De-spreading The improvement of time-domain information rate means that the bandwidth of spectrum-domain information is spread with reference to the signal transmission rate. Figure 4.6 Spectrum before and after transmission An Example of Spreading with Three Subscribers: In this example, three users A, B, and C are assigned three orthogonal codes for spreading purposes: Page 66

9 _ User A signal = 00, Spreading code = 0101 _ User B signal = 10, Spreading code = 0011 _ User C signal = 11, Spreading code = 0000 The analog signal shown on the bottom of the Figure 4.7 is the composite signal when all of the spread symbols are summed together [76-78]. Figure 4.7 composite signal De-spreading At the receiver of user A, the composite analog signal is multiplied by the Walsh code corresponding to user A and the result is then averaged over the symbol time. This process is called correlation. Note that the average voltage value over one symbol time is equal to 1. Therefore, the origi may try to decode the symbols for users B or C in the same manner. This process occurs in the CDMA mobile unit for recovering the signals [79]. Figure 4.8 is about composite signal with product in transmitter. Page 67

10 Figure 4.8 composite signals with product In an actual application, the system implements the modulation in this way: I and Q channel sequences represent two channels of cyclic PN short code sequences [80]. The cyclic period of each channel of PN short codes is 215. For different sectors, there are different starting locations of I and Q sequence cycles (that is to say, different sectors have different time offsets). In this way, different PN short code sequences can be obtained and a mobile station can recognize the information from different sectors. Figure4.9 CDMA Transmission Page 68

11 Figure 4.9 shows the process of a CDMA transmission. The data to be transmitted (a) is spread before transmission by modulating the data using a PN code. This broadens the spectrum as shown in (b). In this example the process gain is 125 as the spread spectrum bandwidth is 125 times greater the data bandwidth. Part (c) shows the received signal. This consists of the required signal, plus background noise, and any interference from other CDMA users or radio sources. The received signal is recovered by multiplying the signal by the original spreading code. This process causes the wanted received signal to be dispread back to the original transmitted data. However, all other signals that are uncorrelated to the PN spreading code become more spread. The wanted signal in (d) is then filtered removing the wide spread interference and noise signals CDMA generation and Forward Link Encoding CDMA is achieved by modulating the data signal by a pseudo random noise sequence (PN code), which has a chip rate higher than the bit rate of the data. The PN code sequence is a sequence of ones and zeros (called chips), which alternate in a random fashion. Modulating the data with this PN sequence generates the CDMA signal. The CDMA signal is generated by modulating the data by the PN sequence. The modulation is performed by multiplying the data (XOR operator for binary signals) with the PN sequence. Figure 4.10 shows a basic CDMA transmitter. The PN code used to spread the data can be of two main types. A short PN code (typically chips in length) can be used to modulate each data bit. The short PN code is then repeated for every data bit allowing for quick and simple synchronization of the receiver. Figure 4.11 shows the generation of a CDMA signal using a 10-chip length short code. Alternatively a long PN code can be used. Long codes are generally thousands to millions of chips in length, thus are only repeated in frequently. Because of this they are useful for added security as they are more difficult to decode. The channel code decoder in a CDMA mobile performs channel state estimation to determine the reliability of each received encoded symbol. This reliability information is typically an estimate of received signal envelope and interference plus noise variance. In order to define the directions of the transmission within a cellular system, the links to and from the base station are defined [81]. The names used within IS-95 and CDMA2000 are different to those used for GSM and UMTS. Page 69

12 Figure 4.10 Simple direct sequence modulator Figure 4.11 Direct sequence signals The forward link, from the base station to the mobile, of a CDMA system can use special orthogonal PN codes, called Walsh codes, for separating the multiple users on the same channel. These are based on a Walsh matrix, which is a square matrix with binary elements and dimensions that are a power of two. It is generated from the basis that Walsh (1) = W1 = 0 and that Where Wn is the Walsh matrix of dimension n. For example: Page 70

13 Walsh codes are orthogonal, which means that the dot product of any two rows is zero. This is due to the fact that for any two rows exactly half the number of bits match and half do not. Each row of a Walsh matrix can be used as the PN code of a user in a CDMA system. By doing this the signals from each user is orthogonal to every other user, resulting in no interference between the signals. However, in order for Walsh codes to work the transmitted chips from all users must be synchronized. If the Walsh code used by one user is shifted in time by more than about 1/10 of chip period, with respect to all the other Walsh codes, it loses its orthogonal nature resulting in inter-user interference. This is not a problem for the forward link as signals for all the users originate from the base station, ensuring that all the signal remain synchronized [82] CDMA Reverse Link Encoding The reverse link is different to the forward link because the signals from each user do not originate from a same source as in the forward link. The transmission from each user will arrive at a different time, due to propagation delay, and synchronization errors. Due to the unavoidable timing errors between the users, there is little point in using Walsh codes as they will no longer be orthogonal. For this reason, simple pseudo random sequences are typically used. These sequences are chosen to have a low cross correlation to minimize interference between users. The capacity is different for the forward and the reverse links because of the differences in modulation. The reverse link is not orthogonal, resulting in significant inter-user interference. For this reason the reverse channel sets the capacity of the system. 4.4 Multipath Detection in CDMA Systems In conventional CDMA receivers, the detection of multipath components and RAKE finger management is normally based on the received signal energy per path. These energy-based schemes essentially overlook the interference component contaminating the Page 71

14 total received power. Consequently, they exhibit poor detection capability especially at low signal-to-interference-plus-noise ratio (SINR). Therefore a better scheme for multipath detection that takes into consideration the interference level in each resolved path individually has been implemented. This scheme estimates and cancels the interference per path before detection [83-84] Energy-Based Multipath Detection Method (EMDM) To improve the probability of multipath detection, N A independent search results are obtained through repeating the search process at different time instants, eg., bits or search blocks, during data frame. In the conventional scheme, the correlation energy is averaged over NA independent search blocks at each delay offset and the results are compared to a threshold. If the average energy at a certain delay offset exceeds the threshold, the path with that delay offset is acquired. This process is done for all delay offsets in the search window. If a wrong delay, which does not contain the desired channel signal, passes the threshold test, the tracking loop will detect it and declare a false alarm state after a relatively long period of processing time. The average correlation energy at the k th delay offset is (4.2) 2 I 2 I (k) and the channel power p(k) obtained from N A observations. In Eqn.(4.2), it is assumed that the interference and the desired channel signal are independent. This is a valid assumption since each multipath component fades independently [85] Improved Multipath Detection Method (IMDM) A better detection scheme should consider the received interference power at the searched delay offsets. For this purpose, the interference power should be estimated at each delay offset in the search window. To obtain an estimate of the interference power, we use the N A search results used by the EMDS scheme as shown: Page 72

15 (4.3) With the assumption that the channel fading coefficients are known at the receiver, this estimator can be shown to be Minimum Variance Unbiased (MVU) estimator1 [86]. The new detection metric is: (4.4 ) Consequently, under the IMDM, the paths with maximum Z will be acquired. Comparing the decision metric given in Eqn.4.4 and the one in Eqn.4.2, we observe that the proposed metric depends on the desired signal power as opposed to the conventional metric which uses the signal and interference power in the detection process. The parameters pertaining to the CDMA system used for simulation are as follows: The system consists of a total of five asynchronous user communicating with a single base station. Each user is assigned a PN code of length N c {32, 128, 256}. The traffic channel carries binary phase shift keying (BPSK) modulated data bits spread by a Walsh code code-multiplexed with the pilot channel. A rectangular pulse is used to shape the transmitted chips. The pilot channel gain G P is set to - db relative to the traffic channel, i.e. G P = 0.5E b. A frequency selective Rayleigh fading channel is considered here with normalized Doppler rate f d {10-2,10-3,10-4 }.The number of paths L is set to four at relative delays dc = {2, 4, 6, } chips with uniform power delay profile. The search step size S is set to half a chip. Table 4.4 shows the results obtained by simulation for EMDM and IMDM methods [87]. For zero thresholds, probability of detection in EMDM and in IMDM is u(t) Searcher - 2 N A f(n) FIR Filter - + h k (n) 1-2 NA 1 Y(k) + - Z(k) Comparator Receive r (MRC) Figure 4.12 Block diagram of IMDM Page 73

16 It can be seen that the in eqn 4.3 uses the output of the FIR filter as rough channel estimates in order to estimate the interference variance and the signal power.in fig we present a schematic diagram of proposed IMDM structure where the simplified estimator is incorporated into the detection method. While the upper branch is used to compute the decision metric of the EMDM, the lower branch is used to estimate the interference variance. (a) (b) Fig Probability of detection for (a) Energy Based Multipath Detection Method (b) Improvised Multipath Detection Method Threshold Fig 4.14 Probability of detection for the first path using EMDM Page 74

17 Fig 4.15 Probability of detection for the first path using IMDM Table: 4.3 Effect of threshold setting on the Probability of detecting the first path for both methods SNR SIR= 0 db, SNR= 10 db Method Conventional EMDM Proposed(IMDM) No Threshold 91.05% 95.40% 99.30% f d = % 90.56% 91.85% Table 4.4: Results obtained by simulation for EMDM and IMDM methods EMDM IMDM Threshold Probability of Detection Threshold Probability of Detection Page 75

18 It is apparent that, for high probability of detection, the approximate results agree very well with the simulation results. To illustrate the detection performance improvement provided by the IMDM when the FIR filter length K f is varied with the Doppler rate, the receiver operational characteristics (ROC) curves, i.e., the probability of detecting the first path as a function of the probability of false alarm, are produced at different Doppler rates for both detection schemes. The filter length K f is increased when the Doppler rate decreases. The combinations {(fd,kf) : (10 2, 3), (10 3, 6), ( 10 4, 9)} are used here. Fig 4.21 and 4.23 shows the ROC curves for both detection schemes at Eb/N0 = 8 db, SIR = 10 db when Nc = 128 and NA = 24. It is seen that, for the same Doppler rates, the IMDSM results in higher detection probability than that of the EMDS. In EMDM, as the threshold value increases the probability detection of signal is accurate and in IMDM, as the threshold value increases the probability detection of signal is more accurate [88] The results show that the proposed scheme provides significant improvements in the detection probability of multipath signal over the energy-based and signal detection methods. 4.5 Experimental set up of modulator and demodulator Figure 4.16 shows the performance evaluation set up to find the highest strength with low BER using Rayleigh fading channel. It was assumed that the transmitted signal takes five different paths to reach the receiver. At the receiver side, the signal having highest received power was compared with threshold value and reordered in descending order using statistical algorithm. The signal having highest receiving power was selected for BER calculation. Simulation environment shows that using Simulink, Matlab tool five Rayleigh channels are developed and CDMA signal is passed and observed the fading signal and BER is calculated for the signal and using the detection algorithm, strength of the signal is known and that signal is consider Page 76

19 for further observations. This is done for both AWGN and Rayleigh channel [89-90].Using MRC diversity in Flat fading Rayleigh channel, incorporating statistical multipath signal detection algorithm, we are getting the signal at receiver. The received power is more for the obtained signal compared in AWGN channel due additive property of noise in the channel. Figure: 4.16 CDMA Signal over Rayleigh channel Page 77

20 . Figure 4.17.Statistically ordered signals in five paths in MRC. Figure 4.16 describes the experimental setup to analyze the transmission of CDMA signal over Rayleigh channel model using five simultaneous paths using Matlab Simulink tool. Each channel has its parameters with different SNR levels. Figure 4.17 is about signals which have been statistically ordered in Maximal Ratio Combiner (MRC) diversity scenario to get one signal with highest received power with low BER. The Probability density function (pdf) of the available average channel capacity per user (in the Shannon sense) for a hybrid DS/FFH-CDMA cellular system when operating in a Rayleigh fading environment is analytically examined. In addition, the probability that the channel capacity per user does not exceed the available average channel capacity per user is derived. The final expressions can be very useful for the quantitative analysis of a DS CDMA cellular system, when operating in a Rayleigh fading environment [91-93] Statistical signal detection algorithm: The algorithm is explained through the following steps: Step 1: start step2: generate the data step3: setting the signal parameters in five channel step4: transmit the data through Rayleigh channel step5: add the AWGN ( noise) and considered Flat fading in Rayleigh channel step6: calculate the BER for both channel(after assigning weighting element in MRC) Page 78

21 step7: compare the BER values maxval = min(y) and BERindx = (Y == maxval) step8: find the step9: weighted BER values in descending order step10: repeat the steps 6 to Results and Analysis In this section, the performance of forward channel and reverse channel with and without channel estimation has been analyzed through simulation. It is assumed that the channel remains unchanged for the length of the signal (i.e., it undergoes slow fading). Fig 4.18, Fig4.19 shows the results for forward channel and reverse channel estimation with Pilot symbols. The graph shows BER v/s E b /N o. Fig 4.20 shows the performance of signal in five channels over fading channel. Figure4.18Channel estimation with pilot symbols ilot Bit Error Rate Calculation for Forward Channel: Common simulation parameters: Frame Length 100 (Bytes) Total Number of Packets 1000 Total Number of Error Packets 100 SNR (Eb/No) is varying from 0 to 20 (db) Maximum Number of Tx Antennas 2 Page 79

22 Maximum Number of Rx Antennas 2 Total Number of Pilot Symbols per Frame 8 Modulation Method BPSK Bit Error Rate Calculation for different Random signals Bit Error Rate: (bits/sec) Bit Error Rate: (bits/sec) Bit Error Rate: (bits/sec) Bit Error Rate: (bits/sec) Bit Error Rate: (bits/sec) Bit Error Rate: (bits/sec) Bit Error Rate: (bits/sec) This section highlights the simulation of coherent binary phase-shift keying (BPSK) modulation over flat-fading Rayleigh channels. The simulation is an end-to-end system showing the encoded and/or transmitted signal, channel model, and reception and demodulation of the received signal. (appendix I and III) We run the system functionality in a range of E b /N o points to generate BER results that allow us to compare the different systems. Figure 4.19 Transmission of signal through Reverse Channel. Bit Error Rate Calculation for Reverse Channel: Page 80

23 Common simulation parameters: Frame Length 100 (Bytes) Total Number of Packets 1000 SNR (Eb/No) is varying from 0 to 20 (db) Maximum Number of Tx Antennas 2 Maximum Number of Rx Antennas 2 Modulation Method BPSK Bit Error Rate Calculation: Bit Error Rate: (bits/sec) Bit Error Rate: (bits/sec) Bit Error Rate: (bits/sec) Bit Error Rate: (bits/sec) Bit Error Rate: (bits/sec) Bit Error Rate: (bits/sec) Bit Error Rate: (bits/sec) Bit Error Rate: (bits/sec) Bit Error Rate: (bits/sec) Bit Error Rate: (bits/sec) Bit Error Rate: (bits/sec) Page 81

24 Figure 4.20 Signal performances over Five path BER at Eb/N : BER of 1 System is : BER of 2 System is : BER of 3 System is : BER of 4 System is : BER of 5 System is : Receivers Power is More BER at Eb/N : BER of 1 System is : BER of 2 System is : BER of 3 System is : BER of 4 System is : BER of 5 System is : Receivers Power is More BER at Eb/N : BER of 1 System is : BER of 2 System is : BER of 3 System is : BER of 4 System is : BER of 5 System is : Receivers Power is More represents lowest BER BER at Eb/N : BER of 1 System is : BER of 2 System is : BER of 3 System is : Page 82

25 BER of 4 System is : BER of 5 System is : Receivers Power is More BER at Eb/N : BER of 1 System is : BER of 2 System is : BER of 3 System is : BER of 4 System is : BER of 5 System is : Receivers Power is More Figure 4.21 The ROC for the EMDM and IMDM when the FIR filter length changes adaptively with the Doppler rate. Page 83

26 Figure 4.22 The ROC for the EMDM and IMDM with different spreading factors.figure 4.21 and Fig 4.22 show the Receiver Operational Characteristic (ROC) curves of the IMDM and EMDM for different spreading factors Nc; 32, 128, and 256. It is apparent that the detection performance, for both schemes, is directly proportional to the processing gain [94]. This is mainly because the processing gain determines, actually scales down, the interference power at the output of the despreader. Hence, at spreading factor 32, which is relatively low, the detection performance deteriorates severely. However, the IMDM maintains a performance improvement in the range of 8 12% for the considered processing gains compared to the EMDS when the probability of false alarm (threshold level) is set to The Probability density function (pdf) pc i,ds (C i,dscdma) of the channel capacity C i,dscdma, is plotted in Figure 4.23 and 4.24 shows the function of the channel capacity per user C i,dscdma for K=10 users per cell as an indicative value (in real cellular systems the actual number K of users per cell is of the order of 50). In addition, the following system parameters are assumed: (i) totally constant allocated 30 KHz, (iii) number of hops per transmitted bit: M=8, (iv) signal bandwidth of DS transmission: Wds=1.25MHz, (v) processing gain: G p =41.6 and (vi) average received SNR over the signal bandwidth Ws: S=20dB. The pdf of the available average channel capacity per user (in the Shannon sense) for a hybrid DSCDMA cellular system when operating in a Rayleigh fading environment Page 84

27 is analytically examined. Then, it is derived theoretically without applying a lengthy simulation process or complex theoretical algorithms, a novel mathematical general parameters (Appendix II) In addition, the probability that the channel capacity per user does not exceed the available average channel capacity per user is derived. The final expressions DSCDMA cellular system, when operating in a Rayleigh fading environment [95]. Figure 4.23 Probability density function (pci,ds) of channel capacity Ci,DSCDMA Figure 4.24 Probability density function (pci, DSCDMA) (Ci, DSCDMA) of channel capacity. Page 85

28 . Channel capacity per user Ci,DSCDMA does not exceed the average channel capacity per user Ci,DSCDMA, Rayleigh (expressed in bits/sec) for a hybrid DSCDMA system when operating in a Rayleigh fading environment for: K=10,S=20dB, Wt=10MHz, Ws=30KHz, Wds=Wt/M, G p =Wds/Ws and for (a): M=5, (b): M=8 and (c): M=10. Recently, a worldwide convergence has occurred for the use of Orthogonal Frequency Division Multiplexing (OFDM) as an emerging technology for high data rates. In particular, many wireless standards (Wi-Max,IEEE802.11a, LTE, DVB) have adopted the OFDM technology as a mean to increase dramatically future wireless communications. OFDM is a particular form of Multi-carrier transmission and is suited for frequency selective channels and high data rates. This technique transforms a frequency-selective wide-band channel into a group of non-selective narrowband channels, which makes it robust against large delay spreads by preserving orthogonality in the frequency domain. Moreover, the ingenious introduction of cyclic redundancy at the transmitter reduces the complexity to only FFT processing and one tap scalar equalization at the receiver. OFDM converts a frequency selective channel into a collection of flat fading channels. As previously stated, one of the attractive features of OFDM is that, for a certain delay spread, the complexity of an OFDM modem vs. sampling rate does not grow as fast as the complexity of a single carrier system with an equalizer (thanks to the use of redundancy). The reason is that when the sampling rate is reduced by a factor of two, an equalizer has to be made twice as long at twice the speed, so its complexity grows quadratically with the inverse of the sampling rate, whereas the complexity of OFDM grows only slightly faster than linear. This makes easier to implement modems, which have to handle data rates exceeding 20 Mb/s. In OFDM systems, only simple (scalar) equalization is performed at the receiver (whereas in the context of single carrier transmission, a matrix inversion is required). Indeed, provided that the impulse response of the channel is shorter than the Guard interval, each constellation is multiplied by the channel frequency coefficient and there is no Inter-Symbol Interference (ISI).However, the channel still has to be compensated by a multiplication of each FFT output by a single coefficient. OFDM does not capitalize on channel diversity, which prohibits the use of plain OFDM schemes in fading environments. The diversity achieved by the OFDM system can be less than a single-carrier system employing the same error control code in a signaling Page 86

29 environment rich in diversity. Indeed, due to frequency flat fading, the transmitted information on one OFDM sub channel can be irremediably lost if a deep fade occurs. Moreover, the Rayleigh behavior of such fading can have a dramatic impact on the performance of uncoded OFDM schemes. Methods based on coding (convolution codes, block codes, multidimensional constellations, turbo-codes are usually employed with the use of interleaving to combat fading. Figure 4.25 OFDM scheme. To get diversity for CDMA signal OFDM is used and also to get orthogonality for each signal transmitted over channel. Table 4.4 PHY MODES OF IEEE802.11A. Modulation Code Rate Net rate on top of PHY Byte per Symbol BPSK ½ 6 Mbit/s 3 BPSK ¾ 9 Mbit/s 4.5 QPSK 1 ½ 12Mbit/s 6 QPSK ¾ 18Mbit/s 9 Page 87

30 This chapter presents the simulation results for the WCDMA system at different channel conditions. Simulation results include BER vs Eb/N0 and BER vs Number of interferers. Improvement in BER due to error correction coding scheme is shown for a 9.6 kbps uplink service. The following figures show the BER vs Eb/N0 curves for different number of users. The spreading factor at the data channel is 32. The channels are Indoor channel and the simulation resolution is 5 samples per chip. We assume that the received signals from all the users at the base station have equal power, i.e. we have perfect power control. Figure 4.26 BER vs Eb/N0 at the WCDMA Uplink for Indoor Channel. (The Spreading Factor of the channel is 32.The Number of Interferers Vary from 0 to 12) We can see from Figure 4.26 that the system becomes interference limited as the number of interferer increases. This is expected for any communications system employing CDMA as the multiple access technique. BER performance at the downlink is also presented in this section (figure4.27). The following figures illustrates the BER vs Eb/N0 curves at the downlink for different number Page 88

31 of interfering users. The simulation resolution is 5 samples per chip. BER for Indoor channel is shown. The spreading factor is 32. We assume that the signals for all the users are transmitted at equal power. Figure 4.27: BER vs Eb/N0 at the WCDMA Downlink for Indoor Channel. (The Spreading Factor of the User is 32. The Number of Interferers Vary from 0 to 12) As in the uplink, the system is interference limited at the downlink for higher number of interfering users. However, since the users are synchronized, the orthogonality among downlink. The system is interference limited at both the links for higher number of channel 4.7 Conclusion In this chapter, the implementation of multipath detection method in CDMA system has been explained. From the results obtained by simulation, it was found that the IMDM method performs better than EMDM method. Further, the signal took five different paths to reach the signal and the BER for the signal having maximum received power was calculated. After finding the lowest BER and highest received power of the set of signals is studied. ROC represents that the detection performance for both methods is directly proportional to the processing gain. Hence, at spreading factor 32, which is relatively low, the detection performance deteriorates severely. However, the IMDM maintains a performance improvement in the range of 8-12% for the considerable processing gains Page 89

32 compared to EMDM when the probability of false alarm is set at A low complexity version of the proposed scheme was also presented and its efficiency has also been demonstrated through computer simulation.. Page 90