Estimation of Design Parameters for Cellular WCDMA Network

The estimation of design parameters for WCDMA network which support different data rate (Multimedia) services is an essential step in network planning. The estimated parameters includes: Cell coverage range using path loss propagation model, average and maximum path loss, throughput, capacity / km, number of users per cells blocking capacity, tracking efficiency. The parameters are computed for different data rates.

The link budget of WCDMA (3G) have some specific parameters that are not used in GSM (2G). The most important are: -Interference-margin: is needed in the link budget because of the loading of the cell, the load factor affects the coverage, the more loading is allowed in the system the large is the interference margin needed in the uplink and the smaller is the coverage area. For coverage limited cases, smaller interference margin is used (1-3 dB) corresponds to 20-50% loading. While in a capacity limited cases a larger interference margin is used.
-Fast fading margin (= power control headroom): Some headroom is needed in the mobile station (MS) transmission power for maintaining closed loop fast power control to effectively compensate the fast fading [1]. Typical values of the fast fading margin are 2-5 dB for slow moving mobiles.
-Soft handover gain: Handover (soft or hard) gives gain against slow fading (Log normal fading) by reducing the required Log-normal margin. Since slow fading is partly uncorrelated between the Base Stations (BSs), by making a handover, the mobile can select a better BS. Soft handover gives an additional macro diversity against fast fading by reducing the signal power to noise power (E b /N o ) relative to a single radio link. The total soft handover gain is assumed to be between 2-3 dB in the example below including the gain against slow and fast fading. Three examples of link budgets are given for typical universal mobile. Telecommunication Services (UMTS) which implements WCDMA.
-144 kbps real time data.
-384 kbps non-real time data. In the urban macro-cellular environment at the planned uplink a noise rise of 3 dB and interference margin of 3 dB is reserved for the uplink noise rise. The assumptions are in the link budget [2] for the receivers and the transmitters are shown in tables 1 and 2. Once the maximum allowed propagation loss in the cell is known, then applying propagation model for cell range estimation. The propagation model is chosen so that it is optionally describes the propagation condition in the area. The restrictions on the model are the carrier frequency used, BS to MS distance, BS and MS antenna heights. For urban and suburban macro-cellular system Okumura-Hata propagation model is used [3][4].  The path loss for suburban area is by adding a correction factor of 8 dB to the L PUA [1].

Maximum
For the allowed path loss [1] in the case of voice and data, the cell coverage range is calculated and the results are shown in The theoretical spectral efficiency of a WCDMA can be calculated using equation (3)  bit rate of user j. j : activity factor for uses j and is 0.67 for voice and for data.
The i factor depends on propagation environment, overlapping cell areas, the number of sector in the base station, the traffic intensity and its distribution, as well as on the distance to the serving and interfering cells [6].
For voice service network, where all users in the cell have low bit rate (R) i.e.
and hence equation 3 is simplified to: The noise rise is defined in [5] as: The load equation (4) is used to make a semi-analytical prediction of the average capacity of WCDMA without going into system level simulation. Equation (6) can be used to predict cell capacity and noise rise. From equation (5) it can be seen that as UL becomes close to one, the corresponding noise rise approaches infinity and the system has reached its pole capacity.
For noise rise calculation, the value of E b /N o is different for different data rate as given in table 4.  (3) and (6) the results are shown in figure (1).
Figure (1) shows the noise rise versus the number of users for different data rate. It can be seen for figure (1) that for constant noise rise value, the number of users decreases for higher bit rate.
Throughput is defined as: Where, N is the number of users and R is the bit rate. Figures (2-a) and (2-b) show the variation of noise rise with throughput for different interference level (i), for voice (12.2 kbps) and data (64 kbps) which show that as interference increases the noise increases and system capacity decreases.
The downlink load factor DL can be defined similarly as for the uplink although the parameters are slightly different [1] [2] [5].
Comparing equations (7) and (3), j is the most important new parameter which represents the orthogonality factor in the downlink, typical j is between 0.4 to 0.9 in multipath radio channel. In downlink the value of i depends on the user location and therefore is different for each user j. The load factor can be approximated by the average value across the cell.
The minimum required transmission power between the base station transmitter and the mobile receiver is the average path loss L and the mobile receiver sensitively in the absences of multipath access interference. The effect of noise rise interference is then added to the minimum power, and the total represents the transmission power required for the user at an average location in the cell. Therefore the total base station transmit power can be represented. -Assuming the required aggregate cell throughput in kbps. -Using the load factor DL from equation 9, then throughput is: Where: R is the bit rate (bps).
BLER is the block error bit.
-Using equation 10, the average path loss L can be calculated.
-The maximum path loss is calculated by adding 6dB from [6] [1] which is the difference between maximum path loss and average path loss.  Figure (3) is throughput versus maximum path loss, and figure (4) is for the number of users versus maximum path loss both are for data of 64 kbps. It can be noticed in both that the load affects the coverage but is not the same for UL and DL. In the DL the coverage depends in the downlink, since in the downlink the power (20 W) is shared between the downlink user, the less power per user. Therefore even with low load in the DL, the coverage decrease as the number of users (or throughput increases).

BS. Tx
When the amount of traffic increases, the downlink capacity can be upgraded in a number of ways [1] [6].
-More power amplifiers if initially the power amplifiers is split between unknown sectors. -Two or more carriers if the operator frequency allocation per units.
-Transmit diversity with a second power amplifier per sector.
The uplink coverage is not affected by these upgrades.
Providing higher capacity is challenging in urbane area (i.e. traffic per km 2 can be very high).
For macrocell layer assuming maximum capacity per site per carrier is 3 Mbps with three sectors, the maximum capacity with maximum site density of 5 sites / km 2 ; with three carriers.
Maximum capacity = number of carrier maximum capacity per site per carrier × maximum dense site density.
Maximum capacity = 3 × 3 × 5 = 45 Mbps/km 2 when capacity in needed but with increased macro site density of e.g. 5 sites/km 2 the maximum site density is 45 Mbps.
For microcell layer with 30 site / km 2 .
The number of channels per cell is calculated for different data rate, as in figure (1) Where: traffic density is in Erlarg.
Call arrival rate and call departure rate both are in calls / sec. If the capacity is hard blocked i.e. limited by the amount of hardware, the Erlarg capacity can be obtained from the Erlarg B model [8]. If the maximum capacity is limited by the amount of interference in the air interface it is a soft capacity and there is no single fixed value for the maximum capacity.
Soft capacity is important for high bit rate real time data users (e.g. for video connections). In the soft capacity calculations below, it is assumed that the number of subscribers are the same in all cells, but the connection starts and ends independently. The call arrival interval is a Poisson distribution.
For WCDM soft capacity is defined in equation (15) below as: The basic Erlarg B formula is then applied to the larger channel pool (interference pool).
-The Erlarg capacity obtained is shared between the cells.
The soft capacity estimation procedure is as follows: -Calculating the number of channel per cell (N) (for equally loaded) based on uplink load factor using equation (4) The various design parameters have been estimated for different data rates (12.2 kbps to 384 kbps for voice and video application). The essential concluding remarks as: -The cell range decreases with increasing data rate in urban and suburban areas. -Noise rise increase with increasing number of users (or throughput). -For a given number of users (or throughput) the noise rise increase with the interference factor (i). -Cell loading affect the coverage range in the down link more than the uplink.