A Cdma Based Mobile Communications System Computer Science Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

A study has been conducted to examine the feasibility of deploying a mobile communications CDMA based technology in Orange Island together with the, deployment and services infrastructure possibilities to enable the government understand the structure of the network and proposed service delivery.

To this end, some major tasks have been undertaken that will:

Analyze the state of mobile communications technology worldwide, understand market trends together with associated challenges and decide on current and future technology that would provide a capacity to enable fixed and mobile voice and data connectivity.

Provide good recommendations and possible alternatives that would help to integrate new and current systems

Two design options were investigated: voice only with 10kbps and one that includes additional high speed data capability of about 150kbps for internet services. The services will be implemented at varying degrees at the urban, suburban and rural areas. As stated by [1] WCDMA supports high data rate up to 384kbps with wide area coverage, 2Mbps with local coverage apart from high service flexibility and the built in support for future characteristic capacity and coverage enhancing technologies that include but not limited to adaptive antennas. WCDMA is an obvious choice for this design that embraces a rich blend of data capabilities and a thought for future expansion.

Various service and business concerns like modulation techniques, error detection and correction techniques, spectrum regulation and future technologies were also investigated. Service deployment options as regards the number of base stations were considered for 2 main regions: urban and suburban. The spectrum allocated as 1920 - 1980MHz for the uplink and 2110 - 2170MHz were managed in such a way that network expansion was adequately catered for . Further feasibility study would be required to provide accurate information about the sparse population distribution of the rural areas. Options were considered that involve the future service possibility deployment at orange island with necessary information about how the present system can cope or be modified.

Document Information

Main Report

3403 words, 13 pages


Matlab codes (3pages)


1 page


A good network plan is essential to building a cost effective quality network. One of the major things to contend with in drafting out a reasonable plan is that the required 3G mobile service mix and usage is often a guess and no one knows exactly what the consumption pattern would ultimately be [2]. According to the International Telecommunication Union, only 84.49 in 100 people will use mobile phones in a typical city in the U.K with 40% opting for additional data service.

Voda being a city on Orange Island comparable to the size of London measuring 1500km2 in area and a city centre of about 100km2 would have the typical structure that most modern cities have in terms of urban(industrial), suburban(residential) and rural(farmlands and holiday resorts). The mobile communications network design for Orange Island would strive to achieve network and data convergence with services that include voice, browsing, global positioning services, live multimedia serving many high data rate between users and various content offerings as part of the design considerations. The network would also consider growth in users as more and more "bandwidth-hungry" applications are deployed


Deploying a mobile communications network is subject to different technologies that embraces diverse coding, access techniques and modulation formats to provide the required service.

One of the acclaimed world-wide standard for mobile communications is the GSM which was originally designed to handle voice and text services. Spanning an operating frequency of 890 - 915MHz in the uplink and 935 - 960MHz in the downlink, and access techniques of FDMA and TDMA, it was established in Europe in 1982 as a standard for mobile communications [4]. The 200kHz channels are able convey 270kbps data rates based on the Gaussian Minimum Shift Keying modulation technique [5]. When GSM was established, it used Cyclic Redundancy Coding and 1/2 convolution encoding for error detection and correction purposes[5]. The demand for various services on the mobile communications network made GSM to be extended in functionality for more data rates thus giving birth to GPRS and EDGE.

A 2nd generation system like GSM was natively designed to deliver only voice and services unlike UMTS which incorporated functionalities for diverse services from the design stage without any independent network optimization [3]. One of the standards developed under UMTS is the WCDMA (Wideband Code Division Multiple Access). The characteristic 3.84MCps chip rate of WCDMA leads to an approximated carrier bandwidth of 5MHz. The size of this bandwidth makes it possible to deploy highly variable user data rates with so many associated benefits to the system performance. Within an operating license, a network operator is able to improve the capacity of the system by deploying multiple 5 MHz carriers. Apart from its compatibility with 2nd generation systems like GSM, the use of advanced CDMA carrier for multiuser detection and smart adaptive antennas makes it an attractive choice when considering expansion in capacity and/or coverage [3]


Given the above background about some of the available standards for mobile communications, the design specification in orange Island requires a network with multiple data rates in terms of circuit and packet switching. 1920 - 1980MHz for the uplink and 2110 - 2170MHz for the downlink exactly matches what is used WCDMA and the design option involving variable user data rates makes WCDMA an attractive option for this network. With this in mind, 2 x 5 MHz, 3.84MCps, CDMA-FDD channel in the WCDMA would be used. From a link budget calculation, a trade off between coverage and capacity would be done to decide the optimum service structure.

3.1 OPTION 1: A voice only network with a required user data rate of 10kbps


The most important parameters for the calculation of the link budget is Interference Margin, Fast fading margin and soft handover gain [3]. The direct dependence of coverage on interference margin makes it have values between 1.0 - 3.0dB corresponding to 20 - 50% loading as stated by [3] based on the equation below. The load on the network will impact the coverage in many ways because as more and more users are added to the network, higher network noise would be detected by the system.

It is expected that the subscribers would use the service on their mobile while they are on the move, hence a fast power control is needed to compensate for the fast fading with typical values of 2.0 - 5.0dB as studied in [6]. The movement of users between different base stations offers them different levels of power and hence a handover technique is adopted to make a selection of a better base stations. In this design, soft handover gain assumed to be between 2.0 and 3.0 dB would be considered since it gives a gain against slow fading [4]

The energy per bit to Noise power density ratio is essential to the determination of the link budget, simulation tools like Simulink in Matlab or other commercially available ones can be used to tailor the Eb/No value since it is a predetermined value by the network operator. According to literature typical values are 4 and 6 which are 5dB and 7dB respectively. For this design 7dB has been adopted as choice. A Modified version of [3] and [2] as shown in the tables below reveals the assumptions that have been used in the link budget for both the receivers and transmitters

Table 1: Basic Assumptions

MOBILE STATION - Transmitter

Maximum Transmission Power

21dBm / 0.125mW

Antenna Gain


Body Loss

2 - 3dB


Noise Figure


Antenna Gain

19dBi(3 Sector Base Station)

Eb/No requirement

7.0dB (Speech)

Cable and Connector Loss


The approach of [3] helped in generating a link budget for 10kbps which based on the OKAMURA-HATA model, gave the table below.

Table 2: Link Budget calculation Results for 10kbps voice service


URBAN : Path Loss = 137.4 + 35.2log10(R) = 139.94dB

Desired Cell Range R


3 Sectored Cell: Coverage Area for a cell = 1.95R2


Required Urban coverage area (km2)


Required Urban Coverage Sites = Coverage Area / Area of a Cell


SUBURBAN: Assume area correction factor = 8dB

Modified Path Loss = 129.4 + 35.2log10(R) = 139.94

Desired Cell Range R


3 Sectored Cell: Coverage Area for a cell = 1.95R2


Required suburban [1] coverage area (km2)


Required suburban Coverage Sites = Coverage Area / Area of a Cell


The link budget calculation so far has given an idea of the coverage requirements which shows that a high speed 3G network requires many base stations.


The capacity of a CDMA system is determined by the systems interference and the planned quality and grade of service. Overall capacity calculations is based on the determination of the capacity of a single cell and how many users each cell can effectively serve According to the load equation in [3], the number of available channels in the 5MHz bandwidth can be computed according to the formular below

Where N= number of Channels, R = Bit Rate, W = Chip Rate, V = Activity Factor,

= Signal to Interference Ratio and i = Interference factor.

Where values have been selected based on a modified version of [4] as in table below








For speech assuming 50% voice activity


0.65 or 65%

Macro cell with 3 sectors



WCDMA Chip Rate



Bit Rate specification; a Service dependent parameter for this design

For a 3dB noise rise, a load factor of 0.5 is selected and for 6dB, load factor of 0.7488 is selected, a measure of the overall throughput per cell of all simultaneous users is inspected for this two cases by modifying equation 1 to form so that throughput = R x N

Figure 1: Uplink noise rise as a function of uplink data throughput for different SIR values.

For this design, a user would not be discouraged by the usage of the service if out of 100 calls, only 1 is unable to connect then, Assuming a blocking probability of about 1 in 100 calls which amounts to about 1% is a reasonable design choice. Assuming the design choice of 6dB Noise Rise and 7dB SIR, a throughput of 531kbps amounts to 53 Channels for the 10kbps bit rate. Based on the Erlang B table, the Traffic Intensity for a 1% blocking probability is 52.808 Erlang and 62.224Erlang for 2% blocking probability. Depending on parameter values chosen initially, the planned capacity is typically from 400kps to 700kbps [2].

Assuming we project that during a busy hour period, a subscriber would make on average, one 75-second voice call every busy hour. The projected service rate for 10kbps voice only network is (75 * 10kbits/s) /(60 * 60) = 0.833kbits/s/user/busy hour. From the design possibilities revealed in the graphs, assuming we consider the 531kbps throughput design, then each cell can be dimensioned for 531 / 0.833 = 637users.

Given a 3 sector cell of area 3 * 2.6R2 = 7.8R2, with the N = 53 known as estimated in load calculations, values of R against the density of users in urban, surburban and rural areas, Range of a cell can be determined for the various regions as shown in figure 2 below

Figure 2: Determination of Cell Range based on User Density and Blocking Probability for 10kbps, 53 Channels voice network.

Since the service rate is fixed at 10kbps, then the design choice at the city center would involve increasing the number of base stations. Despite capacity calculations, mobile networks are usually planned first to meet the coverage objectives because the real traffic load is what determines how capacity sites and transceiver upgrades are installed [2]. Both the urban and suburban are capacity limited hence to optimise the used resources, transmitter diversity will be employed or the number of transceivers is increased gradually until the capacity matches the coverage.

Based on the calculations used in [7], the bandwidth is :

(Number of channels per sector - 1) * Eb/No * ( 1 + Interference Factor) * Voice Activity Factor.= (53 - 1)* 100.7* (1 + 0.65) * 0.5 = 215

Since the bit rate = 10kbps, Required bandwidth = 215 * 10kbps = 2.15MHz, which is well within the allowable spectrum. The throughput designed for the of 531kbps for the 637 users uses up only 43% of a single bandwidth of 5MHz for this W-CDMA system.

3.2 OPTION 2: A voice and data network with a voice user data rate of 10kbps and the data rate for other traffic (e.g. internet data) of 150kbps.

Based on [8] the increase in the demand for real-time data services, would place a high demand on network services in the near future especially for mobile e-commerce applications, to embrace this future move, this design would consider real-time data service for an indoor user covered by an outdoor base station. The coverage calculation would follow up from what was obtained in the last section. Assuming as before the data in the table as obtained modified from those of [3] & [2], taking log normal fading margin of 4.2dB, indoor vehicle loss of 15dB and soft handonver gain of 2dB,

Table 5: Link Budget calculation Results for 150kbps data service


URBAN : Path Loss = 137.4 + 35.2log10(R) = 141.582dB

Desired Cell Range R


Using 3 sectored Cell: Coverage Area for a cell = 1.95R2


Required Urban coverage area (km2)


Required Urban Coverage Sites = Coverage Area / Area of a Cell


SUBURBAN: Assume area correction factor = 8dB

Modified Path Loss = 129.4 + 35.2log10(R) = 141.582

Desired Cell Range R


Using 3 sectored Cell: Coverage Area for a cell = 1.95R2


Required suburban [2] coverage area (km2)


Required suburban Coverage Sites = Coverage Area / Area of a Cell


Using the same link budge approach with the change in parameter requirements of SIR = 1.5dB and The link budget calculations above gives a rough estimate of what the number of base stations required for Coverage.

For this peculiar option that has the need for both voice and data, analysis would proceed by first assuming that for the 5MHz bandwidth, only voice is used and also for the situation when only data is used then tradeoff is done in terms of percentage by expected user pattern to find the optimum bandwidth with which this service can be deployed. Following the method used by [7] as shown in the equation below

Where N =Number of Simultaneous users per sector of a cell. Gp = Process gain, Gp = B/Rb, β = Interference due to mobiles in other cells transmitting on the same channel = 0.85 for 3 sector cells.

where B is the bandwidth and Rb is the information bit rate




Signal to Interference Ratio








DPCCH overhead during data transmission.

For speech assuming 50% voice activity


0.65 or 65%

Macro cell with 3 sectors



WCDMA Chip Rate




Bit Rate specification for data

Bit rate specification for voice

With 100% utilization by 10kbps users, 80 users per sector can be accommodated in the 5MHZ band while 150kbps (100% utilization) can accommodate 12 users per sector. By making a trade off between percentage given to simultaneous users of voice and data within the available 5MHz bandwidth, the design possibilities are shown in the graph below generated by the code in appendix.

Figure 3: Bandwidth Utilization by percentage based on appendix

According to the US 3G and wireless data services: Market analysis and forecasts 2005-2010 [3] , about 30% o f service utilization is given to out to data and 70% goes to voice hence, approx. 1.5MHz is assigned for data services and this lead to a sector capacity of about 4 users and hence 12 users for a 3-sector cell if each user uses the data service for the whole time. Correspondingly 70% capacity would use 3.MHz leading to 56 users per sector and approximately 168 users for a 3 sector cell. Total number of simultaneous voice and data users per sector is 180 The Table below shows the projected number of base station required at the urban, and suburban area considering this capacity allocation by population size. Based on ITU's projection that 85% users use mobile service in a city and that about 30% would use both voice and data service at a time. From the population information given, it is estimated that approximately 150 base station would be required at both urban and suburban areas.

This shows that a service delivery employing both voice and data is capital intensive. This estimate is only an approximate guess based on capacity. By combining both capacity and coverage, the Number of Base stations would change significantly.

Another possible design alternative would be to assign a specific 5MHz WCDMA band to data usage so that only data users are found within that spectrum. Following the Using a load factor of 0.7488 for a 6dB noise rise, (based on equation 1), the throughput is as calculated below

Throughput = R x N =

Using the matlab code of appendix. Uplink noise rise is plotted as a function of throughput in the figure below

Figure 4: Noise Rise as a function of throughput in a 150kbps data network

Capacity improvements can be obtained by increasing the Noise Rise for the data network until the maximum capacity of about 1600kbps .The initial design involving voice only network made a blocking probability approximation of 1%, some systems use 2% and it still provides acceptable performance because Digital voice is robust in terms of error probability when compared to data service. These numbers would not be good for transmission of data and hence a value of blocking probability of 10-7 is proposed.



The fact that this service combines so many data rates and different QOS schemes, it is expedient to consider an Adaptive Modulation technique which will enable the system to choose between different modulation techniques depending on user service need and signal quality. To this end, High speed packet data access (HSPA) would be deployed in the system because it uses QPSK or 16QAM for the downlink and the uplink as stated in [9] so that the required adaptive modulation can be obtained.

To improve the performance of this network by reducing error probabilities for better Quality of service, turbo coding the rate-1/3 turbo coder and decoders would be adopted because as stated by Shannon, it almost attain s the optimum channel capacity limit.


The high price tag placed on bandwidth makes it important to know the exact amount required by the mobile operator which is ultimately determined by the projected number of users and future considerations. The regulatory body in uk has various spectrum allocation for mobile operators like T-mobile and Orange for both 2nd and third generation frequency bands. Based on the results obtained in the initial analysis, the following spectrum will be bought




1920 - 1980MHz

3 x 5MHz



2110 - 2170MHz

3 x 5MHz


The remaining spectrum would be used for future expansion of this network


The nature of the frequency band and service offerings deployed for this network significantly determines the type of handsets, or other network equipments that can be used. This design suggests the use of cellular phones that have dual band beside 3D capabilities to maximize the service offerings. Notable equipment vendor both on the network and consumer side includes Huawei and Motorola respectively.


As the demand for mobile services increases in orange island for services like internet data, live streaming, video on demand and other multimedia services, the present 3G system would run out of capacity and new spectrum would be required for a new generation of mobile systems. The 4G technology presently being standardized is developed from the ground up to include multimedia services for mobile environments at rates up to 2Mb/s for outdoor and up to 20Mb/s for indoor application with the sole aim of providing such services without restriction [7]. According to [7], some of the features expected are unrestricted roaming and mobility for voice and data services over regional and global networks with the possibility of IP architecture apart from interoperability with present 2G and 3G systems


According to [10], a good network plan should be able to address both coverage and capacity requirement of the area under consideration and at the same time ensure possibility of network expansion without a major change to existing sites. The initial design would define a set of parameters for the Coverage and Capacity which in the long run would be lend itself to more subtle adjustment depending on the services usage by users over time.

Based on capacity and coverage analysis done, and cell ranges obtained, pico cells will be used at urban area, macro cells for suburban areas and macro cells for rural areas. Techniques embracing diversity, sectoring and fast link adaptation would be used to obtain required performance.

The population statistics, taxation, income/wealth distribution and spending habits of voda island would also be a useful tool in deciding the future mobile usage in each of the urban, suburban and sparse population environment. This proposal is aimed at defining the initial network structure which would be influenced by customer intake and behaviour towards deciding the network development direction.


[1] Zamzuri, H.B.; Ibrahim, M.B.; , "Simulation on WCDMA for 3G mobile systems," Telecommunication Technology, 2003. NCTT 2003 Proceedings. 4th National Conference on Telecommunications Technology [online] Available at ieeexplore.ieee.org [Accessed May, 2010]

[2]3G Network Planning Basics [online]. Available at www.umtsworld.com [Accessed May, 2010]

[3] Holma, H(2005); " WCDMA for UMTS: Radio Access for Third Generation Mobile Communications". 3rd Edition. NJ, USA: Johly Wiley & Sons, Incorporated.

[4]Anonymous (2009). GSM [online]. Available at: http://en.wikipedia.org/wiki/GSM [Accessed 21 April 2010]

[5] Schwartz, M. (2005); "Mobile Wireless Communications". 2nd edition. New York: Cambridge university press

[6]Sipila K., Laiho-Steffens, J. Ja¨sberg, M. and Wacker, A., 'Modelling the Impact of the Fast Power Control on the WCDMA Uplink', Proceedings of VTC'99, Houston, Texas, May 1999, pp. 1266-1270.[online] Available at ieeexplore.ieee.org [Accessed May, 2010]

[4] Karim, M.R and Sarraf M. (2002); "WCDMA and cdma 2000 for 3G Mobile Networks " McGraw-Hill e-book

[8] kyoung-don kang, Sang h. son and John a. stankovic (2003); "Differentiated Real-Time Data Services for E-Commerce Applications" in Electronic Commerce Research, vol 3 pp 113-142[online] http://www.cs.virginia.edu/papers/diff-rtds-e-comm-ecr03.pdf [accessed May, 2010]

[9]SPG Media Limited(2009):"High Speed Downlink Packet Access Mobile Broadband" [online] www.mobilecomms-technology.com/projects/hsdpa [Accessed April, 2010]

[10]QUALCOMM; "WCDMA Network Planning and Optimization" ESG Engineering Services Group [online] available at : http://www.docstoc.com/docs/14211814/WCDMA-Network-Planning-and-Optimization [Accessed May, 2010]