Introduction of mobile broadband is beneficial for subscriber through which Internet can be used in mobile every place where the services of operators are available. Mobile broadband requires more bandwidth as compare to voice services and this is problematic for operators with respect to revenue while providing flat rate subscription.
In Swedish Telecommunication Market Statistics Report  mobile broadband users in Sweden increased and became as prominent part of mobile service. In Diagaram1, the numbers of mobile broadband subscribers in Sweden before 2007 are low as compare to home based internet service (DSL). From 2007 it tends to increase rapidly and in present they reached to higher number of subscription.
Diagram 1. Number of broadband subscription in Sweden 
On the other side the increment in mobile subscribers affects the traffic in the mobile network. According the PTS statistics , traffic is rapidly increasing every year. The effect of increasing traffic is shown in Diagram 2. In the beginning of 2007 the mobile broadband traffic in Sweden is about 500 Tbytes and this reaches to around 9000 Tbytes in the end of 2008. This drastic increase of traffic is just only for one year and this will increase furthermore in future.
Diagram 2. Total Data Traffic in Mobile Networks in Sweden 
With the increase of mobile data traffic the operator's revenue degraded, as stated earlier mobile broadband requires more bandwidth. Bandwidth can be increased installing more new macro base stations, buying spectrum, enhancing coding schemes but these are not cost optimal techniques. As a consequence the operator revenue which is higher in case of voice users and increases with the rises of voice users is seems to be not valid for the case of mobile broadband user with existing system. Increment in data traffics major factor of operator low revenue (see Diagram 3). Revenue became lower and lower with more GByte .
Diagram 3. Revenue per transmitted GByte in Sweden mobile networks 
Introduction of 3G services beneficial for users for providing mobile broadband services, on the other hand there are some issues with the deployment of 3G systems that cannot be omitted. The performance of 3G with respect to coverage especially inside building or indoor areas and capacity is not reliable when comparing with old 2G technology.
Mobile broadband users mostly active in buildings, offices, centrums and homes because of cold weather situations and during surfing of internet on mobile they are stationary. It implies the indoor users can not be neglected and majority of mobile broadband users are indoor users.
Subscribers expect coverage in every where outdoor and indoor with the new 3G technology which offers mobile broadband. Therefore the services must be at least every place where the existing 2G networks cover. Mostly third generation network are designed to provide services outdoor. As a consequence the third generation indoor services are poor as compare to second generation which is capable to providing services outdoor and indoor .
Diagram 4.Indoor Coverage from Macro Base Station 
Macro cell is common way to provide outdoor and indoor coverage (see Diagram 4). For indoor coverage problems mentioned below may be appearing:
The radio wave from outside to in-building suffers free space losses with additional penetration losses. With advance building architecture and style it causes more difficult to penetrate. The attenuation in building also depends on the type of obstacles because building contains several types of obstacles like concrete walls, woods, glass, and metal. Some obstacles have higher attenuation like concrete walls and some have low attenuation i.e. wood, glass etc.
With the introduction of penetration loss, users which are present at far side of building experience greater penetration losses and bad coverage because they are far from macro base station.
3G uses higher frequency as compare to 2G services. Higher frequencies have a higher penetration loss than lower frequency i.e. the signal strength decrease faster. As the radio frequency increases as a consequence the rate of attenuation increases as compare to attenuation which occurs in low frequencies. Diagram 5 shows the attenuation for lower and higher frequency .
Diagram 5. Higher frequencies have higher attenuation on penetrating obstacles 
Radio wave also tends to reflect when striking with dense surfaces like metallic walls or vessels. The radio wave reflects several times before it reaches to receiver unit. When radio wave reflects some of power is absorbed by the obstacles and this results attenuation in the reflected radio signal. The attenuation due to reflections is also dependent on the frequency. The attenuation due to reflection is much higher when comparing with low frequency (see Diagram 6) .
Diagram 6. Higher frequencies have more losses on reflection 
Turbo 3G provides much higher data rates on excellent RF-link condition. The Macro base station cannot deliver to indoor user.
The Macro base station mounted on the roof of building have problem at their base due to reflections from the neighbors buildings.
Problems mentioned above can be resolve by dedicated in-building solutions which provides better services and satisfies users inside buildings with higher bit rates and coverage requirements. The dedicated indoor solution also helps gaining revenue from the mobile broad users with low installation and operational costs.
Dedicated in-building coverage helps offload the macro network. In-building solutions offload the macro base station by reducing average downlink power levels and indoor solutions are isolated from the macro cells as a result low interference occurs. Better capacity and coverage requirement can be fulfilled by using indoor solutions such as Distributed Antenna Systems, Femtocells, picocells and WLAN. These indoor solutions are low coverage and high data rates solutions .
Distributed Antenna System (DAS)
Distributed Antenna System is a network of antennas distributed in a building to provide higher capacity and coverage requirements. These antennas are connected via coaxial cable or fiber optic to the central base stations which are located in ground location of the building. Moreover Distributed Antenna System consist couplers and power splitters for distributing RF signals .
Femtocells are low-power wireless access points that operate in licensed spectrum to connect standard mobile devices to a mobile operator's network using residential DSL or cable broadband connections. Femtocells are low power radio stations that are designed for indoor solutions. They are operating in licensed as well as unlicensed spectrum. Major advantages of Femtocells are low transmission power, higher capacity, better coverage and economical  .
Picocells technology has been around for many years, providing cellular coverage and capacity in spotty and areas where cellular signals do not reach. The market for picocells has been limited, with scattered usage. Picocells have higher cost of installing and managing . A Picocell is smaller than microcell physically and with respect to coverage area and larger than Femtocells .
WLANs provide wireless network communication over short distances using radio or infrared signals instead of traditional network cabling. WLAN uses a device called Access Point (AP). Terminals communicate with WLAN using WLAN adaptor .
The importance of Indoor users can not be avoided with the introduction of mobile broadband. As a consequence which solution is suitable for providing indoor solution with respect to cost optimal and technical point of view?
The aim of this thesis project is to compare both technical performance like cost and capacity of the wireless indoor access and the transmission as well as the different options for cooperation and business model. The following questions need to be addresses:
When is WLAN better than Femtocell with respect cost optimal solution in higher user density with high wall attenuation or low wall attenuation scenarios?
Is DAS dominates than WLAN in terms of economical solution in average user density with high wall attenuation scenario?
Will a model be established for effective sharing of cost and revenue of resources among actors having some proportion of users?
Indoor techniques Distributed Antenna System, Femtocells, Picocells and WLAN are widely discussed in past. Many business and white papers representing comparison of different RATs with respect to complete cost analysis.
Femtocells are studied in  and , which provides cost analysis and several business models for market actor for sharing resources. The major emphasis of this study is on Femtocells In which the Femtocells are compared with macro cells with respect to cost, capacity and coverage in deployment scenarios cases. Secondly a two dimensioned business model in  is discussed between operator of the indoor network and different kind of users that can access the network.
Studies ,  and  cover WLAN cost analysis. In  WLAN and Picocell is compared with macro and micro base station. The 3G system Evolution-Data Optimized (EV-DO) is compared with WLAN in . Femtocell and WLAN which are indoor solution providing higher data rates are discussed in .
Picocells are covered in  and . Picocell compared with Macro cell in  and in  Picocell compared with Macro and Micro cell.
After all it is concluded many studies cover some part of this thesis with respect to cost analysis. In mostly studies indoor solutions compared with Macro base station and some studies compares indoor solutions as well.
The attenuation in the radio signal depends on the type of obstacles in the building. Two type of wall attenuation is assumed. The attenuation from concrete walls is Strong wall attenuation and from wood and glass is weak wall attenuation.
The measurements will done on Skatterverket building located at Solna . There are 4 floors. This building contains fixed amount of users which are 300. Each floor have different wall attenuation in which 1 floor have strong wall attenuation and remaining three are weak wall attenuation.
A peak hour is defined with maximum number of broadband usage. The total monthly demand for broadband access is 7 GByte per user.
(I am not sure about the architecture now, when I go to building I will update this part)
Network Model and Dimensioning
The analysis is based on user demand i.e. number of users in a building and the data usage (GByte) per user and month. The data usage is converted to capacity per area unit (Mbps/km2) computed for a number of 'busy hour'. It is assumed that the capacity share among users is a 'best effort type' in a cell , e.g. 10 users will get an average data rate of 0,1Mbps generated from 1 Mbps base station. The required number of base station or antenna elements depends on the type of indoor solution with respect to capacity and coverage requirement.
The dimensioning is capacity limited and we need to cover single building. The system is UMTS with single carrier. Femtocell, picocells and DAS will operate in UMTS frequency and data rates. WLAN is operated in 2.4 GHz or 5 GHz. The maximum or peak capacity of single Femtocell, Picocell and DAS base station is 2Mbps in a cell when using UMTS. Single WLAN access point has capacity of 108 Mbps.
The aim is to provide coverage in in all places inside the building. Coverage is depending on number of base station. Since the dimension is capacity limited so deployed number of base stations must satisfies the capacity and coverage. For coverage and capacity its depends on the type of attenuation. In Weak wall attenuation scenario the number of base station are calculated xx (it will decided later after looking building) and for strong wall attenuation the number of base station are xx. Depending on the attenuation type on each floor the number of base station vary, strong attenuation scenarios require more base station as compare to weak attenuation scenarios.
There are various indoor propagation models are used to predict the path loss in indoor environments and no standard model can be predicted for indoor path loss. The models mostly depend on the environment of indoor system.
The free space loss model is use when there is no any object between transmitter and receiver which is ,
In building we have wall attenuation W and floor attenuation F. Attenuation also more dependent on number of walls n and number of floors m through which signal penetrates . The above equation now becomes:
Will be discussed after visiting building
DAS (Distributed Antenna System) Architecture Model
The problems from Macro base station can be minimized by using DAS. The principle of DAS is to subdivide one area into smaller areas and let each DAS antenna cover that area . DAS shows performance improvement in terms of coverage area, path loss and less Tx power from BS and mobile phone . In DAS several antennas are connected to the base station via coaxial cable CAT 5/6 or fiber optics. The architecture for DAS is shown in Diagram 7.
Diagram 7. DAS providing indoor coverage 
DAS is implemented using passive devices is Passive DAS. This is most common type of DAS in which coaxial cable is connected from base station to antennas. Passive DAS consists of coaxial cable, splitter, attenuators and filter. The architecture of passive DAS is shown in Diagram 8. Passive DAS is used in harsh environments e.g. mostly deployed in parking areas and tunnels etc. Although this is simple but due to losses in coaxial cable the losses especially in higher frequencies which results non -uniform coverage and more power required from mobile station .
Diagram 8. Passive DAS Architecture 
Active DAS uses active components with coaxial cable CAT 5/6. Active DAS contains main unit, expansion unit and remote unit. The architecture for active DAS is shown in Diagram 9.
Diagram 9. Active DAS Architecture 
The main is the central part of the network. Main unit provides connectivity between base station and expansion unit via fiber optic cable. It monitor the whole system and provides signals to base station. There are LED's installed on main unit which shows the network status and the power consumption for main unit is approx 30W .
Expansion unit responsible for converting optical signal arrives from the main unit to electrical signals that further use by remote unit. Maximum 4 expansion units can be connected to the main unit and typical the distance between main unit and expansion unit is 1,5 km with multi mode fiber and 6 km for single mode fiber. The expansion unit is also responsible for providing power to remote unit .
Remote units are connected to expansion units through the coaxial cable, maximum 8 remote units can be connected to single expansion unit. The expansion unit delivers DC power to remote unit so there is not any external power required for remote unit. The uplink losses are minimized by connecting antennas near the remote unit .
With the use of active components and fiber cable the losses in the cable and uplink are minimized and hence the battery life is increased with minimization of uplink losses.
The cost structure analysis is a comparison of different indoor solution by maintaining same user demand with respect to capacity and coverage requirement. For the modeling of capacity demand and the cost structure break down we use the methodology proposed and used by Johansson . In this thesis both capital expenditure (CAPEX) and operational expenditure (OPEX) will take account for all indoor solutions. The factors involve in CAPEX and OPEX are different for RAT type which are given below:
The elements for CAPEX and OPEX cost for different RATs are given below :
Access Point ( Base Station)
Radio Network Controller (RNC)
O & M
Access Point (Base Station)
Annual Cost of Broadband/Femtocell
Access Point Controller
Antenna Element Installation
Antenna Elements Maintenance
Backhaul Connection Maintenance
Access Point ( Base Station)
Radio Network Controller (RNC)
O & M
Business Modeling and Analysis
The expensive task for operators to built whole coverage region especially indoor areas for broadband. They may degrade their revenue while providing flat rate for mobile broadband services and it is unnecessary to built same technology several times in a building. There should be some cost optimal business model among operator. The business model requires cooperation among operators and market actors so that they share devices for indoor areas. Several possibilities can be used for creation of business models. Some of them are follows:
In Business Modeling major emphasis in this thesis is consider on sharing cost and revenue. How it is distributed among actors depending on the customers they have?
Business Modeling Scenario
Single building scenario may not perfect for business modeling analysis because in single building some operators have customers and some do not have. So for that the scenario for building is good when considering some large area and group of building so that effective modeling can be done.
Consider a scenario having 10 buildings with same size and structure. There are 3 operators having customers in that 10 buildings. The users from each operator is not equal in 10 buildings, they have some fixed proportion of users in that 10 buildings which is giving below:
Proportion of user in 10 buildings
Operator 3 is the facility owner of all 10 buildings and operator 1 and 2 wishes to share resources with operator 3.
The analysis of different indoor solutions is based on cost and performance. Better coverage and capacity is the requirements for users and with the least CAPEX and OPEX cost.
CAPEX and OPEX Cost
Initially CAPEX and OPEX cost of each indoor solution will represented using Pie diagram. The aim of this Pie diagram is to show elements which are involve in CAPEX and OPEX costs. For different indoor solution different elements exists that representing CAPEX and OPEX cost. An example is shown in Diagram 4 and Diagram 5.
Diagram10. CAPEX Cost
Diagram 11. OPEX Cost
Cost optimal indoor solution is obtained by comparing all indoor solutions with respect to CAPEX and OPEX which are shown in Diagram 6 and 7.
Diagram 12. Femtocells, Picocells, DAS and WLAN CAPEX
Diagram 13. Femtocells, Picocells, DAS and WLAN OPEX
Time and Activity Plan