The thesis titled. Handover Management in GSM Cellular System ..... French Radiocom 2000 and the Italian RTMI/RTMS helped make up Europe's nine analog.
HANDOVER MANAGEMENT IN GSM CELLULAR SYSTEM
May 2011 Department of Electrical, Electronics & Telecommunication Engineering Faculty of Science & Engineering Dhaka International University Dhaka, Bangladesh
H ANDO VE R M A NAG E M E NT I N GSM CELLULAR SYSTEM This dissertation is submitted to the Department of Electrical, Electronics & Telecommunication Engineering of Dhaka International University, Bangladesh, for partial fulfillment of requirements for the degree of B.Sc. in Electrical, Electronics & Telecommunication Engineering.
Submitted by Rakibul Hasan Piyal Halder Shahidul Islam Syed Maruf Hossion
(Roll# 01; Registration# 204201) (Roll# 03; Registration# 204203) (Roll# 04; Registration# 204204) (Roll# 05; Registration# 204146) Supervised by Ashraful Arefin
Lecturer Department of Electrical, Electronics & Telecommunication Engineering Dhaka International University
May 2011 Department of Electrical, Electronics & Telecommunication Engineering Faculty of Science & Engineering Dhaka International University Dhaka, Bangladesh.
Declaration We, Rakibul Hasan, Piyal Halder, Shahidul Islam and Syed Maruf Hossion, the students of B.Sc. in Electrical, Electronics and Telecommunication Engineering, hereby solemnly declare that, the works presented in this project and thesis has been carried out by us and have not previously been submitted to any other University/College/Organization for any academic qualification/ certification/ diploma/ degree.
We warrant that the present work does not break any copyright.
Rakibul Hasan
Piyal Halder
Shahidul Islam
Syed Maruf Hossion
i
Certification The thesis titled
Handover Management in GSM Cellular System Submitted by Rakibul Hasan (Roll # 01; Registration # 204201), Piyal Halder (Roll # 03; Registration # 204203), Shahidul Islam (Roll # 04; Registration # 204204), Syed Maruf Hossion (Roll # 05; Registration # 204146), students of B.Sc. Engineering has been accepted as satisfactory in partial fulfillment for the degree of Bachelor of Electrical, Electronics & Telecommunication Engineering on May 2011.
BORD OF EXAMINERS Prof. Dr. Md. Sana Ullah Dean Faculty of Science & Engineering Dhaka International University Chairman
Saiful Islam Chairman (Acting) Dept. of EETE Dhaka International University Member
Ashraful Arefin Lecturer Dept. of EETE Dhaka International University Supervisor
Prof. Dr. Adnan Kibar Dept. of APECE University of Dhaka External Member
ii
Table of Contents Page No. Declaration ………………………………………………………………………...
i
Certification ………………………………………………………………………..
ii
Table of Contains ………………………………………………………………...
iii-v
Acknowledgement …………………………………………………………………
vi
Abstract ……………………………………………………………………………
vii
List of Figures ……………………………………………………………………
viii
Abbreviations ……………………………………………………………………
ix-x
Chapter 1 Introduction to GSM
1- 11
1.1. GSM Architecture ………………………………………………………...
2
1.2. Open interfaces of GSM …………………………………………………..
3
Mobile Station (MS) ……………………………………….
5
1.3. Subsystems and network elements in GSM ………………………………
6
Network Switching Subsystem (NSS) …………………….
6
1.2.1.
1.3.1.
1.3.2.
1.3.3.
1.3.1.1.
Mobile services Switching Centre (MSC)…
7
1.3.1.2.
Visitor Location Register (VLR)…………...
7
1.3.1.3.
Home Location Register (HLR)……………
7
1.3.1.4.
Authentication Centre (AC)…………………….
8
1.3.1.5.
Equipment Identity Resister (EIR)…………
8
Base Station Subsystem (BSS) …………………………….
8
1.3.2.1.
Base Station Controller (BSC)………………….
9
1.3.2.2.
Base Transceiver Station (BTS)………………..
9
1.3.2.3.
TC(SM) − Transcoder (Sub Multiplexer)………
9
Network Management Subsystem (NMS)…………………
10
iii
Page No. Chapter 2 Concept of Cellular Network
12-29
2.1. Introduction ………………………………………………………………...
12
2.2. Basic Concept ……………………………………………………………...
12
2.3. Shape of a Cell ……………………………………………………………..
16
2.4. Most Efficient Cell Shapes to Cover Large Regions ………………………
18
2.5. Location Management ……………………………………………………...
19
2.6. Hand-Off Strategies and Channel Assignment …………………………….
22
2.7. Evaluation of Cellular Network ……………………………………………
25
2.8. Alternatives to Cellular Network …………………………………………..
28
Chapter 3 Overview of Handover
30-40
3.1. Introduction ………………………………………………………………...
30
3.2. Handover Initiation ………………………………………………………...
30
3.2.1.
Relative Signal Strength …………………………………….
31
3.2.2.
Relative Signal Strength with Threshold …………………...
31
3.2.3.
Relative Signal Strength with Hysteresis …………………...
31
3.2.4.
Relative Signal Strength with Hysteresis and Threshold …...
32
3.3. Measurements - a prerequisite for handover ……………………………….
32
3.4. Handover cases …………………………………………………………….
33
3.4.1.
Activation of new channel ………………………………….
34
3.4.2.
Handover command ………………………………………...
34
3.4.3.
Handover bursts …………………………………………….
34
3.4.4.
Handover complete …………………………………………
34
3.4.5.
Release of old channel ……………………………………...
34
3.5. Inter- BSC Handover ………………………………………………………
34
3.5.1(a).
Handover request …………………………………………...
35
3.5.1(b).
Activation of new channel ………………………………….
35
3.5.2.
Handover command ………………………………………...
35
iv
Page No. 3.5.3.
Handover bursts …………………………………………….
35
3.5.4.
Handover complete …………………………………………
36
3.5.5.
Release of old channel ……………………………………...
36
3.6. Inter- MSC Handover ………………………………………………………
36
3.6.1(a).
Handover request …………………………………………...
36
3.6.1(b).
Activation of new channel ………………………………….
37
3.6.2.
Handover command ………………………………………...
37
3.6.3.
Handover bursts …………………………………………….
37
3.6.4.
Handover complete …………………………………………
37
3.6.5.
Release of old channel ……………………………………...
37
3.7. Handover Types ……………………………………………………………
38
3.7.1.
Hard vs. Soft Handover ……………………………………..
38
3.7.2.
Microcellular vs. Multilayer Handover ……………………..
38
3.7.3.
Horizontal vs. Vertical Handover …………………………..
40
Chapter 4 Mechanism & Analysis of Handover Management
41-51
4.1.Introduction ………………………………………………………………...
41
4.2.Conventional Handover Mechanism ……………………………………….
41
4.3.Channel Carrying Handover Mechanism …………………………………..
42
4.4.GSM Handover Prioritization Schemes ……………………………………
43
4.4.1.
Guard Channel Prioritization Scheme ………………………
43
4.4.2.
Call Admission Control Prioritization Scheme ……………..
45
4.4.3.
Handover Queuing Prioritization Schemes …………………
46
4.4.4.
Cell Overlapping and Load Balancing Scheme (Proposed
Scheme) ……………………………………………………………..
48
4.5. Analysis …………………………………………………………………...
50
Chapter 5 Conclusion ……………………………………………………………..
52-53
References ………………………………………………………………………….
54-55
v
Acknowledgement
We are very grateful to all the people that have supported us during the work with this thesis. First of all, we would like to thank our thesis supervisor Mr. Ashraful Arefin for the support and advice during the work with this thesis.
And also we are grateful to Prof. Dr. Md. Sana Ullah, Dean, Faculty of Science & Engineering, who always responded to various problems during the research and never hesitated to offer guidance. We would also like to thank Mr. Saiful Islam, Chairman, Department of EETE, Dhaka International University, for the kind help that he extended to us. We would also like to thank other faculty members of the Department of EETE.
Special thanks to all of our friends and all other concerned authority for all they are rendering during the course of research work.
Rakibul Hasan Piyal Halder May, 2011 Dhaka
Shahidul Islam Syed Maruf Hossion
vi
Abstract
Handover mechanism is extremely important in cellular network because of the cellular architecture employed to maximize spectrum utilization. Handover is the procedure that transfers an ongoing call from one cell to another as the user moves through the coverage area of cellular system. One way to improve the cellular network performance is to use efficient handover prioritization schemes when user is switching between the cells. In this thesis an analytical framework has been presented that can enhance considerably the handover call mechanism in wireless network. Some advance schemes namely, guard channels, call admission control and handover queuing are discussed. All these of prioritizations schemes have a common characteristic and that is, reducing the call dropping probability at the expense of increased call blocking probability. Efficient prioritization scheme accommodates a number of new calls while guarantees the quality of service (QoS) of handover call. This idea is based on the neighboring cells have an overlapping (the area served by more than one cell) coverage area. Furthermore cell overlap and load balancing scheme is proposed to enhance the GSM cellular capacity using an overlapping coverage area. Capacity enhancement is achieved by balancing the load in neighboring cells.
vii
List of Figures: Figure No.
1.1. 1.2. 1.3. 1.4. 1.5. 1.6. 1.7. 2.1. 2.2. 2.3. 2.4. 2.5. 2.6. 2.7. 2.8. 3.1. 3.2. 3.3. 3.4. 3.5. 4.1. 4.2. 4.3. 4.4. 4.5. 4.6. 4.7.
Figure Name Information required by a mobile communications network The three subsystems of GSM and their interfaces Inserting a SIM card in a mobile phone The Network Switching Subsystem (NSS) The Base Station Subsystem (BSS) Location of Transcoder and Sub multiplexer The NMS and the GSM network
Typical cellular network Frequency reuse Radiation pattern Intersection area of a cell. Intersection area of two cells Intersection area for six cells.
Cell shapes A service area with three location area Movement of a MS in the handover zone Intra-BSC Handover Inter-BSC Handover Inter-MSC Handover A city segment with three BSs deployed on streets Signal Levels for Handover r and (r+1) Channel Carrying Guard Channels for Handover Request State Transition Diagram of Guard Channels Priority Queue System Model for Handover Call State Transition Diagram Areas A, B and C of three Cells
viii
Page No. 3 5 5 6 9 10 11 13 14 16 17 17 18 19 21 31 33 35 36 38 42 42 44 44 46 47 49
Abbreviations:
2G
The Second Generation.
3G
The Third Generation.
BS
Base Station.
BSC
Base Station Controller.
BSS
Base Station Subsystem.
CAC
Call Admission Control.
CDMA
Code Division Multiple Access.
CEPT
Conference of European Post and Telegraphs.
DCRS
Dynamic Channel Reservation Scheme.
ETSIE
European Telecommunication Standards Institute.
FIFO
First In First Out.
FSS
Fully Shared Scheme.
GCS
Guard Channel Scheme.
GSM
Group Special Mobile / Global System for Mobile Communications.
HLR
Home Location Register.
HO
Handover.
HQS
Hand-off Queuing Scheme.
IEEE
Institute of Electrical & Electronic Engineering.
ISDN
Pen-European means European-wide.
ME
Mobile Equipment.
MS
Mobile Station.
MSC
Mobile services Switching Centre.
MTSO
Mobile Telephone Switching Office.
NMS
Network Management Subsystem. ix
NMT
Nordic Mobile Telephone System.
NSS
Network Switching Subsystem.
PSTN
Public Switched Telephone Network.
RSS
Received Signal Strengths.
SCN
Single-hop Cellular Network.
SDCCH
Stand Alone Dedicated Control Channel.
SIM
Subscriber Identity Module.
TACS
Total Access Communication System.
TCH
Traffic channel.
TDMA
Time Division Multiple Access.
VLR
Visitor Location Register.
x
Chapter 1
Introduction to GSM During the early 1980s, analog cellular telephone systems were experiencing rapid growth in Europe, particularly in Scandinavia and the United Kingdom, but also in France and Germany. Each country developed its own system, which was incomplete with everyone else’s in equipment and operation. This was an unwanted situation, because not only was the mobile equipment limited to operation within national boundaries, which in a unified Europe were ever more unimportant. But there was also a very limited market for each type of equipment, so economics of scale and the subsequent savings could not be realized. The Europeans realized this early on and in 1982 the Conference of European Post and Telegraphs (CEPT) formed a study group called the Group Special Mobile (GSM) to study and develop a pan-European public land mobile system. The proposed system had to meet certain criteria: •
Good subjective speech quality
•
Low terminal and service cost
•
Support for international roaming
•
Ability to support handheld terminals
•
Support for range of new services and facilities
•
Spectral efficiency
•
ISDN compatibility
Pen-European means European-wide. ISDN throughput at 64Kbs was never envisioned, indeed, the highest rate of a normal GSM network can achieve is 9.6Kbs. Europe saw cellular service introduced in 1981, when the Nordic Mobile Telephone System or NMT450 began operating in Denmark, Sweden, Finland and Norway in the 450 MHz range. It was the first multinational cellular system. In 1985 Great Britain was started using Total Access Communication System or TACS at 900 MHz range. Later, the West German C-Netz, the
Page 1
French Radiocom 2000 and the Italian RTMI/RTMS helped make up Europe’s nine analog incompatible radio telephone systems. Plans were afoot during the early 1980’s however, to create a single European digital mobile service with advanced features and easy roaming. While North American group concentrated on building out robust but increasingly fraud plagued and featureless analog network, Europe planed for a digital future. In 1989, GSM responsibility was transferred to the European Telecommunication Standards Institute (ETSI) and phase I of the GSM specification were published in 1990. Commercial service was started in mid 1991 and by 1993 there were 36 GSM network in 22 countries. Although standardized in Europe, GSM is not only a European standard. Over 200 GSM network are operation in 110 countries around the world. In the beginning 1994, there were 1.3 million subscriber worldwide, which have grown to more than 55 million by October 1997. With North America making a delayed entry into GSM filed with a derivative of GSM called PCS1900, GSM system exist on every continent and the acronym GSM now aptly stands for Global System For Mobile Communications. According to the GSM Association as of 2002, here are current GSM statistics: •
No. of Countries/Areas with GSM system (October 2001) – 172
•
GSM total subscriber – 590.3 million (to end of September 2001)
•
World subscriber growth – 800.4 million (to end of July 2001)
•
SMS messages send per month – 23 Billion (to end of September 2001)
•
SMS forecast to end December 2001 – 30 Billion per month
•
GSM account for 70.7% of the world’s digital market and 64.6% of the world’s wireless market.
1.1 . GSM Architecture The communication between two people – the caller and the called person – is the basic service of all telephone networks. To provide this service, the network must be able to set up and maintain a call, which involves a number of tasks: identifying the called person, determining the location, routing the call and ensuring that the connection sustained as long as the conversation lasts. After the transaction, the connection is terminated and (normally) the calling user is charged for the service he has used.
Page 2
In a fixed telephone network, providing and managing connections is a relatively easy process, because telephones are connected by wires to the network and their location is permanent from the networks’ point of view. In a mobile network, however, the establishment of a call is a far more complex task, as the wireless (radio) connection enables the users to move at their own free will − providing they stay within the network's service area. In practice, the network has to find solutions to three problems (shown in Figure 1.1) before it can even set up a call:
Figure 1.1: Information required by a mobile communications network.
In other words, the subscriber has to be located and identified to provide him/her with the requested services. In order to understand how we are able to serve the subscribers, it is necessary to identify the main interfaces, the subsystems and network elements in the GSM network, as well as their functions [1].
1.2. Open Interfaces of GSM One of the main purposes behind the GSM specifications is to define several open interfaces, which then limit certain parts of the GSM system. Because of this interface openness, the operator maintaining the network may obtain different parts of the network from different GSM network suppliers. When an interface is open, it also strictly defines what is happening through
Page 3
the interface, and this in turn strictly defines what kind of actions/procedures/functions must be implemented between the interfaces. The GSM specifications define two actual open interfaces within the GSM network. The first one is between the Mobile Station (MS) and the Base Station (BS). This open-air interface is appropriately named the “air interface”. It is relatively easy to imagine the need for this interface to be open, as mobile phones of all different brands must be able to communicate with GSM networks from all different suppliers The second interface is located between the Mobile services Switching Centre, MSC, (which is the switching exchange in GSM) and the Base Station Controller (BSC). This interface is called the “A-interface”. The system includes more than the two defined interfaces, but they are not totally open, as the system specifications had not been completed when the commercial systems were launched. When operating analogue mobile networks, experience has shown that centralized intelligence generates excessive load in the system, thus decreasing the capacity. For this reason, the GSM specification, in principle, provides the means to distribute intelligence throughout the network. Referring to the interfaces, the more complicated the interfaces in use, the more intelligence is required between the interfaces in order to implement all the functions required. In a GSM network, this decentralized intelligence is implemented by dividing the whole network into three separate subsystems: •
Network Switching Subsystem (NSS)
•
Base Station Subsystem (BSS)
•
Network Management Subsystem (NMS)
The actual network needed for establishing calls is composed of the NSS and the BSS. The BSS is responsible for radio path control and every call is connected through the BSS. The NSS takes care of call control functions. Calls are always connected by and through the NSS. BSS the NMS is the operation and maintenance related part of the network and it is needed for the control of the whole GSM network. The network operator observes and maintains network quality and service offered through the NMS [1].
Page 4
Figure 1.2: The three subsystems of GSM and their interfaces.
The three subsystems in a GSM network are linked by the air-, A-, and O&M interfaces as shown in Figure 1.2.
1.2.1. Mobile Station (MS) The MS (Mobile Station) is a combination of terminal equipment and subscriber data. In Figure 1.3 shows, the terminal equipment as such is called ME (Mobile Equipment) and the subscriber's data is stored in a separate module called SIM (Subscriber Identity Module). Therefore, ME + SIM = MS.
Figure 1.3: Inserting a SIM card in a mobile phone
Page 5
1.3. Subsystems and network elements in GSM The GSM network is divided into three subsystems: Network Switching Subsystem (NSS), Base Station Subsystem (BSS), and Network Management Subsystem (NMS). The three subsystems are different network elements and their respective tasks are presented in the following:
1.3.1. Network Switching Subsystem (NSS) In Figure 1.4, the Network Switching Subsystem (NSS) contains the network elements MSC, VLR, HLR, AC and EIR [1].
Figure 1.4: The Network Switching Subsystem (NSS)
The main functions of NSS are: • Call control • Charging • Mobility management • Signaling • Subscriber data handling
Page 6
1.3.1.1. Mobile services Switching Centre (MSC) The MSC is responsible for controlling calls in the mobile network. It identifies the origin and destination of a call (mobile station or fixed telephone), as well as the type of a call. An MSC acting as a bridge between a mobile network and a fixed network is called a Gateway MSC. The MSC is responsible for several important tasks, such as the following. • Call control • Initiation of paging • Charging data collection
1.3.1.2. Visitor Location Register (VLR) Visitor Location Register (VLR) is integrated with the MSC. VLR is a database which contains information about subscribers currently being in the service area of the MSC/VLR, such as: •
Identification numbers of the subscribers
•
Security information for authentication of the SIM card and for ciphering
•
Services that the subscriber can use
1.3.1.3. Home Location Register (HLR) HLR maintains a permanent register of the subscribers, for instance subscriber identity numbers and the subscribed services. In addition to the fixed data, the HLR also keeps track of the current location of its customers. The MSC asks for routing information from the HLR if a call is to be set up to a mobile station (mobile terminated call). The two network elements, Authentication Centre (AC) and Equipment Identity Register (EIR), are located in the HLR [1].
Page 7
1.3.1.4. Authentication Centre (AC) The Authentication Centre provides security information to the network, so that we can verify the SIM cards (authentication between the mobile station and the VLR, and cipher the information transmitted in the air interface (between the MS and the Base Transceiver Station).
1.3.1.5. Equipment Identity Register (EIR) The Equipment Identity Register is used for security reasons. But while the AC provides information for verifying the SIM cards, the EIR is responsible for IMEI checking (checking the validity of the mobile equipment). When performed, the mobile station is requested to provide the International Mobile Equipment Identity (IMEI) number. This number consists of type approval code, final assembly code and serial number of the mobile station. The EIR contains three lists: •
Mobile equipment in the white list is allowed to operate normally.
•
If we suspect that mobile equipment is faulty, we can monitor the use of it. It is then placed in the grey list.
•
If the mobile equipment is reported stolen, or it is otherwise not allowed to operate in the network, it is placed in the black list.
Note that IMEI checking is an optional procedure, so it is up to the operator to define if and when IMEI checking is performed [1]. (Some operators do not even implement the EIR at all.)
1.3.2. Base Station Subsystem (BSS) The Base Station Subsystem (Figure 1.5) is responsible for managing the radio network, and it is controlled by an MSC. Typically, one MSC contains several BSSs. A BSS itself may cover a considerably large geographical area consisting of many cells (a cell refers to an area covered by one or more frequency resources). The BSS consists of the following elements: •
BSC – Base Station Controller
•
BTS – Base Transceiver Station
•
TC – Transcoder
Page 8
Figure 1.5: The Base Station Subsystem (BSS)
1.3.2.1. Base Station Controller (BSC) The BSC is the central network element of the BSS and it controls the radio network. It has several important tasks, some of which are presented in the following: •
Connection establishment between the MS and the NSS
•
Mobility management
•
Statistical raw data collection
•
Air- and A-interface signaling support
•
BTS and TC control
1.3.2.2. Base Transceiver Station (BTS) The BTS is the network element responsible for maintaining the air interface and minimizing the transmission problems (the air interface is very sensitive for disturbances). This task is accomplished with the help of some 120 parameters. These parameters define exactly what kind of BTS is in question and how MSs may "see" the network when moving in this BTS area.
1.3.2.3. TC(SM) − Transcoder (Sub Multiplexer) In the air interface (between MS and BTS), the media carrying the traffic is a radio frequency. To enable an efficient transmission of digital speech information over the air interface, the digital speech signal is compressed. We must however also be able to communicate with and through the fixed network, where
Page 9
the speech compression format is different. Somewhere between the BTS and the fixed network, we therefore have to convert from one speech compression format to another, and this is where the Transcoder comes in (Figure 1.6).
Figure 1.6: Location of Transcoder and Sub multiplexer.
1.3.3. Network Management Subsystem (NMS) The Network Management Subsystem (NMS) is the third subsystem of the GSM network in addition to the Network Switching Subsystem (NSS) and Base Station Subsystem (BSS), which we have already discussed. The purpose of the NMS is to monitor various functions and elements of the network (Figure 1.7). The operator workstations are connected to the database and communication servers via a Local Area Network (LAN). The database server stores the management information about the network. The communications server takes care of the data communications between the NMS and the equipment in the GSM network known as “network elements”.
Page 10
Figure 1.7: The NMS and the GSM network
The functions of the NMS can be divided into three categories: •
Fault management
•
Configuration management
•
Performance management
These functions cover the whole of the GSM network elements from the level of individual BTSs, up to MSCs and HLRs [1].
Page 11
Chapter 2
Concept of Cellular Network 2.1. Introduction Cellular communications has experienced explosive growth in the past two decades. Today millions of people around the world use cellular phones. Cellular phones allow a person to make or receive a call from almost anywhere. Likewise, a person is allowed to continue the phone conversation while on the move. Cellular communications is supported by an infrastructure called a cellular network, which integrates cellular phones into the public switched telephone network. The cellular network has gone through three generations. The first generation of cellular networks is analog in nature. To accommodate more cellular phone subscribers, digital TDMA (time division multiple access) and CDMA (code division multiple access) technologies are used in the second generation (2G) to increase the network capacity. With digital technologies, digitized voice can be coded and encrypted. Therefore, the 2G cellular network is also more secure. The third generation (3G) integrates cellular phones into the Internet world by providing high speed packet-switching data transmission in addition to circuit-switching voice transmission. The 3G cellular networks have been deployed in some parts of Asia, Europe, and the United States since 2002 and will be widely deployed in the coming years. This chapter gives an introduction to cellular networks.
2.2. Basic Concept A cellular network provides cell phones or mobile stations (MSs), to use a more general term, with wireless access to the public switched telephone network (PSTN). The service coverage area of a cellular network is divided into many smaller areas, referred to as cells, each of which is served by a base station (BS). The BS is fixed, and it is connected to the mobile telephone switching office (MTSO), also known as the mobile switching center. An MTSO is in charge of a cluster of BSs and it is, in turn, connected to the PSTN. With the wireless link between the BS and MS, MSs such as cell phones are able to communicate with wire line phones in the PSTN.
Page 12
Both BSs and MSs are equipped with a transceiver. Figure 2.1 illustrates a typical cellular network, in which a cell is represented by a hexagon and a BS is represented by a triangle.
Figure 2.1: Typical cellular network.
The frequency spectrum allocated for cellular communications is very limited. The success of today’s cellular network is mainly due to the frequency reuse concept. This is why the coverage area is divided into cells, each of which is served by a BS. Each BS (or cell) is assigned a group of frequency bands or channels. To avoid radio co-channel interference, the group of channels assigned to one cell must be different from the group of channels assigned to its neighboring cells. However, the same group of channels can be assigned to the two cells that are far enough apart such that the radio co-channel interference between them is within a tolerable limit. Typically, seven neighboring cells are grouped together to form a cluster, as shown in Figure 2.2 The total available channels are divided into seven groups, each of which is assigned to a cell. In Figure 2.2, the cells marked with the same number have the same group of channels assigned to them. Furthermore, the cells marked with different numbers must be assigned different groups of channels.
If there are a total of M channels allocated for cellular communications and if the coverage area consists of N cells, there are a total of MN/7 channels available in the coverage area for concurrent use based on the seven-cell reuse pattern. That is the network capacity of this coverage area. Because of explosive growth of mobile phone subscribers, the current network capacity may not be enough. Cell splitting is one technique that used to increase the network
Page 13
capacity without new frequency spectrum allocation. In this technique, the cell size is reduced by lowering antenna height and transmitter power. Specifically, an original cell is divided into four smaller cells. After cell-splitting, the coverage area with N cells originally will be covered by 4N smaller cells. Therefore, the new network capacity is 4MN/7, which is four times the original network capacity. In reality, bigger cells are not completely replaced by smaller cells. Therefore, cells of different sizes (e.g., Pico, micro, and macro cells) may coexist in one area. This allows high-speed subscribers to use bigger cells, which reduces the number of hand-over.
Sectoring is another technique to increase the network capacity. In sectoring, the cell size remains the same, but a cell is divided into several sectors by using several directional antennas at the BS instead of a single Omni-directional antenna. Typically, a cell is divided into three 120o sectors or six 60o sectors. The radio cochannel interference will be reduced by dividing a cell into sectors, which reduces the number of cells in a cluster. Therefore, the network capacity is increased. Digital technology can also be used to increase the network capacity. Transmission of digitized voice goes through three steps before the actual transmission.
Figure 2.2: Frequency reuse
Speech coding, channel coding, and modulation. Speech coding is to compress voice. For example, a short voice segment can be analyzed and represented by a few parameter values. These values cannot be transmitted directly because wireless transmission is error prone, and a small change in these values may translate into a big change in voice. Therefore, data representing compressed voice should be arranged carefully, and redundancy should be Page 14
introduced such that a transmission error can be corrected or at least detected. This process is called channel coding. Finally, the output data from channel coding are modulated for transmission. A good speech-coding scheme combined with a good channel coding scheme will greatly reduce the amount of bandwidth needed by each phone user and therefore increase the network capacity while keeping the quality of voice unchanged.
The channels assigned to a cell are used either for voice or for control. A voice channel is used for an actual conversation, whereas a control channel is used to help set up conversations. Both voice and control channels are further divided into forward (or downlink) and reverse (or uplink). A forward channel is used to carry traffic from the BS to the MS, and a reverse channel is used to carry traffic from the MS to the BS. The channels assigned to a cell are shared by MSs located in the cell. Multiple access methods are used to share the channels in a cell.
Each MS has a home MTSO, which is the MTSO where the mobile user originally subscribed for wireless services. If an MS moves out of the home MTSO area, it is roaming. A roaming MS needs to register in the visited MTSO. An MS needs to be authenticated against the information kept in its home MTSO before any service can be rendered by the network. The services include making a call, receiving a call, registering the location, and so forth. These services are possible because of a widely used global, common channel-signaling standard named SS7 (Signaling System 7).
To make a call from an MS, the MS first needs to make a request using a reverse control channel in the current cell. If the request is granted by MTSO, a pair of voice channels (one for transmitting and the other for receiving) is assigned for the call. Making a call to an MS is more complicated. The call is first routed to its home MTSO or its visited MTSO if it is roaming. The MTSO needs to know the cell in which the MS is currently located. Finding the residing cell of an MS is the subject of location management. Once the MTSO knows the residing cell of the MS, a pair of voice channels is assigned in the cell for the call.
If a call is in progress when the MS moves into a neighboring cell, the MS needs to get a new pair of voice channels from the BS of the neighboring cell so the call can continue. This process
Page 15
is called handover. A BS usually adopts a channel assignment strategy that prioritizes hand-over calls from neighboring cells over the new calls initiated in the current cell [2].
2.3. Shape of a Cell •
We will assume that cellular towers use Omni-directional antennas that are installed in a vertical form. Such antennas produce the radiation pattern shown previously and is repeated here for convenience. This radiation pattern indicates that (Figure 2.3) equal power is radiated horizontally in all directions around the antenna (0° to 360°). Vertically, the radiated power has maximum power on the plain parallel to the earth surface. As the observation point goes above or below this plain, the radiated power drops. In fact, no power is radiated above or below the antenna directly [3].
Figure 2.3: Radiation pattern •
Assuming that no geographical features or buildings exist around the antenna and that only one cellular tower exist in the region, the radiation pattern of the Omni‐directional antenna produces circles with equal power at ground‐level as shown in the figure 2.4. The power drops as the radius of the circles increases. It may appear that no radiation will reach the region very close to the cellular tower because of the fact that the antenna is installed on top of a tower. But remember that distances away from the antenna are measure in kilometers while the height of the antenna usually around 10m – 30m only.
Page 16
Figure 2.4: Intersection area of a cell. •
As distance between the cellular phone and the tower increases the signal drops quickly until at some distance from the tower, the signal transmitted by the tower becomes so weak that the cellular phone can no longer demodulate the signal. At that distance, a mobile phone wills indicate that there is no service.
•
When two towers are close to each other, the transmitted power from each one enters the coverage area of the other. A cellular phone will connect to the cellular tower from which it receives a stronger signal. Assuming that both towers transmit equal power, the curve at which equal power is received from both towers is a straight line as shown Figure 2.5.
Figure 2.5: Intersection area of two cells. Page 17
•
When multiple towers are close to each other, a cell phone will communicate with the tower that provides it with the highest power. The curves that separate different regions belonging to different towers based on the highest power criterion are straight lines as shown in the following Figure 2.6.
Figure 2.6: Intersection area for six cells.
•
In practice, cell boarders are never straight lines because of uneven power transmission by different towers, uneven earth surface, the existence of geographical features, trees, or buildings make the cell barriers random in shape. In fact, cell phone companies usually test cell barriers by having some drive around with sophisticated cell phones to determine were cell barriers are located for proper planning [3].
2.4. Most Efficient Cell Shapes to Cover Large Regions The question we would like to answer now is: What is the most efficient theoretical cell shape that will allow us to provide full coverage to a large area by stacking the minimum number of cells possible [3].
Ideally, a circle is the best shape because it is the natural reach of electromagnetic waves when they are transmitted from an Omni‐directional antenna. However, it is not possible to stack circles near each other and cover the whole region of interest without leaving some gaps [3]. For proper cell shapes, let us observe the following points:
Page 18
•
Boarders of cells are straight lines and cell shapes are polygons (Polygons are geometric shapes with all edges being straight lines like triangles, rectangles, pentagon and so on which is shown in Figure 2.7).
•
Full coverage of the whole region is necessary without leaving any uncovered spots.
•
We will assume that all cells have the same shape.
•
Cells should have some symmetry (cells can be rotated in place at angles less than one complete rotation without affecting cells layout)
Cell shapes that meet the above constraints are one of the following:
a) Equilateral Triangles
b) Square
c) Hexagons
Figure 2.7: Cell shapes. •
Equilateral Triangles: An equilateral triangle is one with all sides having equal lengths and all angles being 60°.
•
Hexagons: A shape with six equal sides and six equal angles.
2.5. Location Management This section describes how to track an MS. How to keep track of an MS is the subject of location management in cellular networks. Because the exact location of an MS must be known to the cellular network when a call is in progress, location management tracks an active MS that is not in a call. (An MS is active when it is powered on.) Specifically, the cellular network needs to find out the exact cell in which an MS is located when an incoming call arrives for the MS.
An extreme case, known as “Never-Update” in Bar-Noy, Kessler, and Sidi (1995), is that an MS never tells the cellular network its location when it moves around. When an incoming phone call
Page 19
arrives for the MS, the cellular network needs to page all cells in the service area to find out the cell in which the MS is currently located so the incoming call can be routed to the BS of that cell. It will cost a great deal to page all cells in the service area. The other extreme case is known as “Always-Update,” in which an MS needs to update its location whenever it moves into a new cell. When an incoming phone call arrives for the MS, the cellular network can just route the incoming call to the cell that is last reported by the MS. Obviously, there is no paging cost involved. However, the MS needs to tell the cellular network its location when it moves from cell to cell, which can also be very expensive.
There are two basic operations involved with location management: location update and paging. The paging operation is performed by the cellular network. When an incoming call arrives for an MS, the cellular network will page the MS in all possible cells to find out the cell in which the MS is located so the incoming call can be routed to the corresponding BS. The number of all possible cells to be paged is dependent on how frequent the location update operation is performed. The location update operation is performed by an MS. Both operations consume wireless bandwidth: location update uses reverse control channels, whereas paging utilizes forward control channels. Although the cost of location management involves both the wire line portion and the wireless portion of the cellular network, only the wireless portion is usually considered. This is mainly because the radio-frequency bandwidth is limited, whereas the bandwidth of the wire line network is easily expandable.
Therefore, the location management cost is measured by the total wireless bandwidth consumed by location update and paging operations, that is, the location update cost and the paging cost. There is a trade-off between the location update cost and the paging cost. For example, if an MS updates its location more frequently, the network knows the location of the MS better. It incurs a higher location update cost, but the paging cost will be lower when an incoming call arrives for the MS.
The location-areas scheme is used in current cellular networks. In the location-areas scheme, the whole service coverage area is partitioned into location areas, each of which consists of several contiguous cells. The BS of each cell broadcasts the ID of the location area to which the cell
Page 20
belongs. An MS knows the location area it is in by listening to the broadcasting from the BS. An MS updates its location (i.e., location area) whenever it moves into a cell that belongs to a new location area. When an incoming call arrives for an MS, the cellular network pages all the cells of the location area that was last reported by the MS. Figure 2.8 illustrates a coverage area with three location areas separated by wide lines.
Figure 2.8: A service area with three location area.
When an MS moves from cell A to cell B in the figure, it will report its new location area because cell A and cell B are in different location areas. No location update is necessary if the MS moves from cell A to cell C because cell A and cell C are in the same location area.
The location-areas scheme is a generalization of both Never-Update and Always-Update extreme cases. In Never-Update, all the cells in the coverage area belong to the same location area, and in Always-Update, each cell forms a location area. How to divide the whole coverage area into locations greatly affects the location management cost. In reality, a location area consists of all the cells under the control of an MTSO. Each MTSO consists of two main databases: HLR (Home Location Register) and VLR (Visitor Location Register). HLR contains information for each MS that considers the location area covered by the MTSO as its home. Each MS can subscribe to one MTSO as its home. VLR records a list of the MSs that are currently located in the location area covered by the MTSO but do not consider the MTSO as their home.
Page 21
When an MS moves out of its home MTSO and visits another MTSO, the MS needs to register with the visited MTSO. The visited MTSO contacts the home MTSO about the MS. The home MTSO records the visited MTSO of the MS in its HLR. The visited MTSO updates its VLR about the MS. When a call is made to the MS, it is first addressed to the home MTSO of the MS. By checking the information stored in HLR about the MS, the call can eventually be routed to the visited MTSO. Finally, the visited MTSO pages the MS in the cells of its covered location area.
The location-areas scheme is global in the sense that all MSs transmit their location updates in the same set of cells, and it is static in the sense that the location areas are fixed. A locationupdate scheme can be classified as either global or local. A location-update scheme is global if all MSs update their locations at the same set of cells, and a scheme is local (or individualized) if an individual MS is allowed to decide where to perform location update. From another point of view, a location-update scheme can be classified as either static or dynamic. A location-update scheme is static if there is a predetermined set of cells at which location updates must be generated by an MS regardless of its mobility. A scheme is dynamic if a location update can be generated by an MS in any cell depending on its mobility. Recently there was a swarm of research on location management, especially how to reduce the cost of location management [2].
2.6. Hand-Off Strategies and Channel Assignment The previous section described how to track the movement of an MS when it is not in a call. This section deals with the movement of an MS when it is in a call. When an MS is in a call, it has acquired two channels (one for transmitting and the other for receiving) from the current cell for communication with the BS. When the MS moves out of the current cell and enters a neighboring cell, the MS needs to acquire two channels from the neighboring cell in order for the call to continue. The process of transferring a call from the current cell to a neighboring cell is called hand-off. To a cell, a hand-off call is a call that is in progress in a neighboring cell and needs to be continued in the cell because of the movement of an MS. In contrast, a new call is a call that is started in the cell.
Page 22
As mentioned earlier, the number of channels assigned to each cell is limited. Channel assignment deals with how to assign available channels to new calls as well as hand-off calls. The simplest channel assignment scheme is the fully shared scheme (FSS), in which all available channels are shared by hand-off calls and new calls. No distinction is made between a hand-off call and a new call. FSS is widely used in the current cellular networks because of its simplicity. In addition, FSS has the advantage of maximizing the utilization of wireless channels. The disadvantage is the increased dropping rate of hand-off calls. In general, it is less desirable to drop a hand-off call than to block a new call. The dropping probability of hand-off calls is considered as one of major metrics that measure the quality of service of calls.
Recently intensive research on channel-assignment schemes has been conducted to decrease the dropping probability of hand-off calls. One such scheme is the hand-off queuing. When an MS detects that the received signal strength from the current BS is below a certain level, called the hand-off threshold, a hand-off operation is initiated. The hand-off operation first identifies the new BS into which the call is moving. If the new BS has unused channels, the call will be transferred to the new BS. If there is no unused channel available, the hand-off call will be queued until a channel is released by another call. The HQS scheme is feasible because there is a difference between the signal strength at the hand-off threshold and the minimum acceptable signal strength for voice communication. This gives an MS some time to wait for a channel at the new BS to become available. A new call will be blocked in the new cell until all the hand-off calls in the queue are served. Therefore, the HQS scheme decreases the dropping probability of hand-off calls while increasing the blocking probability of new calls because the scheme gives higher priority to hand-off calls.
When a hand-off call is not able to acquire the necessary channels in the new cell, the call is allowed to carry the channels in the current cell to the new cell, a concept the authors called channel carrying. However, carrying the channels from the current cell to the new cell may reduce the reuse distance of the channels and violate the minimum reuse distance requirement. To ensure the minimum reuse distance requirement is not violated, an (r + 1)-channel assignment scheme is used. That is, the same channels are reused exactly (r + 1) cells apart.
Page 23
Here r is the minimum reuse distance. In this scheme, a channel can only be carried from the assigned cell to a neighboring cell, and the carried channel will be returned as soon as a local channel is available. By using channel carrying, a hand-off call can continue, even if there is no channel available in the new cell. Therefore, the dropping probability of hand-off calls will be reduced. However, the capacity of the cellular network is reduced because the (r + 1)-channel assignment scheme is used instead of r-channel assignment.
Another channel assignment scheme that prioritizes the hand-off call is referred to as the guard channel scheme (GCS). In GCS, each BS reserves a fraction of wireless channels exclusively for hand-off calls, and the remaining channels, called normal channels, are shared between hand-off calls and new calls. Both hand-off calls and new calls use the normal channels first. When the normal channels are used up, a new call will be blocked, but a hand-off call can still use the reserved channels. In this way, the dropping probability of hand-off calls will be reduced. The improvement in the dropping probability of hand-off calls is dependent on the number of channels reserved. However, a new call is blocked if there are only reserved channels left even though no hand-off calls exist. Therefore, the total utilization of wireless channels is decreased. There is a tradeoff between decreasing the blocking probability of hand-off calls and increasing the total channel utilization. The number of channels that should be set aside for hand-off calls depends on a lot of factors such as the mobility of MSs, the call duration, and so forth.
Kim propose a dynamic channel reservation scheme (DCRS) based on mobility. Their goal is to guarantee the required dropping probability of hand-off calls while minimizing the blocking probability of new calls. As in GCS, the normal channels of DCRS are shared by both hand-off calls and new calls, and the guard channels are reserved for hand-off calls. However, the guard channels can also be used by new calls whose request probability is a function of the mobility of calls. The mobility of calls in a cell is defined as the ratio of the hand-off call arrival rate to the new call arrival rate. If there is no arrival of hand-off calls, the request probability will be one, and the guard channels will be used by new calls. If there is no arrival of new calls, the request probability will be zero, and the guard channels will be used by hand-off calls. When the arrival rate of hand-off calls is larger than that of new calls, the request probability is decreased quickly so the hand-off calls can use the guard channels. In this way, the dropping probability of handoff Page 24
calls is guaranteed. When the arrival rate of hand-off calls is lower than that of new calls, the request probability is decreased slowly so new calls have a chance to use available guard channels. This decreases the blocking rate of new calls without sacrificing the dropping probability of hand-off calls, which increases the total channel utilization [2].
2.7. Evaluation of Cellular Network Cellular systems became popular because of radio-frequency reuse, which allows more cell phone users to be supported. The cellular concept was first used in the AMPS in the United States. As a first generation of cellular systems, AMPS is a FDMA-based analog system. The 2G of cellular systems uses digital technologies. Two interim standards, IS-95 (CDMA-based) and IS-136 (TDMA based), are used in the United States, and TDMA-based GSM is used in European countries. It is clear that the 3G of cellular systems will be CDMA-based. However, the GSM community is developing WCDMA to be backward compatible with GSM while the CDMA community tries to evolve CDMA into CDMA2000. Currently researchers are studying technologies for beyond 3G (B3G) or fourth generation (4G) networks.
In the late 1970s and early 1980s, AT&T Bell Laboratories developed the AMPS, which was the first-generation cellular system used in the United States. It was first deployed in Chicago and Washington, DC, then in all the major U.S. cities. Currently it is still used in many rural areas. The Federal Communications Commission (FCC) initially allocated a 40-MHz spectrum in the 800-MHz band for the AMPS in 1983, and later in 1989 added an additional 10 MHz to accommodate the increasing demand for cellular phone services. AMPS are FDMA-based, with each channel occupying a narrow band of 30 kHz. AMPS are an analog system. It transmits 3kHz voice signal over the 30-kHz channel using frequency modulation. IS-41 was originally developed to support the operations with AMPS. However, as an analog system, AMPS does not support voice encryption.
To overcome the limited capacity of AMPS, especially in large cities, D-AMPS (IS-54) was developed in the early 1990s (EIA/TIA, 1990). D-AMPS inherited a lot of features from AMPS. Specifically, in D-AMPS, the same AMPS allocation of frequency spectrum is used, and each channel is still 30 kHz wide. However, in D-AMPS, a 30- kHz channel can be shared by three
Page 25
users through the 2G TDMA digital technology. In a typical D-AMPS cell, some of the 30-kHz channels are assigned for analog AMPS traffic, whereas the others are for digital TDMA traffic. It means that D-AMPS allow a service provider to migrate from the first-generation analog technology to the 2G digital technology on a gradual basis. In a less densely populated area, all the channels can be assigned for AMPS traffic. When the demand increases, some channels will be converted from analog to digital.
In a densely populated area, all the channels need to be assigned for TDMA traffic to meet the demand. IS-136, another prominent TDMA-based cellular system in the United States, is built on D-AMPS. Whereas DAMPS provides dual-mode operations (both analog and digital), IS-136 provide pure digital operations. All the 30- kHz channels are shared by three users via TDMA digital technology. In addition, unlike D-AMPS, IS-136 also uses the digital control channels. IS-136 was initially developed on the 800-MHz cellular spectrum. It can be adopted onto the 1900-MHz PCS spectrum.
GSM is a 2G system developed to solve the incompatibility problem of different first-generation systems in Europe. It is now widely deployed around the world including in the United States. GSM was first developed for Europe in the 900-MHz band (GSM 900), and then expanded to the 1800-MHz band (1710–1880 MHz), which is named DCS 1800, and later renamed to GSM 1800. The North America version of GSM is called PCS 1900 because of its use of the 1900MHz PCS spectrum. GSM uses the TDMA digital technology. The allocated spectrum is divided into multiple channels of 200 kHz using FDMA, and each 200- kHz channel is shared by as many as eight users using TDMA. One feature of GSM worth mentioning is the SIM card that can be inserted into a cellular phone to provide the owner’s identity information. A cell phone without a SIM card inserted does not work. A SIM card can be inserted into any cell phone to make the phone usable.
Whereas IS-54, IS-136, and GSM are all TDMA-based, IS-95 is CDMA-based (EIA/TIA, 1995). As mentioned earlier, each user in CDMA is assigned a unique code to encode the data to be transmitted. Knowing the code of the transmitter, the receiver is able to recover the original data from the received data. CDMA is a very new 2G digital technology. Since its first launch in
Page 26
1995, CDMA quickly became one of the world’s fastest-growing wireless technologies. CDMA uses channels that are 1.25 MHz wide, and it is able to support up to 64 users with orthogonal codes. With CDMA, the same channel can be reused in a neighboring cell. CDMA is superior to FDMA and TDMA.
In fact, CDMA provides roughly 10 times more capacity than analog systems, whereas TDMA provides 3 to 4 times more capacity than analog systems. In 1999, CDMA was selected by the International Telecommunications Union as the industry standard for new 3G cellular systems. The goal of a 3G cellular system is to provide all kinds of services: voice, high-speed data, audio/video, and so forth. The high-speed data transmission is the main development focus. The CDMA and GSM communities are two major players in this effort. The 3G path adopted by the GSM community is first to GPRS, then to EDGE, and ultimately to WCDMA. Currently GSM provides data services of 9.6 Kbps using a single TDMA channel. Although multiple TDMA channels can be combined to provide high-speed data service, it is circuit switched. A service called general packet radio service (GPRS) is first developed to allow users to connect to packetswitched data networks via a different connection from the voice network. With GPRS, the raw data rate increases to approximately 170 Kbps. The next step is enhanced data rates for global evolution (EDGE).
EDGE provides practical raw data rates of up to 384 Kbps using a new high-speed physical layer. Finally, WCDMA is used for the 3G version of the GSM community. In WCDMA, CDMA is used instead of TDMA, and the carrier bandwidth jumps from 200 kHz to 5 MHz to provide data rates of up to 2 Mbps. However, new frequency allocations, BSs, and MSs are required because of the change from TDMA to WCDMA.
The 3G path adopted by the CDMA community is first to CDMA2000 1x, then to 1xEV, and ultimately to 3x. The first step is to use one 1.25-MHz carrier to support packet data. In fact, CDMA phones and networks are already capable of handling packet data. The CDMA2000 1xEV consists of two phases. In Phase 1, one carrier (1.25 MHz) is dedicated to high-speed packet data, and one or more additional carriers are used for voice. In Phase 2, packet data and
Page 27
voice can be combined in the same carrier. Finally, CDMA200 3x can use up to three 1.25-MHz carriers.
When CDMA2000 uses three 1.25-MHz carriers, its total bandwidth approaches that of WCDMA. However, 3x is more flexible because three channels can be used independently or together as a single 3.75-MHz channel. The main advantage of the 3x approach is that it can be implemented in existing frequency allocations in CDMA, which also uses 1.25-MHz carriers. Now researchers are developing technologies for B3G or 4G networks. It is expected all the 4G network elements are digital and the entire network is packet-switched. 4G networks will integrate wireless local area networks (LANS) such as IEEE 802.11 and Bluetooth with wide area cellular networks. The data transmission rate of 4G communications will be much higher than 3G, at 20 to 100 Mbps in mobile mode [2].
2.8. Alternatives to Cellular Network This section describes alternatives to cellular networks. Although cellular networks use either the 800-MHz cellular spectrum or the 1900-MHz PCS spectrum, most of the alternatives use the license-free 2.4-GHz industrial, scientific, and medical (ISM) band. The alternatives introduced here include IEEE 802.11 wireless LAN, IEEE 802.16 wireless MAN (metropolitan area network), mobile ad hoc networks, and multichip cellular networks. IEEE 802.11 wireless LAN is an infrastructure wireless network, similar to a cellular network. Access points (APs) that are analogous to BSs in cellular networks are established to provide an extension to the wired network. The area covered by an AP is called the basic service area, which is analogous to a cell in cellular networks. The nodes in a basic service area form the basic service set. Any node in the basic service set can communicate with the AP directly. Their radio communication uses the license-free ISM band with direct sequence spread spectrum. Under 802.11b (a variation of the IEEE 802.11), the communication is kept at a maximum speed of 11 Mbps whenever possible. It drops back to 5.5 Mbps, then 2 Mbps, and finally down to 1 Mbps if signal strength or interference is corrupting data. New IEEE 802.11g extends the data rate to 54 Mbps from 11 Mbps of IEEE 802.11b. With IEEE 802.11 wireless LAN, a node is connected to the Internet world, and it is able to get any service the Internet provides.
Page 28
IEEE 802.11 works well for wireless access in a local area. To use it for wireless access in a metropolitan area, its bandwidth is often insufficient, and it can experience interference from competitors because they use the same license-free ISM band. IEEE 802.16 standards are for broadband wireless access in a metropolitan area, and they use licensed bands.
IEEE 802.16a is just for fixed broadband wireless access. IEEE 802.16e supports both fixed and mobile broadband wireless access. Therefore, hand-off between two towers is allowed in IEEE 802.16e. Unlike cellular networks, mobile ad hoc networks (MANET for short) are infrastructure less. All nodes within the network can be mobile, and there is no fixed BS or centralized control (Macker & Corson, 1998). A mobile node can communicate with another one directly if one is within the transmission range of the other. Because the network topology for a MANET can dynamically change as a result of node mobility, a MANET is usually modeled by a dynamically changing graph. A MANET finds its applications in an environment where it may not be economically practical or physically possible to provide the necessary infrastructure, or the expediency of the situation may not permit its installation. For example, in a battlefield, it is not possible to install an infrastructure wireless network in the enemy’s territory. In such a situation, a MANET is a good solution. Unlike in the cellular network, the route for communication between two nodes is not fixed in MANET. To send a message from one mobile node (the source) to another mobile node (the destination), a route between the source and the destination must be found. Designing a good routing protocol in a dynamically changing network is very challenging.
The traditional cellular network is called a single-hop cellular network (SCN) because an MS in a cell can reach the corresponding BS with a single wireless hop, and a MANET is considered a multichip wireless network because intermediate nodes are required for communication between two nodes that are not within the transmission range of each other. Lin and Hsu (2000) proposed a new architecture, referred to as multichip cellular network (MCN), which combines the features of SCN and MANETs. MCN has many advantages over SCN. In SCN, two MSs communicate only via a BS even though they are in the same cell and mutually reachable. In MCN, these two MSs can communicate directly. In MCN, if a MS is not reachable from a BS in one hop, the BS will seek intermediate nodes to forward, which is not feasible in SCN [2].
Page 29
Chapter 3
Overview of Handover
3.1. Introduction The handover procedure is used when there is a need for a cell change when the MS is busy. The network is responsible for making the handover decision and performing the actual handover. To assist in the handover decision the MS will provide the network a feedback with measurements made on the downlink. Measurements will also be made on the network side.
Three types of handover can be distinguished depending on the network structure: i) Intra-BSC Handover – an MS move between two cells, belonging to the same BSC. In this case the BSC has full control over the handover.
ii) Inter-BSC Handover – an MS move between two cells belonging to different BSCs under the same MSC/VLR. In this case the “old” BSC will take the decision and initiate the handover.
iii) Inter-MSC Handover – an MS move between two cells belonging to different BSCs under different MSC/VLRs. In this case the “old” BSC will take the decision and initiate the handover. The “old” MSC, called anchor-MSC, and the new MSC together with the new BSC will be parts of the link procedure to commit handover [4].
3.2. Handover Initiation Handover initiation is the process of deciding when to request a handover. Handover decision is based on received signal strengths (RSS) from current BS and neighboring BSs. In Figure 3.1 RSSs of current BS (BS1) and one neighboring BS (BS2) is examined. The RSS gets weaker as MS goes away from BS1 and gets stronger as it gets closer to the BS2 as a result of signal propagation. The received signal is averaged over time using an averaging window to remove momentary fading due to geographical and environmental factors. Below, the four main
Page 30
handover initiation techniques mentioned in relative signal strength, relative signal strength with threshold, relative signal strength with hysteresis, and relative signal strength with hysteresis and threshold will be examined.
Figure 3.1: Movement of a MS in the handover zone 3.2.1. Relative Signal Strength In relative signal strength, the RSSs are measured over time and the BS with strongest signal is chosen to handover. In Figure 3.1 BS2’s RSS exceeds RSS of BS1 at point A and handover is requested. Due to signal fluctuations, several handovers can be requested while BS1’s RSS is still sufficient to serve MS. These unnecessary handovers are known as ping-pong effect. As the number of handovers increase, forced termination probability also increases. So, handover techniques should avoid unnecessary handovers.
3.2.2. Relative Signal Strength with Threshold Relative signal strength with threshold introduces a threshold value (T1 in Figure 3.1) to overcome the ping-pong effect. The handover is initiated if BS1’s RSS is lower than the threshold value and BS2’s RSS is stronger than BS1’s.The handover request is issued at point B in Figure 3.1.
3.2.3. Relative Signal Strength with Hysteresis This technique uses a hysteresis value (h in Figure 3.1) to initiate handover. Handover is requested when the BS2’s RSS exceeds the BS1’s RSS by the hysteresis value h (point C in Figure 3.1).
Page 31
3.2.4. Relative Signal Strength with Hysteresis and Threshold The last technique combines both the threshold and hysteresis values concepts to come with a technique with minimum number of handovers. The handover is requested when the BS1’s RSS is below the threshold (T1 in Figure 3.1) and BS2’s RSS is stronger than BS1’s by the hysteresis value h (point C in Figure 3.1). If a lower threshold than T1 (but higher than T2) is chosen then the handover initiation will be somewhere at the right of point C.
All the techniques discussed above initiate handover before point D where it is the “receiver threshold”. Receiver threshold is the minimum acceptable RSS for call continuation (T2 in Figure 3.1). If RSS drops below receiver threshold, the ongoing call is than dropped. The time interval between handover request and receiver threshold enable cellular systems to delay the handover request until the receiver threshold time is reached when the neighboring cell does not have any empty channels.
In this thesis, a handover algorithm using multi-level thresholds is proposed which assigns different threshold values to the users according to their speed. Since low speed users spend more time in handover zone they are assigned a higher threshold to distribute high and low speed users evenly. High speed users are assigned lower thresholds. The obtained performance results show that an 8-level threshold algorithm operates better than a single threshold algorithm in terms of forced termination and call blocking probabilities [4].
3.3. Measurements – a Prerequisite for Handover When the MS is busy the decision about which cell is the best for the dedicated connection is done by the BSS. The procedure for the system to figure out which cell is the most suitable, and evaluate the measurements, is called handover preparation. This is normally done in the BSC although the option of doing most of it in the BTS is also available. In the following text it has been assumed that the BSC to be responsible for the evaluation of the measuring reports. The change of the dedicated connection (SDCCH or TCH) is called handover which is handled by the BSC and MSC in one case. To be able to make the right handover decision, the BSC needs
Page 32
measurements on the connection to the serving cell as well as to the possible handover candidates, (i.e. the neighboring cells) [5].
Measurements in Busy Mode To perform a fast handover procedure the BSC is provided with all the necessary information about the serving and neighboring cells beforehand. Measurements are done by the BTS and MS during the call and will be reported to the BSC every 480 ms. The BTS will measure signal strength and bit error rate on the uplink while the MS will measure the same parameters on the downlink. In addition the MS will measure signal strength on the BCCH-carriers of the neighboring cells. The MS will send its reports on SACCH.
3.4. Handover Cases The measurement reports from the BTS and MS, together with system parameters set by the operator, are used in the preparation algorithm in the BSC. The outcome could be a handover if this is judged necessary. Three handover cases will be discussed. The difference between them is due to where the cells are located in the network structure, and thus how many nodes will be involved in the handover. One thing in common for all three cases is that the BSC that makes the handover decision will order and control the handover procedure from start to finish [5].
Intra- BSC Handover In this case the handover is controlled by the BSC internally and the MSC will only be informed for statistical reasons (Figure 3.2).
Figure 3.2: Intra-BSC Handover.
Page 33
If a better candidate for the connection is identified, based on the measurements carried out by the MS and BTS, the following will happen:
3.4.1. Activation of New Channel BSC allocates a TCH in the new cell and orders that BTS to activate it. The chosen HO will be part of the activation message. The BTS in the new cell will acknowledge that the TCH has been activated.
3.4.2. Handover Command After the activation the BSC commands (Figure 3.2) the MS to change to the new channel. The message is sent on FACCH and will contain a full description of the new channel and the HO.
3.4.3. Handover Bursts The MS will tune in to the new channel and send handover bursts on the new channel. The bursts are as short as the access bursts, since the MS does not know the new Timing Advance (TA) value yet. On the detection of the handover bursts, the new BTS will send the new TA to the MS.
3.4.4. Handover Complete Now the MS is ready to continue the traffic and will send a handover complete message addressed to the BSC.
3.4.5. Release of Old Channel When the BSC receives the handover complete message from the MS, the BSC will know that the handover was successful. The BSC orders the old BTS to release the TCH and the BTS will acknowledge [5].
3.5. Inter- BSC Handover In this case BSC1, (old BSC) does not control the letter cell which is the target for the handover. This means that the MSC (Figure 3.3) will be part of the link procedure between BSC1 and BSC2 (new BSC) [5].
Page 34
3.5.1(a). Handover Request BSC1 will use the MSC to send a handover request to BSC2 (Figure 3.3). The MSC will know which BSC controls that cell.
Figure 3.3: Inter-BSC Handover.
3.5.1(b). Activation of New Channel BSC2 will allocate a TCH in the target cell and then order the BTS to activate it. The chosen HO will be part of the activation message. The BTS will acknowledge that the activation has been made.
3.5.2. Handover Command After the activation the new BSC commands the MS to change to the new channel. The message is sent on FACCH via the old channel and will contain a full description of the new channel and the HO.
3.5.3. Handover Bursts When the MS has changed to the new channel, it will send handover bursts on the new channel. The information content is the HO. The bursts are as short as the access bursts (Figure 3.3). This is because the MS does not know the new Timing Advance (TA) value yet. On the detection of the handover bursts, and check of HO, the new BTS will send the new TA.
Page 35
3.5.4. Handover Complete Now the MS is ready to continue the traffic and will send a handover complete message, which will be addressed to the old BSC as a clear command.
3.5.5. Release of Old Channel When the old BSC receives the clear command from the MSC, the BSC (Figure 3.3) knows that the handover was successful. The BSC orders the BTS to release the TCH and the BTS will acknowledge.
3.6. Inter- MSC Handover In this case the old BSC is connected to a different MSC than the BSC that controls the target cell. This means that a new MSC will be part of the procedure (Figure 3.4). The old MSC will be called anchor-MSC and the new MSC will be called the target MSC [5].
Figure3.4: Inter-MSC Handover.
3.6.1(a). Handover Request The old BSC will use the anchor-MSC to send a request to the new BSC (Figure 3.4) for a handover to the target cell. The anchor-MSC knows which MSC to contact and the target-MSC in turn knows which BSC that controls the target cell.
Page 36
3.6.1(b). Activation of New Channel The new BSC allocates a TCH in the target cell and order the BTS to activate it. The chosen HO will be part of the activation message. The BTS will acknowledge that the activation has been made.
3.6.2. Handover Command After the activation the new BSC commands the MS to change to the new channel. The message is sent on FACCH via the old channel and will contain a full description of the new channel and the HO. In order to reroute the call, the target- MSC will also send a handover number, similar to the MSRN, to the anchor-MSC.
3.6.3. Handover Bursts When the MS has changed to the new channel, it will send handover bursts on the new channel. The information content is the HO. The bursts are as short as the access bursts as the MS does not know the new Timing Advance (TA) value yet. On the detection of the handover bursts, and check of HO, the new BTS will send the new TA.
3.6.4. Handover Complete Now the MS is ready to continue the traffic and will send a handover complete message, which will be addressed to the old BSC (Figure 3.4) as a clear command.
3.6.5. Release of Old Channel When the old BSC receives the clear command from the anchor MSC, the BSC knows that the handover was successful. The BSC orders the BTS to release the TCH and the BTS will acknowledge. •
Note that the MSC that performed the call setup will be in charge of the call until it is released, no matter handovers. The call is always routed to the target-MSC through the anchor-MSC.
•
Also note that the MS after call release has to perform a “location updating, type normal”. As a location area is part of only one MSC/ VLR Service area the MS must be in a new location area after the handover.
Page 37
3.7. Handover Types In this section will be discussed different types of handovers. First, channel usage will be considered. Then, will be investigated handover in microcells and multilayered systems [4]. Finally, handover in homogeneous and heterogeneous systems will be explained.
3.7.1. Hard vs. Soft Handover The hard handover term is used when the communication channel is released first and the new channel is acquired later from the neighboring cell. Thus, there is a service interruption when the handover occurs reducing the quality of service. Hard handover is used by the systems which use time division multiple access (TDMA) and frequency division multiple accesses (FDMA) such as GSM and General Packet Radio Service (GPRS).
In contrast to hard handover, a soft Handover can establish multiple connections with neighboring cells. Soft Handover is used by the code division multiple access (CDMA) systems where the cells use same frequency band using different code words. Each MS maintains an active set where BSs are added when the RSS exceeds a given threshold and removed when RSS drops below another threshold value for a given amount of time specified by a timer. When a presence or absence of a BS to the active set is encountered soft Handover occurs. The sample systems using soft Handover are Interim Standard 95 and Wideband CDMA (WCDMA) [4].
3.7.2. Microcellular vs. Multilayer Handover In this section will be first looked at the handover issues in microcellular environments. Later, some systems that use microcells overlaid by microcells in order to minimize number of Handovers will be investigate [4].
Figure 3.5: A city segment with three BSs deployed on streets Page 38
Microcellular Handover The microcells are cells with small radii and employed in highly populated areas such as city buildings and streets to meet high system capacity by frequency reuse. In Figure 3.5 there are two streets intersecting with three BSs employed on streets. BS1 and BS3 have line-of-sight (LOS) with each other. The handover between BS1 and BS3 is called LOS handover while the handover between BS1 and BS2 is a non-LOS (NLOS) handover since they don’t have LOS. In NLOS handovers, when a MS lose LOS (by turning the corner) with current BS, a drop in RSS (20-30 dB) occurs. This effect is called corner effect and needs faster handover algorithms since the RSS can drop quickly below receiver threshold resulting in a call drop. Two types of handovers, LOS and NLOS have different characteristics where LOS Handovers try to minimize the number of unnecessary handovers between BSs and NLOS must be as quickly as possible because of the corner effect.
In a fast handover algorithm for hard handovers is proposed to remove fast fading fluctuations resulting in algorithm that reacts more quickly to corner effect. They propose a technique called local averaging, in which the averaging time interval is smaller than averaging time interval of common handover algorithms and improve handover performance. A direction biased algorithm is proposed in where all the BSs in handover decision are grouped in two groups. One set of BSs are those in which MS is approaching and the other set includes the BSs in which the MS moves away. In handover initiation an encouraging hysteresis (e h) is used to first group where a discouraging hysteresis (d h) is applied to the second one. The relation between these hysteresis values are e d h ≤ h ≤ h. A signal strength based direction estimation method is used for determining the mobile positions [4].
Multilayer Handover Some designs used a multilayer approach in order to decrease the number of handovers and to increase system capacity. A number of microcells are overlaid by a microcell and the users are assigned to each layer according to their speeds. Microcells and microcells coverage area are respectively about 500 meters and 35 km for GSM900 in. Since slow users are assigned to the microcells and fast users are assigned to the microcells, the total number of handover requests is decreased. Microcells not only serve the fast users but also serve slow users when the microcells
Page 39
are congested. When a microcell allocates all of its channels, the new and handover calls are overflowed to the microcell layer. When the microcells load decreases it is possible to assign slow users back to the microcell. This type of handover is called take-back. So far, we have four types of Handovers: microcell-to-microcell, microcell-to-microcell, microcell-to-microcell, and microcell-to-microcell. In a bonus-based algorithm is proposed where it is compared with classical and macro algorithms. In the classical algorithm, in the case of new call request a user is assigned to microcell or overflowed to microcell if capacity of microcell is full. After the user speed estimation is done, the user is assigned to the appropriate layer using overflow and takeback. This scheme results in too many handovers known as the ping-pong effect. Macro algorithm is similar to classical algorithm with one exception. When a user is assigned to the microcell it is not permitted to take-back to microcell which decreases the number of handovers. The bonus-based algorithm tries to prevent unnecessary handovers to microcell when fast users temporarily slow down. For each fast user a time bonus is given and user can use this time bonus during temporary slowdowns. If a user exceeds the timer then it is assigned as a slow user and is taken-back to the microcell layer. Microcells and microcells are terrestrial part of the network whereas spot beams correspond to satellite part. The users can be overflowed from low layers to the uppers but take-back is not allowed here [4].
3.7.3. Horizontal vs. Vertical Handover Handover between homogenous networks where one type of network is considered is called horizontal handover. On the other hand, handover between different types of networks is also possible. A handover in such a heterogeneous environment is named vertical handover and it is out of scope of this paper. All the issues described in this paper are related to horizontal handover [6].
Page 40
Chapter 4
Mechanism & Analysis of Handover Management 4.1. Introduction In this section a couple of enhancements which can improve the performance of GSM handover algorithms are presented and studied.
4.2. Conventional Handover Mechanism In GSM cellular network both the mobile station and the BTS regularly measure the radio signal strength. The mobile station transmits its measurements reports continuously to the BTS. If the BTS detects a decrease in radio signal under a minimal level durge, it initiates a handover request as shown in Figure 4.1. The BTS then informs the BSC about the request, which then verifies if it is possible to transfer the call into a new adjacent cell.
Actually the BSC checks weather a free channel is available in the new adjacent cell or not. In this situation the BSC does not differentiate between the channel requests either for fresh call or handover. If a free channel is available in the new adjacent cell then handover request can be satisfied, and the mobile station switch to new cell.
If there is no free channel in the adjacent cell then it increases the dropping probability of handover call. The drawback of this handover procedure is the fact that the handover request for channel is same as used for fresh calls. In conventional handover mechanism it is very problematic from the users quality of service perspective, since user can much prefer block a fresh call rather than to be dropped a call in the middle of transmission [7].
Page 41
Figure 4.1: Signal Levels for Handover
4.3. Channel Carrying Handover Mechanism The channel carrying mechanism allows a mobile station to carry its current channel from one cell to another when it moves across the boundaries under specific conditions. The channel carrying mechanism uses a linear cellular system model in which cells or BTS are arranged in linear configuration with minimum reuse distance ‘r’ as shown in the Figure 4.2. Suppose N is the total number of channels available for use in cellular system. Two cells can use the same set of channel as they are apart by distance r.
Figure 4.2: r and (r+1) Channel Carrying
To avoid the co-channel interference an advance solution is proposed in which the distance of identical sets of channels is increased to r+1 instead of r. The distance r is the minimum reuse distance or reuse factor. According to the figure the total number of available channels in each
Page 42
cell is now reduced by amount of N/r+1 where N is the total number of available channels. In typical situation the smaller the reuse distance the more amounts of channels is to be lost. The channel carrying mechanism does not require the complex power control or global channel coordination which simplifies its implementation. Handover request are greatly favored over new calls compared to the conventional handover mechanism. The main drawback of this handover procedure is not suitable for metropolitan environment due to the great amount of channels lost [8].
4.4. GSM Handover Prioritization Schemes Different ideas and approaches are proposed to reduce the handover dropping probability. One approach is to reduce the handover failure rate is to prioritize handover call over new calls. Handover prioritization schemes have a significant impact on the call dropping probability and call blocking probability. Such scheme permits high utilization of bandwidth while guaranteeing the quality of service of handover calls. Basic method of handover prioritization schemes are guard channels (GC), call admission control (CAC) and handover queuing schemes. Sometimes these schemes are combined together to obtain better results [9].
4.4.1. Guard Channel Prioritization Scheme The guard channel scheme was introduced in 80s for mobile cellular systems. However the guard channel schemes are still used in telecommunications with the name of Cutoff Priority Schemes. GC scheme improves the probability of successful handover by simply reserving a number of channels exclusively for handover in each cell.
The remaining channels can be shared equally between handover and new calls. GC are established only when the number of free channels is equal to or less than the predefined threshold G as shown in Figure 4.3. In this situations fresh calls are bypassed and only handover request are served by the cell until all channels are occupied. The GC scheme is feasible because new calls are less sensitive to delay than the handover calls [11].
Page 43
Figure 4.3: Guard Channels for Handover Request
A cellular network is considered with C, the total number of channels in a given cell. According to GC scheme reserve channels for handover are C-T, where T is the predefined threshold. The GC will not accept any new call until the channel occupancy goes below the threshold. Suppose the arrival of new and handover call is denoted with λ and v respectively. The call holding and call residency for both call is exponentially distributed with 1/μ and 1/η respectively. The total traffic can be calculated as ρ= (λ + v) / (μ+ η). Figure 4.4 shows the the state transition diagram of guard channels. r+d
µ+n
r
(T+1)(µ+n)
r
C (µ+n)
Figure 4.4: State Transition Diagram of Guard Channels
Therefore according to the cell occupancy by Markov chain it is straight forward to derive the steady state probability Pn that n channels are busy.
Page 44
𝑃𝑃𝑛𝑛 = Where,
⎧ ⎪
𝜌𝜌𝑛𝑛 � � 𝑃𝑃0 , 𝑛𝑛!
⎨ 𝜌𝜌𝑇𝑇 �𝑣𝑣 � 𝑃𝑃 , 0 ⎪ 𝑛𝑛! ⎩ 𝑛𝑛−𝑇𝑇
0 ≤ 𝑛𝑛 ≤ 𝑇𝑇
𝑇𝑇 ≤ 𝑛𝑛 ≥ 𝐶𝐶
𝑇𝑇
𝐶𝐶
𝑛𝑛 =0
𝑛𝑛 =𝑇𝑇+1
𝜌𝜌𝑛𝑛 𝑣𝑣 𝑛𝑛 −𝑇𝑇 𝑃𝑃0 = � + 𝜌𝜌𝑇𝑇 � 𝑛𝑛! 𝑛𝑛! Pb =ΣC n=T+1, Pn and pf= Pc. In fact, there is a tradeoff between minimizing Pd and minimizing Pb. If the number of channels is static chosen then the admission call control fails to satisfy the specified Pd. The static channels reservation shows results poor utilization of bandwidth [10].
4.4.2. Call Admission Control Prioritization Scheme The call admission control scheme refers to the task of deciding whether new call requests are admitted into the network or not. In the CAC the arrival of new calls are estimated continuously and if they are higher than the predefined threshold level then some calls are restricted (blocked) irrespective of whether a channel is available or not to decrease the probability of dropping handover calls. In the CAC both the new and handover calls have to access to all channels. If a new call that is generated in cell cannot find an idle channel the call is discarded immediately. There is no queue provided for the new calls to wait. The CAC scheme can be classified into different schemes that consider the local information like (the amount of unused bandwidth in cell where the user currently resides), remote information (the amount of unused information bandwidth in the neighboring cells) or local or remote information to determine whether to accept or reject a call. CAC based on knowledge of both network and user characteristics, keeps the track of available system capacity and accommodates new call request while ensuring quality of service for all existing users. Decisions in CAC are performed in each BSC in a distributed manner and there is no central coordination. The CAC scheme can be evaluate on the basis of efficiency, fairness, stability and Flexibility [12].
Page 45
4.4.3. Handover Queuing Prioritization Schemes Queuing handover call prioritization scheme queues the handover calls when all the channels are occupied in the BSC. When a channel is released in the BSC, it is assigned to one of the handover calls in the queue. The handover queuing scheme reduces the call dropping probability at the expense of the increased call blocking probability and decrease in the ratio of carried to admitted traffic since new call are not assigned a channel until all the handover request in the queue are served. In the handover queuing schemes when the received signal strength of the BSC in the current cell reaches to certain defined threshold, the call is queued from service of a neighboring cell. A new call request is assigned a channel if the queue is empty and if there is at least one free channel in the BSC. The call remains queued until either a channel is available in the new cell or the power by the base station in the current cell drops below the receiver threshold. If the call reaches the receiver threshold and no free channel is found then the call is terminated. Queuing handover is possible due to the overlap regions between the adjacent cells in which the mobile station can communicate with more than one base station. This makes provision of the queuing the handover requests for certain time period equal to the time of the mobile host existence in the overlapping area. Queuing is effective only when the handover requests arrive in groups and traffic is low. First in first out (FIFO) scheme is the most common queuing scheme where the handover requests are ordered according to their arrival. To analyze this scheme it is necessary to consider the handover procedure in more detail. Infinite queue size at the base station is assumed having the FIFO queuing strategy, which is shown in Figure 4.5.
Figure4.5: Priority Queue System Model for Handover Call
The handover of the mobile station depends on the system parameters such as moving speed, the direction of the mobile station and the cell size. Suppose the state i (i=0, 1, 2 …, ∞) of a cell as Page 46
the sum of the channels being used and the number of the handover call request in the queue. Then it is clear from the Markov chain that ί one- dimensional. Figure 4.6 shows the sate transition diagram of the cells from the above Figure 4.5. The equilibrium probabilities are related to each other through the following equations [14].
Figure 4.6: State Transition Diagram
𝑖𝑖𝑖𝑖𝑖𝑖(𝑖𝑖 ) = (𝜆𝜆0 + 𝜆𝜆𝐻𝐻 )𝑃𝑃(𝑖𝑖 − 1)
0 ≤ 𝑖𝑖 ≤ 𝑆𝑆𝑐𝑐
�𝑆𝑆𝜇𝜇 + (𝑖𝑖 − 𝑆𝑆)(𝜇𝜇𝐶𝐶 + 𝜇𝜇ℎ −𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 )�𝑃𝑃(𝑖𝑖 ) = 𝜆𝜆𝐻𝐻 𝑃𝑃(𝑖𝑖 − 1)
𝑆𝑆 < 𝑖𝑖 ≤ ∞
𝑖𝑖𝑖𝑖𝑖𝑖(𝑖𝑖 ) = 𝜆𝜆𝐻𝐻 𝑃𝑃(𝑖𝑖 − 1)
𝑆𝑆𝑐𝑐 ≤ 𝑖𝑖 ≤ 𝑆𝑆
Then the steady state probability is found as follows according to the Markov chain [14] (𝜆𝜆 + 𝜆𝜆 )2 ⎧ 0 𝑡𝑡 𝐻𝐻 𝑃𝑃(0), ⎪ 𝑖𝑖! 𝜇𝜇 1−𝑆𝑆 ⎪( 𝜆𝜆0 + 𝜆𝜆𝐻𝐻 )𝑆𝑆𝑐𝑐 𝜆𝜆𝐻𝐻 𝑐𝑐 𝑃𝑃 (0), 𝑃𝑃(𝑖𝑖 ) = 𝑖𝑖! 𝜇𝜇 𝑡𝑡 ⎨ 1−𝑆𝑆 ⎪ (𝜆𝜆𝑜𝑜 + 𝜆𝜆𝐻𝐻 )𝑆𝑆𝑐𝑐 𝜆𝜆𝐻𝐻 𝑐𝑐 ⎪ 𝑃𝑃(0), 𝑆𝑆! 𝜇𝜇 𝑠𝑠 ∑𝑖𝑖−𝑠𝑠 ⎩ 𝑗𝑗 =1�𝑆𝑆𝜇𝜇 + 𝑗𝑗 (𝜇𝜇𝑐𝑐 + 𝜇𝜇ℎ −𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 )�
0 ≤ 𝑖𝑖 ≤ 𝑆𝑆𝑐𝑐
𝑆𝑆𝑐𝑐 < 𝑖𝑖 ≤ 𝑆𝑆
𝑆𝑆 < 𝑖𝑖 ≤ ∞
Where, 𝑆𝑆𝑐𝑐
𝑆𝑆
𝑡𝑡=0
𝑡𝑡=𝑆𝑆𝐶𝐶 +1
𝑡𝑡−𝑆𝑆 𝐶𝐶
(𝜆𝜆0 + 𝜆𝜆𝐻𝐻 )𝑆𝑆𝑐𝑐 𝜆𝜆 𝐻𝐻 (𝜆𝜆0 + 𝜆𝜆𝐻𝐻 )𝑡𝑡 𝑃𝑃(0) = �� + � 𝑖𝑖! 𝜇𝜇 𝑡𝑡 𝑖𝑖! 𝜇𝜇 𝑡𝑡 ∞
1+𝑆𝑆
(𝜆𝜆0 + 𝜆𝜆𝐻𝐻 )𝑆𝑆𝐶𝐶 𝜆𝜆𝐻𝐻 𝑐𝑐 + � ∏𝑡𝑡−𝑆𝑆 𝑆𝑆! 𝜇𝜇 𝑠𝑠 𝑗𝑗 =1[𝑆𝑆𝜇𝜇 + 𝑗𝑗 (𝜇𝜇𝑐𝑐 + 𝜇𝜇ℎ −𝑑𝑑𝑑𝑑𝑑𝑑 𝑙𝑙𝑙𝑙 )] 𝑡𝑡=𝑆𝑆+1
Page 47
Therefore the blocking probability Bo for an originating call is 𝑆𝑆
And termination probability of the call is
𝐵𝐵0 = � 𝑃𝑃(𝑖𝑖) 𝑡𝑡=𝑆𝑆𝑐𝑐
∞
𝑃𝑃′𝑓𝑓 = � 𝑃𝑃 (𝑆𝑆 + 𝑘𝑘)𝑃𝑃𝑓𝑓ℎ/𝑘𝑘 𝑘𝑘 =0
In the FIFO handover Prioritization scheme the probability of the forced termination is decreased however the handover call may be dropped because the handover request can only wait until the receiver threshold is reached [14].
Some new queuing schemes like Measurement based prioritization scheme (MBSP), very early assignment (VEA), early assignment (EA) and most critical first (MCF) are proposed to improve the performance to improve of the handover queuing scheme by modifying the queuing discipline. In the MBSP the handover calls are added to the queue and the priorities of the calls changes dynamically based on the power level they have. The call with the power level close to the receiver threshold has the highest priority. This scheme is produce better results than the FIFO queuing schemes. Each of these schemes has its advantages and disadvantages in the term of capacity and services. Like the VEA gives the shortest call setup but is most capacity inefficient [13].
4.4.4. Cell Overlapping and Load Balancing Scheme (Proposed Scheme) In order to improve the handover call prioritization scheme it is advisable to equalize the traffic load over the cells. Traffic region and directed retry handover make use of this principal. First the new call to be served and if the receiver is able hear a neighboring cell and are not considered in this situation. Traffic reason handover can be used to transfer traffic from one cell to another neighboring when they are closed to the blocking. The traffic reason handover idea is based on the neighboring cell having an overlapping service area. The overlapping service area arises naturally in GSM cellular system especially in small-cell high capacity micro cellular configurations. The small-cells are captured by subdividing a congested cell each with its own base station. The call arising in the common area (overlapping) of cells have access to channels Page 48
more than one base station. By appropriate control strategy a cell may select the base station to establish a connection and contribute to efficient spectrum management. By subdividing a congested cell into small- cell the frequency reuse distance is effectively increased which reduce the level of interference and increase the carrier to interference ratio at both side the mobile station and base station. From the previous work it has been proven the directed retry an increase in the overlapping between cells leads to increases the quality of service of the cellular system. A large overlapping area gives more capacity than a smaller overlap, but even by just having a small overlap a significant gain is achieved. The overlap of 0.1R (where R is the radius of cell) results an overlapping area equal to 9% of the cell area gives a gain of at least 6% whereas if the overlap is equal to 0.5R means overlapping area is 75% of the cell area then the capacity gain is boost to 27%. The performance of this functionality is very dependent on the existing overlapping between cells since it is required that at least one neighboring cell has sufficient signal level for the mobile station to be redirected [13].
Figure 4.7: Areas A, B and C of three Cells
According to the concept of cell radius when two or more adjacent cells overlap they form a set of individual regions which can be categorized into three types A, B, and C according to the number of cells they overlap as shown in Figure 4.7. These regions can be assigned a channel from one of three cells. The importance of the regions and areas is to perform the channel allocation scheme based on either through the region or area. The number of channels for specific region depends on the size of the regions and specified channel can be used in that area. The blocking probability of the cell can be calculated from those users who are able of choosing a channel from cells A, B, and C. This maintains the same lowest blocking probability and load balancing in every area [15]. Page 49
4.5. Analyses In this paper both the prioritized and the non prioritized handover scheme are presented. Moreover different prioritization schemes and there extensive classification are presented as well. All the handover prioritization schemes allocate channels to handovers more frequently than the new call to guarantee the users QoS perspective because new calls are less sensitive to delay than the handover calls. One of the simplest ways introduced in the above literature of giving priority to the handover calls is to reserve a number of channels exclusively for the handover in each cell to improve the performance of the cellular system. The guard channel prioritization schemes are established only when the number of free channels is less or equal to predefined threshold. The value of the threshold directly affects the probability of the call blocking and call dropping. According to the cell channel occupancy by Markov chain it is straight forward to derive the steady state probability P n that n channels are busy and then P b =Σ
𝐶𝐶 n = T+1 P and Pf= P . The equation Pf = P shows that the handover failure probability is n c c
equal to the call completion probability. A critical parameter is the number of channel guard
channel exclusively for handover in each cell. In fact there is relationship between minimizing
Pd (call dropping probability) and minimizing Pb (call blocking probability). The guard channel prioritization scheme has the risk of underutilizing the frequency channels or insufficient spectrum utilization.
According to the equation if the number of guard channels are conservatively chosen then the admission control fails to satisfy the specified the call dropping probability. Therefore an efficient estimation method for the optimum number of guard channel is essential. In addition when a user moves into new cell, bandwidth is reserved in the new neighboring cell and the reserved bandwidth in the cell which are no longer used to the new cell is released. However from the operation point of view there are several weaknesses in CAC scheme. First it seems difficult for CAC scheme to handle the user’s request where the capacity is not enough to deal with all the requirements. In the situation the calls blocking and dropping probability increases this affects QoS of the wireless cellular network. But if the user has predefined priority then the CAC can distribute capacity according to each user so that the requirements of each with higher
Page 50
priority will be fulfilled prior to any other user. In such situation the CAC algorithms becomes more complicated and complex to meet with these requirements. Secondly during rush calling events such is at the end of concerts or cricket or football match or other such big events the network is flooded by the high volume of user calls. Blocking of user during such situation will help lighten the cellular network congestion but will result unsatisfied customers and lost of revenue. Furthermore some calls are blocked unnecessary in the neighboring cells due to the sudden fluctuation in the new call and the handover arrival rates.
Finally as the CAC uses the high level of frequency reuse this leads to costly implementation and more interference in the wireless cellular network. Several other strategies to allocate channel for the handover request in the queue discipline have been proposed. For example queuing of new call arrivals is possible and is less sensitive regarding the queuing time than the case of handover. Queuing of the new call request shows more improvement than queuing of handover calls. In this scheme new call will be accepted if the number of free channels apart of those reserved for handover is enough for the new request otherwise the call be placed in the queue. As soon as the channel is released by the completing a call or outgoing of the handover request then the new call is served immediately from the FIFO queue. The queuing of the new calls involves the concept of the guard channels and queuing schemes. The performance analysis of queuing new call shows; that the blocking of the handover call decreases with the queuing probability of the new calls and increased in the total carried traffic because new calls will be ultimately served. This scheme also achieves less force termination probability compared to other schemes.
Page 51
Chapter 5
Conclusion In this thesis an introduction to GSM with architecture, concept of cellular network, overview of handover, handover mechanism and analysis of handover management are discussed elaborately. This will be helpful in the future when anyone will work in this practical field of GSM handover management. A clear view of GSM cellular concept and overview of handover management is depicted here. Also the shape of cell, location management and evaluation of cellular network, handover initiation, inter-BSC and MSC handover, types of handover are discussed in the report.
Finally as the CAC uses the high level of frequency reuse, this leads to costly implementation and more interference in the wireless cellular network. Several other strategies to allocate channel for the handover request in the queue discipline have been proposed. For example queuing of new call arrivals is possible and is less sensitive regarding the queuing time than the case of handover. Queuing of the new call request shows more improvement than queuing of handover calls. In this scheme new call will be accepted if the number of free channels apart of those reserved for handover is enough for the new request otherwise the call be placed in the queue. As soon as the channel is released by the completing a call or outgoing of the handover request then the new call is served immediately from the FIFO queue. The queuing of the new calls involves the concept of the guard channels and queuing schemes. The performance analysis of queuing new call shows; that the blocking of the handover call decreases with the queuing probability of the new calls and increased in the total carried traffic because new calls will be ultimately served. This scheme also achieves less force termination probability compared to other schemes.
In this thesis the greater depth in the GSM network architecture and handover process is in illustrated which emphasizes the architecture, the several functional network elements and their dedicated channels associated with the call. Furthermore different performance metrics used to make the handover decision is also discussed. Then the most important procedure of GSM
Page 52
handover initiation, handover types and their measurements is discussed to ensure mobility in GSM network and to emphasis the fact that handover in GSM network are very important to maintain the quality of a call.
The call handover prioritization schemes that prioritize handover calls in order to enhance the quality of service (QOS) of GSM wireless network is investigated. Extensive survey and analysis of the handover prioritization schemes that is guard channels, call admission control and handover queuing has been provided. Furthermore this research indicates that different system uses different schemes to execute the handover mechanism for a couple of enhancements to the handover mechanism. The idea of the cells overlap and load balancing scheme which tries to equalize the traffic over cells has been introduced. It has been analyzed theoretically and mathematically that capacity depends on the size of the overlapping area between adjacent cells, the numbers of channels per cells and distribution of traffic. The higher the overlapping area, the trucking efficiency is increased. The overlapping area can be used to reduce the call blocking and dropping probabilities. The attractive feature of this scheme is that it organizes traffic in distributed manner and doesn’t increase the system complexity. It can be concluded that the implementation of mathematical formulas as mentioned in research will be great contribution in call handoff and for QOS’s. Most preferably, after completing this thesis it will be clear to anyone how the handover management is work in GSM system. The specific purpose of this thesis was about the working processes of handover management system in GSM network which was studied thoroughly for this practical environment.
Page 53
References: [1]
GSM Architecture – Training Document, TC Finland Issue January - 2002.
[2]
Jingyuan Zhang, “Cellular Networks”, University of AlabamaIvan Stojmenovic, University of Ottawa, Cananda
[3]
Dr. Wajih A. Abu-Al-Saud, “Wireless & Mobile Communications (082)” - EE 499.
[4]
Nasıf Ekiz, Tara Salih, Sibel Küçüköner, and Kemal Fidanboylu, “An Overview of Handoff Techniques in Cellular Networks” - (World Academy of Science, Engineering and Technology 6 2005).
[5]
Michael Quck – “Handover”.
[6]
M. Stemm and Randy H. Katz, “Vertical handoffs in wireless overlay networks”, Mobile Networks and Applications, vol. 3, 1999, pp. 335- 350.
[7]
Mandjes, M., and Tutschku, K. (1996) Efficient Call Handling Procedures in Cellular Mobile Networks, [Online].Report No.144, Available from : http://www3.informatik.uniwuerzburg. de/publications/reports.shtml [Accessed 4 March 2007].
[8]
Li, J., Shorff, N and Chong, E (1996) Channel Carrying: A Novel Handoff Scheme for Mobile Cellular Networks. [Online]. ECE Technical Reports, Available from: http://docs.lib.purdue.edu/ecetr/98/ [Accessed 6 March 2007].
[9]
Tripathi, N., Nortel, Jerry, R and Vanlandingham. (1998) Handoff in Cellular Systems. [Internet] .IEEE PersonalCommunications. Available from : https://www.cs.tcd.ie/~htewari/papers/tripathi98.pdf [Accessed 7 March 2007].
[10]
Ghaderi, M and Boutaba. “Call Admission Control in Mobile Cellular Networks: A Comprehensive Survey”. Wireless Communications and Mobile Computing, 6(1), pp 6993. Available from : http://bcr2.uwaterloo.ca/~rboutaba/Papers/Journals/WCMC- 04.pdf [Accessed 9 March 2007], (2006).
[11]
Katzela, I and Naghshineh. “Channel Assignment Scheme for Cellular Mobile Telecommunication Systems- A Comprehensive Survey”, IEEE Personal Communications, 3 (3): pp.10-31 (2006).
Page 54
[12]
Yliopisto, O. “Smart Packet and Call Admission Control for Efficient Resource Management in AdvancedWireless Networks”. PhD Dissertation. University of Oulu, Finland (2006).
[13]
Nielsen, T., and Wigard, J. “Performance Enhancements in a Frequency Hopping GSM Network”. Kluwer Academic Publisher, Netherlands (2000).
[14]
Zeng, Q., Agrawal, D. “Handoff in Wireless Mobile Networks”. [Internet]. Ohio.USA(2001). Available from : http://media.wiley.com/product_data/excerpt/28/04714190/04714 19028.pdf [Accessed 20 Jan 2007].
[15]
Katzis, K. “Resource Allocation Techniques for High Altitude Platforms”. PhD Dissertation, University of York, UK (2005).
Page 55
