`
`Aachen Technical University (RWTH)
`
`Prof. Dr.-Ing. B. Walke
`
`Diploma Paper
`
`Simulative and analytical study of measures
`supporting the quality of service in a radio-
`based ATM network
`
`by
`
`Ulrich Vornefeld
`
`Student ID No. 187020
`
`Aachen, April 1, 1997
`
`Mentors:
`
`Prof. Dr.-Ing. B. Walke, Associate Professor
`Dipl.-Ing. D. Petras
`Dipl-Ing. A. Hettich
`
`This paper is for internal use only. All copyrights reserved by the mentoring department. We give
`
`no warranty for its content. Reproduction and publication of any kind are subject to the
`
`Department’s permission.
`
`Broadcom v. Ericsson
`
`|PR2013-00636
`
`ERicsson Ex. 2019
`
`
`
`I hereby confirm that l have conducted this work independently Without the assistance from third
`
`parties — except for the official mentoring by the department. All literature used for this paper is
`
`listed completely in the Bibliography section.
`
`Aachen, April 1, 1997
`
`(Ulrich Vomefeld)
`
`
`
`
`
`TABLE OF CONTENTS
`
`Introduction
`
`Asynchronous Transfer Mode, ATM
`
`3
`
`5
`
`2.1
`
`Structure of the ATM cell and meaning of the control information ......................... 6
`
`2.2 ATM switching ......................................................................................................... 7
`
`2.3
`
`The ATM reference model ........................................................................................ 7
`
`2.4 Quality of service in the ATM landline network ...................................................... 9
`
`Architecture of the ATM radio interface
`
`13
`
`3.1
`
`The nctwork cell as a distributed ATM multiplcxcr ............................................... 14
`
`3.2
`
`The protocol stack of the ATM radio interface ....................................................... 17
`
`The layers of the protocol stack and their functions
`
`19
`
`4.1 Operating strategies of the distributed virtual multiplexer ...................................... 19
`
`4.2
`
`The MAC protocol and its services ......................................................................... 20
`
`4.2.1
`
`Length of the time slots and physical layer ................................................ 22
`
`4.2.2
`
`Signaling of capacity requests .................................................................... 23
`
`4.3
`
`4.4
`
`Structure of the LLC layer ...................................................................................... 26
`
`The cell scheduler ................................................................................................... 28
`
`4.4.1
`
`The reservation phase of the scheduler ...................................................... 29
`
`4.4.2
`
`The transmission phase of the scheduler .................................................... 30
`
`The SR/D ARQ protocol
`
`35
`
`5.1
`
`Error control measures ............................................................................................ 35
`
`5.2
`
`Link access protocols .............................................................................................. 36
`
`5.2.1
`
`Go back n (continuous) ARQ ..................................................................... 36
`
`5.2.2
`
`Selective Repeat (SR) ARQ ....................................................................... 37
`
`5.3 ARQ protocols for the real—time oriented service classes ....................................... 38
`
`5.3.1 Discarding ATM cells ................................................................................ 38
`
`5.3.1.1
`
`Explicit recipient notification using Discard messages ............. 39
`
`
`
`Table ofcomenm
`
`5.3.1.2
`
`Implicit discarding by moving the window ............................... 40
`
`5.3.2 Methods for reducing transmission delays ................................................. 41
`
`5.4 No reject acknowledgment for implicit discarding ................................................. 42
`
`5.5
`
`Long waiting times in the recipient with implicit discarding .................................. 44
`
`6
`
`Acknowledgment strategies
`
`47
`
`6.1
`
`Priority of acknowledgments .................................................................................. 47
`
`6.1.1
`
`Priority of positive acknowledgments ........................................................ 48
`
`6.1.2
`
`Priority of negative acknowledgments ....................................................... 48
`
`6.1.3
`
`Priority of acknowledgments with a set poll bit ......................................... 48
`
`6.1.4
`
`Priority of discard messages ....................................................................... 49
`
`6.2 Methods for acknowledgment transmission examined ........................................... 49
`
`6.2.1
`
`Reservation decisions due to acknowledgment transmissions ................... 51
`
`6.2.2
`
`Capacity requests from terminals for acknowledgment
`transmission ............................................................................................... 52
`
`6.2.3
`
`Influence of acknowledgment transmissions on the transmission
`phase .......................................................................................................... 52
`
`6.3 Acknowledge algorithm of the reservation phase ................................................... 54
`
`6.4 Acknowledge algorithm of the transmission phase ................................................. 58
`
`6.4.1
`
`Bundle acknowledgment ............................................................................ 61
`
`6.4.2 Determining the SREJ to be sent ............................................................... 61
`
`6.4.3
`
`Channel monitoring .................................................................................... 61
`
`6.5
`
`Parameters studied and assessed ............................................................................. 63
`
`7
`
`Integrated stochastic simulation model
`
`65
`
`7.1
`
`Software architecture of the simulation model ....................................................... 65
`
`7.2 Modeling the application processes ........................................................................ 66
`
`7.2.1
`
`The deterministic stochastic process as a model for CBR sources ............ 66
`
`7.2.2
`
`The Poisson process as a model for VBR sources ..................................... 67
`
`7.2.3 Video source with an autoregressive process as a model of a typical
`VBR application ......................................................................................... 67
`
`7.2.4 Non-time-critical ABR and UBR services ................................................. 68
`
`7.3
`
`7.4
`
`The physical channel ............................................................................................... 68
`
`Transmission errors ................................................................................................. 68
`
`7.4.1 Uncorrelated channel errors ....................................................................... 69
`
`7.4.2
`
`The Gilbert model for correlated channel errors ........................................ 70
`
`7.5
`
`Implementation aspects
`
`
`
`Table ofcomenm
`
`8 Performance evaluation
`
`71
`
`8.1
`
`Evaluation goals and performance characteristics to be determined ...................... 71
`
`8.1.1
`
`8.1.2
`
`Performance characteristics for real-time oriented CBR/VBR
`connections ................................................................................................. 71
`
`Performance characteristics for non-time-critical ABR/UBR
`connections ................................................................................................. 71
`
`8.2 Confidence level of the simulation results .............................................................. 72
`
`8.3
`
`The ideal ARQ protocol .......................................................................................... 72
`
`8.4 Greater delays for uncorrelated channel errors ....................................................... 73
`
`8.5
`
`8.6
`
`Explicit, implicit discarding and the various levels of discarding .......................... 73
`
`Performance evaluation of the acknowledge procedure .......................................... 76
`
`8.6.1
`
`LLC acknowledgments in the period Ctrl PDU ......................................... 76
`
`8.6.2
`
`Inserting short poll time slots ..................................................................... 78
`
`8.6.3
`
`Random access due to acknowledgment transmissions ............................. 80
`
`8.6.4
`
`Bundle acknowledgment ............................................................................ 81
`
`8.7 Realistic simulation scenarios with multimedia services ........................................ 82
`
`8.7.1
`
`Size of the ARQ window ........................................................................... 83
`
`8.7.2 Maximum number of reservations ............................................................. 83
`
`8.7.3
`
`Correlated and uncorrelated channel errors ............................................... 85
`
`8.7.4
`
`Channel monitoring .................................................................................... 87
`
`8.7.5
`
`Comparing the SR/D protocol and the ideal ARQ protocol ....................... 89
`
`9
`
`Summary and outlook
`
`93
`
`9.1
`
`9.2
`
`Summary ................................................................................................................. 93
`
`Evaluation of the findings ....................................................................................... 93
`
`9.3 Outlook .................................................................................................................... 94
`
`List of figures ......................................................................................................................... 99
`
`List of tables ......................................................................................................................... 101
`
`List of abbreviations ............................................................................................................ 104
`
`Bibliography ......................................................................................................................... 105
`
`
`
`Table ofcomenm
`
`
`
`Table ofcomenm
`
`
`
`
`
`CHAPTER 1
`
`Introduction
`
`Increasing globalization and internationalization of society and the demand for information that
`
`goes with it constantly increase the information provision and information transmission
`
`requirements that communication networks have to meet. The development of these networks is
`
`essentially driven by two trends: by integration of the various telecommunications services within
`
`one network and by accounting for demands for individual mobility and universal accessibility.
`
`These trends have resulted in the development and introduction of narrowband ISDN (Integrated
`
`Services Digital Network) and digital mobile telephone networks based on the GSM standard
`
`(Global System for Mobile Communication). ISDN eliminates the large number of different
`
`network user interfaces and assigns voice, text, and data services via a uniform interface.
`
`The introduction of new interactive and real-time oriented services and the constantly growing
`
`need for bandwidth of the Internet ensure continuous development of the worldwide cable-based
`
`infrastructure towards multimedia broadband ISDN, which is popularly called the “information
`
`highway.” This network utilizes more and more of a new transmission and switching technology
`
`called ATM (Asynchronous Transfer Mode) [6, 9, 29]. It is expected that the trend towards
`
`universal accessibility in conjunction with wireless access to multimedia services will be one of
`
`the major drivers of telecommunications in the near future [18].
`
`Progress made in recent years in microelectronics and signal processing has made the
`
`implementation of wireless terminals for B-ISDN based on ATM technology a realistic
`
`possibility. The industry and the universities are responding with increased research activities in
`
`the field of wireless ATM networks (for an overview, see the Introduction of [16]). Wireless
`
`ATM workgroups (WAG) were formed at the European Telecommunications Standards Institute
`
`(ETSI) and the ATM Forum, who study the requirements and architectures of W-ATM systems
`
`[26, 4]. It is planned to adopt first technical standards by the end of 1998. The technological
`
`conditions for wireless communications systems with transmission rates of several 10 Mbit/s per
`
`connection were created by releasing the respective frequency range for personal communications
`
`systems in the 5 GHz range.
`
`This paper interprets the radio interface as a distributed ATM multiplexer that performs statistical
`
`multiplexing of ATM cells on the radio channel. An operating strategy with static and dynamic
`
`priorities is used to determine the transmission sequence of ATM cells. The system and protocol
`
`aspects of the W-ATM radio interface that determine the transmission sequence of the ATM cells
`
`and perfomi the multiplexing functions are examined in order to meet the quality of service
`
`requirements of all ATM service classes through suitable extensions, algorithms, and
`
`
`
`6
`
`1. Introduction
`
`modifications. These extensions, algorithms, and modifications are called measures supporting
`
`the quality of service in this paper.
`
`Overview of the paper
`
`After a brief introduction to ATM technology in Chapter 2, Chapter 3 presents the architecture of
`
`the ATM radio interface. The protocol stack derived from it is described in Chapter 4, with a
`
`focus on controlling the transmission sequence of the ATM cells. Chapter 5 explains the link
`
`access protocols at the radio interface, Chapter 6 details the acknowledge strategies and measures
`
`supporting the quality of service that result from these protocols. Chapter 7 provides an overview
`
`of the various aspects of the simulation model used and implemented. Chapter 8 ( Performance
`
`evaluation) presents the results of the simulation.
`
`
`
`
`
`CHAPTER 2
`
`Asynchronous Transfer Mode, ATM
`
`ATM stands for asynchronous transfer mode and denotes a new asynchronous transmission,
`
`switching, and multiplexing method that was developed to meet the continually increasing
`
`requirements posed by new services and more and more powerfiil computers in future broadband
`telecommunications networks in accordance with the 1.300 series of ITU1 Recommendations.
`
`In ATM networks, the data streams to be transmitted are divided into fixed-length packets (cells)
`
`and transmitted asynchronously. This means that the cells of a data stream do not have fixed time
`
`slots. The cells from different connections are transmitted via a physical channel in a time-
`
`interleaved manner in the order of their arrival. The cells of one connection may not pass one
`
`another (virtual connection concept). The ATM multiplexer inserts “idle cells” into the joint
`
`output data stream if none of the connections needs transmission capacity (Figure 2.1). In this
`
`way, the physical channel may provide just as much capacity for a connection as this connection
`
`actually needs at any given point in time. In particular, this allows a response to the dynamic
`
`communication behavior of connections with changing bandwidth requirements over time (see
`
`Table 2.1). This method called statistical multiplexing utilizes the statistical properties of data
`
`traffic to transmit larger amounts of data than synchronous methods over a physical channel with
`
`the same capacity.
`
`Transmission via ATM networks is based on virtual channel connections (VCCs). When the
`
`connection is
`
`set up,
`
`the ATM network determines a route via switches between the
`
`communicating terminals that is stored in the internal routing tables of these switches. All cells of
`
`the associated virtual connection are transmitted via the virtual channel formed in this way. The
`
`cells are transmitted based on the routing parameters contained in their headers which are used as
`
`pointers to the respective entries in the routing tables of the nodes.
`
`1 International Telecommunication Union
`
`
`
`50-250 ms
`50-250 ms
`—
`
`2 - 200
`
`Sen/ice
`
`Voice, tele hon
`)
`Video conference (low uali
`Video conference hih ualitv
`Data transmission
`* Maximum to minimum bit rate ratio
`
`64 kbit/S
`128 kbit/s
`1-10 MbiUs
`0.1 — 30 MbiUs
`
`* * Max. tolerable end-to-end delay
`*** without echo compensation
`
`Table 2.1: Characteristics of some typical services [6, 17]
`
`
`
`8
`
`2. Asynchronous TransferMoa’e, ATM
`
`
`vconnection A
`
`1
`___H L_.-
`
`
`
`
`
`Econnection B
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`[1
`'
`a
`‘
`
`
`
`ce 1 stream behinc the multiplexer
`_
`
`connection C
`‘ a
`
`connection A
`
`
`i__x_1
`k ”We:
`connection B
`Ev
`
`
`
`
`VE‘connection C
`
`See idle cell
`
`
`Figure 2.1: Time multiplex in the ATM method
`
`2.1
`
`Structure of the ATM cell and meaning of the control information
`
`The header structure depends on whether the cells are transmitted within the network (between
`ATM switches) or between the network and the user or terminal (see Fig. 2.2). Two interfaces are
`distinguished:
`
`0 User network interface (UNI)
`
`o
`
`Inter network interface (INI)
`
`
`User interface (UNI)
`
`Inter network interface (INI)
`
`gym;
`
`{3sz
`VFE
`VPI
`_ ‘3
`
`WW
`Véfit
`RP!
`1:12!
`
`Will
`‘e’CE
`
`
`
`
`
`fiEC
`
`HE:
`
`48 bytes of user data
`
`48 bytes of user data
`
`Bell Less Priority
`CL?
`{BFC Geeeric Fm»? Conéml
`HEC Heme: Error Control
`
`{Dagmar} F; pa
`PT
`WEI Virtual Chance? ifiemfifier
`VP: Virtual Path identifier
`
`UM weer-Nemom firterrace
`WI
`inter-Network Entefiam
`formerly NNI - Network Node Interface
`
`Figure 2.2: Structure of ATM cells at the UNI and INI interfaces
`
`GFC Generic Flow Control, 4 bits (UNI only)
`
`This field is used for access control of terminals at the UNI.
`
`VCI
`
`Virtual Channel Identifier, 2 bytes
`
`The reference to the Virtual channel is used to distinguish among different simultaneous
`connections and for assigning cells to connections. The VCI
`is assigned to one
`connection section only.
`
`VPI
`
`Virtual Path Identifier, 8 or 12 bits
`
`
`
`vol
`
`97
`
`or}
`
`we:
`
`9?
`
`61.3:
`
`
`
`
`
`2
`
`3
`
`4
`
`S
`
`6 - 53
`
`
`
`Identifier for Virtual paths identify a channel bundle, allowing to distinguish
`
`
`
`between routes going in the same direction that each contain multiple virtual channels.
`
`Switches will be able to identify and forward channels of the same bundle faster.
`
`PT
`
`Payload vae, 4 bits
`
`This field identifies the type of information field for distinguishing payload and signaling
`
`information. The latter is used for operations and maintenance (OAM) or for resource
`
`management (RM). A switch has to analyze both the header and the user data field in the
`
`body of the cell.
`
`CLP
`
`Cell Loss Priority, 1 bit
`
`This bit is used to identify low-priority cells that are preferably discarded in the event of an
`overload condition in the network.
`
`HEC Header Error Control, 1 byte
`
`Unlike the user data, the header of an ATM cell is protected by a check sequence. This
`
`sequence recognizes two-bit errors and corrects one-bit errors and is recalculated in each
`
`network node. In addition, HEC is used for identifying the beginning of a cell. Higher-order
`
`protocol layers are responsible for error control of user information.
`
`2.2
`
`ATM switching
`
`ATM switching is based on virtual connections. The cells contain a virtual channel identifier (VCI)
`
`and a virtual path identifier (VPI) that a switching node uses to determine the next node and enters it
`
`in the header of the cell. Each switching node only knows the respective next node. The complete
`
`origin and destination addresses are sent once during connection setup to establish these virtual
`
`connections, i.e. the routing tables in the network nodes. When a cell arrives in a network node, the
`
`information in the header of the cell (VCI and VPI) is analyzed and the next network node is
`
`determined using the routing table. This results in new values for the VP] and, depending on the
`
`switching element, for the VCI as well, and this new information is entered into the cell header. Then
`
`the cell is routed to the respective output of the switching network.
`
`Splitting the channel number into VPI and VCI requires two types of network nodes. An ATM cross
`
`connect switches the channel bundle into the respective directions based on the VP1 field. The VCI
`
`ficld remains unchanged. An ATM switch makcs virtual connections, and the VP1 and VCI are
`
`analyzed and changed in the cross connect process, see Fig. 2.3.
`
`2.3
`
`The ATM reference model
`
`Based on the 180/081 reference model, a four-layer reference model has been defined for ATM (see
`Fig. 2.4). These include the physical layer, the ATM layer, the ATM adaptation layer (AAL), and a
`layer that represents the fiinctions of higher-order layers.
`
`
`
`10
`
`2. Asynchronous transfer mode, ATM
`amt-Switch
`
`a:
`1
`a
`3
`
`..
`a f. a
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`”1.x”
`r
`
`y
`an
`T
`m‘
`“a
`‘T:
`WM F '33
`, My”
`k”
`L WEN
`J
`“.3
`Win:
`$352 1 mg;
`‘1
`"_ ‘1
`a”
`;'
`a
`, a
`r,
`“a,
`m w“ .w—
`
`“3‘5 M1
`-
`W15
`“E3
`,
`‘l‘FIZ
`if:
`(‘1
`g- mm min
`V34 I'll-J
`xx x
`“a
`“—“~-w"‘”
`x W
`was —£-‘e m '
`”1%
`.
`,
`“‘1— van
`x“
`~
`”PIE
`.
`-»
`were i! \’
`'3 x!"
`\ “
`Jab mt:
`.4F
`
`gem
`fizz
`
`I_. “‘1
`“5:
`was
`2— WM
`
`W
`Pg
`_
`.
`%
`
`It,
`1 WP”
`I
`
`Ema
`
`l wrpm
`“afF‘EE:
`
`'
`
`a
`39- wow
`x,
`
`\— vim
`L var:
`
`ATM-switching network
`
`ATM-exchange
`
`Figure 2.3: Switching through channel bundles (virtual paths) and connections in ATM-switching
`fields and ATM-
`exchanges
`
`Three different protocol planes (planes) have been implemented: the user plane, the control plane and
`the management plane. The management plan includes two functions: the plane management dealing
`with the entire system and the management of the individual layers (layer management).
`
`.
`
`V
`
`.
`
`V
`
`.
`
`V
`
`.
`
`V
`
`.
`
`we” ..........
`
`
`
`
`
`
`
`
`
`
`
` ighar layers; .igher gay-am
`
`
`
`ATM ada {ablation layer EARL} g.
`
`
`
`ATM layer
`
`
` Physical layer
`
`
`
`Plane management
`
`Layer managgment
`
`Figurer 2.4: ATM Reference model
`
`Physical layer: The physical layer is in regards to the standard ISO/OSI layer l-functions like bit
`synchronization,
`line code and monitoring functions. Their functionality is determined by the method
`of transfer.
`
`ATM-layer: The ATM-layer corresponds with the ISO/OSI-layer 3. It contains the ATM-specific
`functions for cell transport:
`
`0
`
`Control of the VP1 and VCI oriented functions for the differentiation between different
`
`connections in the ATM exchange
`Multiplexes and demultiplexes of the cells from different connections
`Creation and extraction of information in the cell head
`
`Priority control in order to minimize cell loss and waiting times
`Functions to avoid buffer overflows (congestion control)
`Connection-specific momtoring of the cell rate in accordance with the service level
`agreements with the participants (usage parameter control, UPC)
`
`
`
`2.4 Service goods in the ATM-fixed network
`
`eneric flow control (GFC) to the UNI
`
`ATM-adaptation layer: The adaptation layer (AAL) affects the service-specific requirements and
`corresponds with the ISO/0S1 layer 4. It provides services, which they provide with the help of the
`ATM-layer by executing the required segmentation of data streams (segmentation and reassembly
`sublayer, SAR) and balances delay deviations that occur with synchronous services by the ATM-
`network or allows for the recognition of cell losses by data services (convergence sublayer, CS). The
`following AAL-types were defined:
`
`AAL 1
`
`AAL 2
`
`AAL 3/4
`
`allows for synchronous transfer with constant data rates by balancing out delay
`deviations in the receiver
`
`allows for synchronous transfer with variable data rates by balancing out delay
`deviations in the receiver
`
`for data services and other non-real-time oriented services; allows for the
`recognition of cell and package losses; type 3 is connection-oriented, type 4
`allows for unconnected transfers
`
`
`
`reduced
`
`ue 3 for siHnalizin traffic
`
`Table 2.2: Types of the ATM-adaption layer
`
`2.4 Service levels in the ATM-fixed network
`
`The service levels in ATM-networks describe the connection quality, which the participant determines
`during the establishment of connection with the network provider in a so-called service level
`agreement and which must then be guaranteed by the network.
`
`The ATM-network must be able to fulfill the requirements of different services for the service levels
`(quality of service, QoS) of the ATM-layer. The lTU-T recommendation E.800 defines the service
`levels as such:
`
`“Collective effect ofservice performance which determine the degree ofsatisfaction ofa user of
`service ”
`
`In this work, however, the implied subjective impressions of the service user should not be further
`considered in the definition above. The service levels should instead be described with the help of
`network performance parameters, which are measurable and objectively comparable to the user
`interfaces. The service level requirements of a connection are described with the help of this network
`parameters, which are defined in (12) and (3):
`
`CER Cell Error Ratio:
`
`Quotient from incorrectly transferred cells and all transferred cells
`
`
`
`12
`
`CLR. Cell Loss Ratio:
`
`2. Asynchronous transfer mode, ATM
`
`Quotient from lost cells and all transferred cells
`
`CMR. Cell Misinsertion Rate:
`
`Share of cells, which were transferred to the incorrect recipients
`
`CTD mean Cell Transfer Delay;
`Arithmetic method for the end-to-end-cell transfer duration
`
`maxCTD maximum Cell Transfer Delay:
`Cells whose transfer delay exceeds maxCTD are evaluated as lost
`
`CDV Cell Delay Variation:
`Deviation of the end-to-end-cell transfer duration
`
`In order to be able to satisfy the traffic characteristics from different sources, four service classes were defined
`on the ATM-layer, which are different in their requirements for service levels, see also table 2.3:
`
`CBR. Constant Bit Rate
`
`The sources of these service classes transmit with a constant bit rate, which corresponds with the
`maximum bit rate, which was negotiated at the time of the connection establishment. The source has
`real-time requirements for the network and requires compliance of an upper limit for the maximum
`delay of their cells through the network. (For example: speech transmission)
`
`VBR. Variable Bit Rate
`
`VBR-sources transmit with variable bit rate. A difference is made between real-time oriented (real-
`time, RT) services with a defined maximum cell delay and less critical (non real-time, nRT) services).
`(example: video transmission)
`
`ANR. Available Bit Rate
`
`In this service class, the traffic sources adapt their bit rate dynamically to the available open resources.
`No requirements are placed on the real-time behavior and delay. Only the cell loss probability is
`defined as a parameter for the service level. (Example: data transmission)
`
`UBR. Unspecified bit Rate
`This service class does not offer any traffic-related service guarantees. Neither the cell loss probability
`nor a limit for the maximum delay are guaranteed (Best Effort).
`
`ATM Service Class
`
`-- W WW
`
`
`
`Undefined Undefined Undefined
`Defined
`Defined
`
`
`Table 2.3: ATM Service Classes and their Service Level Parameters
`
`
`
`2.4 Service Levels in the ATM-Fixed Network
`
`13
`
`The monitoring and control of the service levels requires functions in end devices, access and
`switching nodes, which are summarized under the general topic of Traffic Management (3).
`
`The connection admission control (CAC) checks before the admission of a desired connection to see if
`the authorization of another connection is possible without hindering the existing connections (14, 32).
`The CAC can only work correctly if the traffic of a connection does not exceed the parameters agreed
`upon in the service level contract. Thus, a monitoring of the connection parameters occurs on the edge
`of the network in order to take immediate countermeasures in the case of an exceeding the thresholds,
`i.e. through the discarding of all cells, which are contributing to exceeding the parameters (Usage
`Parameter Control (UPC), policing frmction).
`
`In ATM-multiplexes and ATM-exchanges, ATM-cells with different connections compete with each
`other for the capacity of the outgoing multiplex lines. The processing of the cells waiting in the buffers
`occurs through a control unit, called a scheduler, which is based on an ideal operating strategy. The
`task of the scheduler is to allocate the transfer capacity of the multiplex lines corresponding with the
`requirements from the different connections.
`
`
`
`14
`
`2. Asynchronous transfer mode, ATM
`
`
`
`CHAPTER 3
`
`Architecture of the ATM Wireless Interface
`
`In this chapter, initially the architecture of the ATM-wireless interface between the wireless or mobile
`ATM-terminal (wireless terminal, WT) and the ATM-fixed network is introduced; after this, the
`structure of the protocol plane will be explained and the areas will be characterized in which the
`measures being investigated in this work to support the service levels are used.
`
`Image 3.1 schematically shows the structure of the basic mobile wireless network (23, 24, 26). The
`access points to the ATM-fixed network form base stations, which are established from one or multiple
`transceivers and a base station controller (B SC), which connects the base stations with the ATM-fixed
`network and executes the protocols of the base station.
`
`3
`
`
`
`I ATM-Network
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`ATM-Cellular Radio
`
`:ATM—Wireless Interface
`l
`
`Exchange
`
`
`Figure 3.1: Architecture of a cellular ATM-cellular radio network
`
`Such a function allows for wireless ATM-access in select areas, i.e. in buildings, in open areas or near
`buildings. Handover functions allow for the free mobility of the terminals within their areas of supply.
`Due to the architecture of this wireless network, there are two types of handovers, see figure 3.2:
`
`Wireless handover: The wireless handover occurs between two transceivers from the same base
`
`station. The switching of virtual connections is executed within the base station controller and is
`independent of the ATM-fixed network.
`Network handover: The network handover occurs between two transceivers from different base
`
`stations and requires the rerouting of Virtual connections within the ATM-network. A special ATM-
`mobile radio exchange is required for this corresponding with the mobility functions.
`
`
`
`16
`
`Wireless handover
`
`3. Architecture of the ATM Wireless Interface
`Network handover
`
`ATM network
`——————
`f
`
`xl/ My
`/
`ATM mobile
`ireless
`
`.W
`
`
`
`M /
`3’3“
`
`Emi‘1‘“
`
`,
`
`\
`
`
`l Wnetwork
`Ell,
`
`r’m“
`ATM
`
`NJ
`
`
`
`my“
`i
`
`Tax ’
`
`83::
`
`l
`
`
`
`C
`
`‘
`
`IR;
`
`~.
`x
`
`
`
`exchange
`
`AAL
`_
`
`firm
`_
`ML
`\
`.
`__
`ware,
`AW
`A"h'l
`
`
`MM
`W—LLC Aim
`w-me-
`will:
`
`W-PHY
`PH‘?’
`Al’s-MAC,
`W—MAC
`
`
`
`
`
`
`
`
`
`WLPHY
`W—PHY
`PHY
`PHY PH”?
`PH‘X’
`
`
`
`
`
`
`. N —
`
`
`
`
`
`I
`W—LLC
`
`
`
`
`W—MAC W—MAC
`
`
`
`
`
`
`W-PHY W-F‘H‘s’l
`
`
` ]
`
`
`
`
`
`A"Wil.a’r \
`+~m
`
`A’ML’WW
`
`Figure 3.2: Handover to the ATM-wireless interface (wireless handover) and in the ATM-fixed
`network (network handover)
`
`Within the scope of this work, the handover is not discussed in further detail. Detailed investigations
`about the network architectures and protocols for handover execution are found, for example, in (7,
`3 5).
`
`The technical data from the wireless interface are oriented to the recommendations of the wireless
`
`ATM group (RESlO WAG) of the European Telecommunications Standards Institute (ETSI) (26) and
`are summarized in table 3.1
`
`5.2 GHz
`
`Multiplex procedure
`FDM
`Access rocedure
`TDMA
`
`
`
`TDD
`Duplex procedure
`23.5 MHz
`Bandwidth of a freuenc channel (carrier)
`50,000 ATM cells/s (= 20 Mbit / s)
`Gross data rate on a channel
`Table 3.1: Technical data from the ATM wireless interface
`
`Aspects of the frequency planning and procedure for the dynamic channel assignment are not
`considered. This allows for the isolated observation of individual wireless cells with a central base
`
`station and multiple wireless or mobile terminals.
`
`3.1 The wireless cells as distributed ATM-multiplexers
`
`A typical area of use for cellular ATM mobile wireless networks includes wireless local networks
`(Wireless Local Area Networks, W-LAN). Here it is desirable that through the wireless terminal in the
`area, it can provide the same services despite is possibilities (limited operating duration through battery
`controlled energy supply and lower data rates due to wireless transmission) as through ATM-tenninals
`with a fixed network connection. In particular, all available ATM applications should be usable
`
`
`
`without changes, this means both can be used in Wireless and wired terminals for the same services of
`the AAL.
`
`
`
`3.1 The wireless cells as distributed ATM multiplexers
`
`17
`
`Virtual ATM-
`
`multiplexer of a
`wireless cell
`
`,x'
`
`[fr/
`
`
`
`.
`
`mobile ATM-termlnal
`
`.
`
`t:::
`ATM—
`
`‘
`
`i
`.
`
`l
`,
`
`i
`
`, g" _
`
`_
`
`,
`
`
`application
`\ii
`
`
`a.
`X
`. XX
`xx
`"xx
`*1.
`Hun I'- Vii—[M
`.m;
`In mu nu .4
`.‘fi‘
`3%..
`ft}? Wag
`,
`j
`l
`
`, .ww
`(x,fATM-fixed network
`if“.
`
`[if i- .
`
`mm..-
`”Wren“,
`
`
`
`
`
`4
`"
`i cilfi‘»9‘