CONTENTS

 1     Scope
 2     References
 3     Abbreviations
 4     Performance model
        4.1     Measurement points and connection portions
        4.2     Cell-based ATM reference events
        4.3     Frame-based ATM reference events
        4.4     Layered nature of B-ISDN performance
 5     ATM cell transfer outcomes
        5.1     Successful cell transfer outcome
        5.2     Tagged cell outcome
        5.3     Errored cell outcome
        5.4     Lost cell outcome
        5.5     Misinserted cell outcome
        5.6     Severely errored cell block outcome
        5.7     Successful and corrupted frame outcomes
 6     ATM performance parameters
        6.1     Cell error ratio
        6.2     Cell loss ratio
                  6.2.1     CLR₀
                  6.2.2     CLR₀₊₁
                  6.2.3     CLR₁
        6.3     Cell misinsertion rate
        6.4     Severely errored cell block ratio
        6.5     Cell transfer delay
                  6.5.1     Mean cell transfer delay
                  6.5.2     Cell delay variation
        6.6     Cell flow related parameters
        6.7     Frame-based ATM parameters
 7     Network performance when some cells are non-conforming
        7.1     A method for computing the number of non-conforming cells
        7.2     Upper bounding the number of non-conforming cells
                  7.2.1     Bounding cell non-conformance for a single peak cell rate specification
                  7.2.2     Bounding cell non-conformance for independent peak cell rate specifications
                  7.2.3     Bounding cell non-conformance for dependent, coordinated cell rate specifications defined on the aggregate cell stream
                  7.2.4     Bounding cell non-conformance for dependent, coordinated cell rate specifications defined on the aggregate and on the CLP = 0 substream
        7.3     Adjusted CLR performance when there are non-conforming cells
                  7.3.1     CLR when the conformance specification applies to a single cell stream
                  7.3.2     CLR when the conformance specification applies to the aggregate and the CLP = 0 substream
        7.4     Non-conforming cells and the severely errored cell block outcome
 8     Network performance objectives
        8.1     General discussion of per-connection QoS
        8.2     QoS classes
                  8.2.1     Nature of the network performance objectives
                  8.2.2     Statistical estimation issues
                  8.2.3     Unbounded (unspecified) performance
                  8.2.4     Default values for cell error ratio, cell misinsertion rate, and severely errored cell block ratio
                  8.2.5     Association of QoS classes with ATM transfer capabilities
        8.3     Alternative QoS negotiation procedures
 9     Allocation of the performance objectives
        9.1     General principles of allocation
        9.2     Route length calculation
        9.3     Allocation of the QoS class 1 and QoS class 5 CTD objectives
        9.4     Allocation of the QoS class 1 CDV objective
        9.5     Allocation of the QoS class 5 CDV objective
        9.6     Allocation of the SECBR and CER objectives
        9.7     Allocation of the QoS class 1 and QoS class 5 CLR objectives
        9.8     Allocation of the QoS class 2 and QoS class 3 CLR objectives
        9.9     Allocation of the CMR objective
       9.10     Concatenation of QoS values
Annex A - Relationship between ATM layer NP and the NP of AAL type 1 for CBR services
        A.1     Possible AAL functions and their effects
                  A.1.1     Lost cell outcome and misinserted cell outcome
                  A.1.2     Errored cell outcome
                  A.1.3     Cell transfer delay
        A.2     Bounding relationships between NP parameters and binary errors
Annex B - Cell transfer delay, 1-point CDV, and 2-point CDV characteristics
        B.1     Components of delay associated with ATM-based user information transfer
        B.2     Relationship between cell clumping and cell queues
        B.3     1-point CDV and non-conformance
        B.4     Relationship between cell transfer delay and cell loss in a single shared buffer
Annex C - Cell transfer performance measurement methods
        C.1     Performance measurements and availability
        C.2     General aspects of performance monitoring using OAM cells
        C.3     General aspects of out-of-service performance measurement
        C.4     Cell error ratio
        C.5     Cell loss ratio
        C.6     Cell misinsertion rate
        C.7     Severely errored cell block ratio
        C.8     Cell transfer delay
        C.9     Cell delay variation
       C.10     Estimation of CLR and SECBR in case of lost FM OAM cells
Annex D - Signalling QoS Preferences
        D.1     Signalling preferences for a QoS class
        D.2     Signalling of individual parameters
                  D.2.1     Interpretation of the fields that contain supplementary information about individual parameters
                  D.2.2     Limits on the applicability of fields containing supplementary information about individual parameters
Appendix I - Assessing the performance of a UPC/NPC mechanism
Appendix II - Hypothetical reference connections for validating the ATM performance objectives
       II.1     Number ATM nodes in the HRXs
       II.2     Switching speeds in the HRXs
       II.3     Loading within the HRXs
       II.4     Geostationary satellites within the HRXs
       II.5     Other aspects of the HRXs
Appendix III - Example applications of the allocation rules of 9.6, 9.7 and 9.8
Appendix IV - Interworking with downstream ATM Forum domains
       IV.1     ATM Forum service categories
       IV.2     ATM Forum fields for signalling of individual performance parameters