draft-ietf-ippm-loss-pattern-05.txt   draft-ietf-ippm-loss-pattern-06.txt 
IP Performance Metrics (IPPM) WG Rajeev Koodli IPPM Working Group Rajeev Koodli
INTERNET DRAFT Nokia Research Center INTERNET DRAFT Nokia Research Center
20 July 2001 R. Ravikanth 5 December 2001 R. Ravikanth
Axiowave Axiowave
One-way Loss Pattern Sample Metrics One-way Loss Pattern Sample Metrics
<draft-ietf-ippm-loss-pattern-05.txt> draft-ietf-ippm-loss-pattern-06.txt
STATUS OF THIS MEMO Status of This Memo
This document is an Internet-Draft and is in full conformance with all This document is an Internet-Draft and is in full conformance with
provisions of Section 10 of RFC2026. all provisions of Section 10 of RFC2026.
Internet-Drafts are working documents of the Internet Engineering Task Internet-Drafts are working documents of the Internet Engineering
Force (IETF), its areas, and its working groups. Note that other groups Task Force (IETF), its areas, and its working groups. Note that
may also distribute working documents as Internet- Drafts. other groups may also distribute working documents as Internet-
Drafts.
Internet-Drafts are draft documents valid for a maximum of six months Internet-Drafts are draft documents valid for a maximum of six
and may be updated, replaced, or obsoleted by other documents at any months and may be updated, replaced, or obsoleted by other documents
time. It is inappropriate to use Internet-Drafts as reference material at any time. It is inappropriate to use Internet-Drafts as
or to cite them other than as "work in progress." reference material or to cite them other than as "work in progress."
The list of current Internet-Drafts can be accessed at The list of current Internet-Drafts can be accessed at:
http://www.ietf.org/ietf/1id-abstracts.txt http://www.ietf.org/ietf/1id-abstracts.txt
The list of Internet-Draft Shadow Directories can be accessed at:
The list of Internet-Draft shadow directories can be accessed at http://www.ietf.org/shadow.html.
http://www.ietf.org/shadow.html
This memo provides information for the Internet community. This memo This memo provides information for the Internet community. This memo
does not specify an Internet standard of any kind. Distribution of does not specify an Internet standard of any kind. Distribution of
this memo is unlimited. this memo is unlimited.
Abstract Abstract
The Internet exhibits certain specific types of behavior (e.g., bursty The Internet exhibits certain specific types of behavior (e.g.,
packet loss) that can affect the performance seen by the users as bursty packet loss) that can affect the performance seen by the users
well as the operators. Currently, the focus has been on specifying as well as the operators. Previously, the focus of the IPPM had
base metrics such as delay, loss and connectivity under the been on specifying base metrics such as delay, loss and connectivity
framework described in [frame-work]. It is useful to capture under the framework described in [10]. However, specific Internet
specific Internet behaviors under the umbrella of IPPM framework, behaviors can also be captured under the umbrella of IPPM framework,
specifying new concepts while reusing existing guidelines as much as specifying new concepts while reusing existing guidelines as much as
possible. This draft proposes the use of "derived metrics" to possible. This document defines metrics derived from the previously
accomplish this, specifically providing means for capturing the loss specified base metrics to capture loss patterns experienced by
pattern on the Internet. streams of packets on the Internet.
Contents
Status of This Memo i
Abstract i
1. Introduction 2
2. Terminology 2
3. The Approach 2
4. Basic Definitions 3
5. Definitions for Samples of One-way Loss Distance, and One-way
Loss Period 4
5.1. Metric Names . . . . . . . . . . . . . . . . . . . . . . 4
5.1.1. Type-P-One-Way-Loss-Distance-Stream . . . . . . . 4
5.1.2. Type-P-One-Way-Loss-Period-Stream . . . . . . . . 4
5.2. Metric Parameters . . . . . . . . . . . . . . . . . . . . 4
5.3. Metric Units . . . . . . . . . . . . . . . . . . . . . . 4
5.3.1. Type-P-One-Way-Loss-Distance-Stream . . . . . . . 4
5.3.2. Type-P-One-Way-Loss-Period-Stream . . . . . . . . 4
5.4. Definitions . . . . . . . . . . . . . . . . . . . . . . . 5
5.4.1. Type-P-One-Way-Loss-Distance-Stream . . . . . . . 5
5.4.2. Type-P-One-Way-Loss-Period-Stream . . . . . . . . 5
5.4.3. Examples . . . . . . . . . . . . . . . . . . . . 5
5.5. Methodologies . . . . . . . . . . . . . . . . . . . . . . 6
5.6. Discussion . . . . . . . . . . . . . . . . . . . . . . . 7
5.7. Sampling Considerations . . . . . . . . . . . . . . . . . 7
5.8. Errors and Uncertainties . . . . . . . . . . . . . . . . 7
6. Statistics 8
6.1. Type-P-One-Way-Loss-Noticeable-Rate . . . . . . . . . . . 8
6.2. Type-P-One-Way-Loss-Period-Total . . . . . . . . . . . . 8
6.3. Type-P-One-Way-Loss-Period-Lengths . . . . . . . . . . . 9
6.4. Type-P-One-Way-Inter-Loss-Period-Lengths . . . . . . . . 9
6.5. Examples . . . . . . . . . . . . . . . . . . . . . . . . 9
7. Security Considerations: 10
8. IANA Considerations 11
9. Acknowledgements 11
Addresses 13
1. Introduction 1. Introduction
In certain real-time applications (such as packet voice and video),
the loss pattern or loss distribution is a key parameter In certain real-time applications (such as packet voice and
that determines the performance observed by the users. For the same video), the loss pattern or loss distribution is a key parameter
loss rate, two different loss distributions could potentially produce that determines the performance observed by the users. For the
widely different perceptions of performance. The impact of loss pattern same loss rate, two different loss distributions could potentially
is also extremely important for non-real-time applications that use produce widely different perceptions of performance. The impact
an adaptive protocol such as TCP. There is ample evidence in the of loss pattern is also extremely important for non-real-time
literature indicating the importance and existence of loss burstiness applications that use an adaptive protocol such as TCP. Refer
and its effect on packet voice and video applications to [2], [3], [5], [12] for evidence as to the importance and
[Bolot], [Borella], [Handley], [Yajnik]. existence of loss burstiness and its effect on packet voice and video
applications.
In this document, we propose two derived metrics, called "loss distance" In this document, we propose two derived metrics, called "loss
and "loss period", with associated statistics, to capture packet loss distance" and "loss period", with associated statistics, to capture
patterns. The loss period metric captures the frequency and length packet loss patterns. The loss period metric captures the frequency
(burstiness) of loss once it starts, and the loss distance metric and length (burstiness) of loss once it starts, and the loss
captures the spacing between the loss periods. It is important to note distance metric captures the spacing between the loss periods. It is
that these metrics are derived based on the base metric important to note that these metrics are derived based on the base
Type-P-One-Way-packet-Loss. metric Type-P-One-Way-packet-Loss.
2. The Approach 2. Terminology
This document closely follows the guidelines specified in [frame-work]. The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL
Specifically, the concepts of "singleton, sample, statistic", NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", "OPTIONAL", and
"silently ignore" in this document are to be interpreted as described
in RFC 2119 [4].
3. The Approach
This document closely follows the guidelines specified in [10].
Specifically, the concepts of singleton, sample, statistic,
measurement principles, Type-P packets, as well as standard-formed measurement principles, Type-P packets, as well as standard-formed
packets all apply. However, since the draft proposes to capture packets all apply. However, since the draft proposes to capture
specific Internet behaviors, modifications to the sampling process specific Internet behaviors, modifications to the sampling process
may be needed. Indeed, this is mentioned in [AKZ], where it is noted MAY be needed. Indeed, this is mentioned in [1], where it is
that alternate sampling procedures may be useful depending on specific noted that alternate sampling procedures may be useful depending
circumstances. This draft proposes that the specific behaviors be on specific circumstances. This draft proposes that the specific
captured as "derived" metrics from the base metrics the behaviors behaviors be captured as "derived" metrics from the base metrics the
are related to. The reasons for adopting this position are the behaviors are related to. The reasons for adopting this position are
following the following:
- it provides consistent usage of singleton metric definition for - it provides consistent usage of singleton metric definition for
different behaviors (e.g., a single definition of packet loss different behaviors (e.g., a single definition of packet loss is
is needed for capturing burst of losses, 'm out of n' losses needed for capturing burst of losses, 'm out of n' losses etc.)
etc. Otherwise, the metrics would have to be fundamentally
different)
- it allows re-use of the methodologies specified for the singleton - it allows re-use of the methodologies specified for the singleton
metric with modifications whenever necessary metric with modifications whenever necessary
- it clearly separates few base metrics from many Internet behaviors
Following the guidelines in [frame-work], this - it clearly separates few base metrics from many Internet
translates to deriving sample metrics from the respective behaviors
singletons. The process of deriving sample metrics from the singletons
is specified in [frame-work], [AKZ], and others. Following the guidelines in [10], this translates to deriving
sample metrics from the respective singletons. The process
of deriving sample metrics from the singletons is specified
in [10], [1], and others.
In the following sections, we apply this approach to a particular In the following sections, we apply this approach to a particular
Internet behavior, namely the packet loss process. Internet behavior, namely the packet loss process.
3. Basic Definitions: 4. Basic Definitions
3.1. Bursty loss:
The loss involving consecutive packets of a stream.
3.2. Loss Distance:
The difference in sequence numbers of two successively lost Sequence number: Consecutive packets in a time series sample
packets which may or may not be separated by successfully are given sequence numbers that are consecutive
received packets. integers. This document does not specify exactly
how to associate sequence numbers with packets. The
sequence numbers could be contained within test
packets themselves, or they could be derived through
post-processing of the sample.
Example. Let packet with sequence number 50 be considered lost Bursty loss: The loss involving consecutive packets of a stream.
immediately after packet with sequence number 20 was
considered lost. The loss distance is 30.
Note that this definition does not specify exactly how to Loss Distance: The difference in sequence numbers of two
associate sequence numbers with test packets. In other words, from successively lost packets which may or may not be
a timeseries sample of test packets, one may derive the sequence separated by successfully received packets.
numbers. However, these sequence numbers must be consecutive
integers.
3.3. Loss period: Example: In a packet stream, the packet with sequence number 20
is considered lost, followed by the packet with
sequence number 50. The loss distance is 30.
Let P_i be the i'th packet. Loss period: Let P_i be the i'th packet. Define f(P_i) = 1 if P_i
Define f(P_i) = 1 if P_i is lost, 0 otherwise. is lost, 0 otherwise. Then, a loss period begins if
Then, a loss period begins if f(P_i) = 1 and f(P_(i-1)) = 0 f(P_i) = 1 and f(P_(i-1)) = 0
Example. Consider the following sequence of lost (denoted by x) Example: Consider the following sequence of lost (denoted by x)
and received (denoted by r) packets. and received (denoted by r) packets.
r r r x r r x x x r x r r x x x r r r x r r x x x r x r r x x x
Then, with i assigned as follows Then, with `i' assigned as follows,
1 1 1 1 1 1 1 1 1 1 1 1
i: 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 i: 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5
f(P_i) is, f(P_i) is,
f(P_i): 0 0 0 1 0 0 1 1 1 0 1 0 0 1 1 1 f(P_i): 0 0 0 1 0 0 1 1 1 0 1 0 0 1 1 1
and there are four loss periods in the above sequence and there are four loss periods in the above sequence
beginning at P_3, P_6, P_10, and P_13. beginning at P_3, P_6, P_10, and P_13.
4. Definitions for Samples of One-way Loss Distance, 5. Definitions for Samples of One-way Loss Distance, and One-way
and One-way Loss Period. Loss Period
4.1 Metric Name: 5.1. Metric Names
4.1.1 Type-P-One-Way-Loss-Distance-Stream 5.1.1. Type-P-One-Way-Loss-Distance-Stream
4.1.2 Type-P-One-Way-Loss-Period-Stream
4.2 Metric Parameters 5.1.2. Type-P-One-Way-Loss-Period-Stream
+ Src, the IP address of a host 5.2. Metric Parameters
+ Dst, the IP address of a host
+ T0, a time
+ Tf, a time
+ lambda, a rate in reciprocal of seconds
+ Path, the path from Src to Dst (See [AKZ] for comments)
4.3 Metric Units Src, the IP address of a host
4.3.1 Type-P-One-Way-Loss-Distance-Stream: Dst, the IP address of a host
A sequence of pairs of the form <loss distance, loss>, where loss T0, a time
is derived from the sequence of <time, loss> in [AKZ], and loss
Tf, a time
lambda, a rate of any sampling method chosen in reciprocal of
seconds
5.3. Metric Units
5.3.1. Type-P-One-Way-Loss-Distance-Stream
A sequence of pairs of the form <loss distance, loss>, where
loss is derived from the sequence of <time, loss> in [1], and loss
distance is either zero or a positive integer. distance is either zero or a positive integer.
4.3.2 Type-P-One-Way-Loss-Period-Stream 5.3.2. Type-P-One-Way-Loss-Period-Stream
A sequence of pairs of the form <loss period, loss>, where loss is A sequence of pairs of the form <loss period, loss>, where loss is
derived from the sequence of <time, loss> in [AKZ], and loss period derived from the sequence of <time, loss> in [1], and loss period is
an integer. an integer.
4.4. Definitions: 5.4. Definitions
4.4.1 Type-P-One-Way-Loss-Distance-Stream 5.4.1. Type-P-One-Way-Loss-Distance-Stream
When a packet is considered lost (using the definition in [AKZ]), we When a packet is considered lost (using the definition in [1]),
look at its sequence number and compare it with that of the we look at its sequence number and compare it with that of the
previously lost packet. The difference is the loss distance between previously lost packet. The difference is the loss distance between
the lost packet and the previously lost packet. The sample would the lost packet and the previously lost packet. The sample would
consist of <loss distance, loss> pairs. This definition assumes that consist of <loss distance, loss> pairs. This definition assumes that
sequence numbers of successive test packets increase monotonically by sequence numbers of successive test packets increase monotonically by
one. The loss distance associated with the very first packet loss is one. The loss distance associated with the very first packet loss is
considered to be zero. considered to be zero.
The sequence number of a test packet can be derived from the timeseries The sequence number of a test packet can be derived from the
sample collected by performing the loss measurement according to the timeseries sample collected by performing the loss measurement
methodology in [AKZ]. For example, if a loss sample consists of according to the methodology in [1]. For example, if a loss sample
{<T0,0>, <T1,0>, <T2,1>, <T3,0>, <T4,0>}, the sequence numbers of the consists of <T0,0>, <T1,0>, <T2,1>, <T3,0>, <T4,0>, the sequence
five test packets sent at T0, T1, T2, T3, and T4 can be 0, 1, 2, 3 and numbers of the five test packets sent at T0, T1, T2, T3, and T4 can
4 respectively, or 100, 101, 102, 103 and 104 respectively, etc. be 0, 1, 2, 3 and 4 respectively, or 100, 101, 102, 103 and 104
respectively, etc.
4.4.2 Type-P-One-Way-Loss-Period-Stream 5.4.2. Type-P-One-Way-Loss-Period-Stream
We start a counter 'n' at an initial value of zero. This counter is We start a counter 'n' at an initial value of zero. This
incremented by one each time a lost packet satisfies the Definition 3.3. counter is incremented by one each time a lost packet satisfies the
The metric is defined as <loss period, loss> where definition outlined in 4. The metric is defined as <loss period,
"loss" is derived from the sequence of <time, loss> in loss> where "loss" is derived from the sequence of <time, loss> in
Type-P-One-Way-Loss-Stream [AKZ], and Type-P-One-Way-Loss-Stream [1], and loss period is set to zero when
loss period is set to zero when "loss" is zero in "loss" is zero in Type-P-One-Way-Loss-Stream, and loss period is set
Type-P-One-Way-Loss-Stream, and loss period is set to 'n' (above) to 'n' (above) when "loss" is one in Type-P-One-Way-Loss-Stream.
when "loss" is one in Type-P-One-Way-Loss-Stream.
Essentially, when a packet is lost, the current value of "n" indicates Essentially, when a packet is lost, the current value of "n"
the loss period to which this packet belongs. For a packet that is indicates the loss period to which this packet belongs. For a packet
received successfully, the loss period is defined to be zero. that is received successfully, the loss period is defined to be zero.
4.4.3 Example: 5.4.3. Examples
Let the following set of pairs represent a Type-P-One-Way-Loss-Stream. Let the following set of pairs represent a Type-P-One-Way-Loss-Stream.
{<T1,0>,<T2,1>,<T3,0>,<T4,0>,<T5,1>,<T6,0>,<T7,1>,<T8,0>,<T9,1>, {<T1,0>,<T2,1>,<T3,0>,<T4,0>,<T5,1>,<T6,0>,<T7,1>,<T8,0>,<T9,1>,<T10,1>}
<T10,1>}
where T1, T2,..,T10 are in increasing order. where T1, T2,..,T10 are in increasing order.
Packets sent at T2, T5, T7, T9, T10 are lost. The two derived metrics Packets sent at T2, T5, T7, T9, T10 are lost. The two derived
can be obtained from this sample as follows. metrics can be obtained from this sample as follows.
(i) Type-P-One-Way-Loss-Distance-Stream: (i) Type-P-One-Way-Loss-Distance-Stream:
Since packet 2 is the first lost packet, the associated loss distance Since packet 2 is the first lost packet, the associated loss
is zero. For the next lost packet (packet 5), loss distance is 5-2 or 3. distance is zero. For the next lost packet (packet 5), loss distance
Similarly, for the remaining lost packets (packets 7, 9, and 10) their is 5-2 or 3. Similarly, for the remaining lost packets (packets
loss distances are 2, 2, and 1 respectively. Therefore, the 7, 9, and 10) their loss distances are 2, 2, and 1 respectively.
Type-P-One-Way-Loss-Distance-Stream is: Therefore, the Type-P-One-Way-Loss-Distance-Stream is:
{<0,0>,<0,1>,<0,0>,<0,0>,<3,1>,<0,0>,<2,1>,<0,0>,<2,1>,<1,1>} {<0,0>,<0,1>,<0,0>,<0,0>,<3,1>,<0,0>,<2,1>,<0,0>,<2,1>,<1,1>}
(ii) The Type-P-One-Way-Loss-Period-Stream: (ii) The Type-P-One-Way-Loss-Period-Stream:
The packet 2 sets the counter 'n' to 1, which is incremented by one The packet 2 sets the counter 'n' to 1, which is incremented
for packets 5, 7 and 9 according to Definition 3.3. However, for by one for packets 5, 7 and 9 according to the definition in 4.
packet 10, the counter remains at 4 satisfying Definition 3.3 again. However, for packet 10, the counter remains at 4, again satisfying
Thus, the Type-P-One-Way-Loss-Period-Stream is: the definition in 4. Thus, the Type-P-One-Way-Loss-Period-Stream is:
{<0,0>,<1,1>,<0,0>,<0,0>,<2,1>,<0,0>,<3,1>,<0,0>,<4,1>,<4,1>} {<0,0>,<1,1>,<0,0>,<0,0>,<2,1>,<0,0>,<3,1>,<0,0>,<4,1>,<4,1>}
4.5. Methodologies: 5.5. Methodologies
The same methodology outlined in [AKZ] can be used to conduct the The same methodology outlined in [1] can be used to conduct the
sample experiments. sample experiments. A synopsis is listed below.
4.6 Discussion: Generally, for a given Type-P, one possible methodology would
proceed as follows:
- Arrange that Src and Dst have clocks that are synchronized with
each other. The degree of synchronization is a parameter of the
methodology, and depends on the threshold used to determine loss
(see below).
- At the Src host, select Src and Dst IP addresses, and form a test
packet of Type-P with these addresses.
- At the Dst host, arrange to receive the packet.
- At the Src host, place a timestamp in the prepared Type-P packet,
and send it towards Dst.
- If the packet arrives within a reasonable period of time, the
one-way packet-loss is taken to be zero.
- If the packet fails to arrive within a reasonable period of
time, the one-way packet-loss is taken to be one. Note that the
threshold of "reasonable" here is a parameter of the methodology.
5.6. Discussion
The Loss-Distance-Stream metric allows one to study the separation The Loss-Distance-Stream metric allows one to study the separation
between packet losses. This could be useful in determining a between packet losses. This could be useful in determining a "spread
"spread factor" associated with the packet loss rate. For factor" associated with the packet loss rate. In conjunction, the
example, for a given packet loss rate, this metric Loss-Period-Stream metric allows the study of loss burstiness for
indicates how the losses are spread. On the other hand, each occurrence of loss. A single loss period of length 'n' can
the Loss-Period-Stream metric allows the study of loss burstiness account for a significant portion of the overall loss rate. Note
for each occurrence of loss. Note that a single loss period of that it is possible to measure distance between loss bursts separated
length 'n' can account for a significant portion of the overall by one or more successfully received packets. (Refer to Sections 6.4
loss rate. Note also that it is possible to measure distance between and 6.5).
loss bursts seprated by one or more successfully received packets: See
Section 5.4, and 5.5
4.7 Sampling Considerations: 5.7. Sampling Considerations
The proposed metrics can be used independent of the The proposed metrics can be used independent of the particular
particular sampling method used. We note that Poisson sampling sampling method used. We note that Poisson sampling may not
may not yield appropriate values for these metrics for yield appropriate values for these metrics for certain real-time
certain real-time applications such as voice over IP, as well as to applications such as voice over IP, as well as to TCP-based
TCP-based applications. For real-time applications, it may be more applications. For real-time applications, it may be more appropriate
appropriate to use the ON-OFF [Sriram] model, in which an ON period to use the ON-OFF [11] model, in which an ON period starts with
starts with certain probability 'p', during which certain number of certain probability 'p', during which certain number of packets
packets are transmitted with mean 'lambda-on' according to geometric are transmitted with mean 'lambda-on' according to geometric
distribution and an OFF period starts with probability '1-p' and distribution and an OFF period starts with probability '1-p' and
lasts for a period of time based on exponential distribution with lasts for a period of time based on exponential distribution with
rate 'lambda-off'. rate 'lambda-off'.
For TCP-based applications, one may use the model proposed in For TCP-based applications, one may use the model proposed in [7].
[Padhye1]. See [Padhye2] for an application of the model. See [8] for an application of the model.
5. Statistics: 5.8. Errors and Uncertainties
5.1 Type-P-One-Way-Loss-Noticeable-Rate The measurement aspects, including the packet size, loss
threshold, type of the test machine chosen etc, invariably influence
the packet loss metric itself and hence the derived metrics described
in this document. Thus, when making assessment of the results
pertaining to the metrics outlined in this document, attention must
be paid to these matters. See [1] for a detailed consideration of
errors and uncertainties regarding the measurement of base packet
loss metric.
Define loss of a packet to be "noticeable" [RK97] if the distance 6. Statistics
between the lost packet and the previously lost packet is no
greater than delta, a positive integer, where delta is the
"loss constraint".
Example. Let delta = 99. Let us assume that packet 50 is lost 6.1. Type-P-One-Way-Loss-Noticeable-Rate
followed by a bursty loss of length 3 starting from
packet 125.
All the *four* losses are noticeable.
Given a Type-P-One-Way-Loss-Distance-Stream, this statistic Define loss of a packet to be "noticeable" [6] if the distance
can be computed simply as the number of losses that violate some between the lost packet and the previously lost packet is no greater
constraint delta, divided by the number of losses. (Alternately, it than delta, a positive integer, where delta is the "loss constraint".
can also be defined as the number of "noticeable losses" to the number
of successfully received packets). This statistic is useful when the Example: Let delta = 99. Let us assume that packet 50 is lost
followed by a bursty loss of length 3 starting from packet 125. All
the three losses starting from packet 125 are noticeable.
Given a Type-P-One-Way-Loss-Distance-Stream, this statistic can be
computed simply as the number of losses that violate some constraint
delta, divided by the number of losses. (Alternatively, it can also
be defined as the number of "noticeable losses" to the number of
successfully received packets). This statistic is useful when the
actual distance between successive losses is important. For example, actual distance between successive losses is important. For example,
many multimedia codecs can sustain losses by "concealing" the effect many multimedia codecs can sustain losses by "concealing" the effect
of loss by making use of past history information. Their ability to of loss by making use of past history information. Their ability to
do so degrades with poor history resulting from losses separated by do so degrades with poor history resulting from losses separated by
close distances. By chosing delta based on this sensitivity, one can close distances. By choosing delta based on this sensitivity, one
measure how "noticeable" a loss might be for quality purposes. can measure how "noticeable" a loss might be for quality purposes.
The noticeable loss requires a certain "spread factor" for losses The noticeable loss requires a certain "spread factor" for losses
in the timeseries. In the above example where loss constraint is equal in the timeseries. In the above example where loss constraint is
to 99, a loss rate of one percent with a spread of 100 between equal to 99, a loss rate of one percent with a spread of 100 between
losses (e.g., 100, 200, 300, 400, 500 out of 500 packets) may be more losses (e.g., 100, 200, 300, 400, 500 out of 500 packets) may be more
desirable for some applications compared to the same loss rate with a desirable for some applications compared to the same loss rate with a
spread that violates the loss constraint spread that violates the loss constraint (e.g., 100, 175, 275, 290,
(e.g., 100, 175, 275, 290, 400: losses occuring at 175 and 290 400: losses occurring at 175 and 290 violate delta = 99).
violate delta = 99).
5.2 Type-P-One-Way-Loss-Period-Total 6.2. Type-P-One-Way-Loss-Period-Total
This represents the total number of loss periods, and can be derived This represents the total number of loss periods, and can be
from the loss period metric Type-P-One-Way-Loss-Period-Stream as derived from the loss period metric Type-P-One-Way-Loss-Period-Stream
follows: as follows:
Type-P-One-Way-Loss-Period-Total = maximum value of the first entry Type-P-One-Way-Loss-Period-Total = maximum value of the first
of the set of pairs, <loss period, loss>, representing the loss metric entry of the set of pairs, <loss period, loss>, representing the loss
Type-P-One-Way-Loss-Period-Stream. metric Type-P-One-Way-Loss-Period-Stream.
Note that this statistic does not describe the duration of each loss Note that this statistic does not describe the duration of each
period itself. If this statistic is large, it does not mean that the loss period itself. If this statistic is large, it does not mean
losses are more spread out than they are otherwise; one or more that the losses are more spread out than they are otherwise; one
loss periods may include bursty losses. This statistic is generally or more loss periods may include bursty losses. This statistic is
useful in gathering first order of approximation of loss spread. generally useful in gathering first order of approximation of loss
spread.
5.3 Type-P-One-Way-Loss-Period-Lengths 6.3. Type-P-One-Way-Loss-Period-Lengths
This statistic is a sequence of pairs <loss period, length>, with the This statistic is a sequence of pairs <loss period,
"loss period" entry ranging from 1 - Type-P-One-Way-Loss-Period-Total. length>, with the "loss period" entry ranging from 1 -
Thus the total number of pairs in this statistic equals Type-P-One-Way-Loss-Period-Total. Thus the total number of
Type-P-One-Way-Loss-Period-Total. In each pair, the "length" is pairs in this statistic equals Type-P-One-Way-Loss-Period-Total. In
obtained by counting the number of pairs, <loss period, loss>, in the each pair, the "length" is obtained by counting the number of pairs,
metric Type-P-One-Way-Loss-Period-Stream which have first entry equal <loss period, loss>, in the metric Type-P-One-Way-Loss-Period-Stream
to "loss period." which have first entry equal to "loss period."
Since this statistic represents the number of packets lost in each Since this statistic represents the number of packets lost in each
loss period, it is an indicator of burstiness of each loss period. In loss period, it is an indicator of burstiness of each loss period.
conjunction with loss-period-total statistic, this statistic is generally In conjunction with loss-period-total statistic, this statistic is
useful in observing which loss periods are potentially more influential generally useful in observing which loss periods are potentially more
than others from a quality perspective. influential than others from a quality perspective.
5.4 Type-P-One-Way-Inter-Loss-Period-Lengths 6.4. Type-P-One-Way-Inter-Loss-Period-Lengths
This statistic measures distance between successive loss periods. It This statistic measures distance between successive loss
takes the form of a set of pairs periods. It takes the form of a set of pairs <loss period,
<loss period, inter-loss-period-length>, with the inter-loss-period-length>, with the "loss period" entry ranging from
"loss period" entry ranging from 1 - Type-P-One-Way-Loss-Period-Total, 1 - Type-P-One-Way-Loss-Period-Total, and "inter-loss-period-length"
and "inter-loss-period-length" is the loss distance between the last is the loss distance between the last packet considered lost in "loss
packet considered lost in "loss period" 'i-1', and the first packet period" 'i-1', and the first packet considered lost in "loss period"
considered lost in "loss period" 'i', where 'i' ranges from 2 to 'i', where 'i' ranges from 2 to Type-P-One-Way-Loss-Period-Total.
Type-P-One-Way-Loss-Period-Total. The "inter-loss-period-length" The "inter-loss-period-length" associated with the first "loss
associated with the first "loss period" is defined to be zero. period" is defined to be zero.
This statistic allows one to consider, for example, two loss periods each This statistic allows one to consider, for example, two loss
of length greater than one (implying loss burst), but separated by a periods each of length greater than one (implying loss burst), but
distance of 2 to belong to the same loss burst if such a consideration separated by a distance of 2 to belong to the same loss burst if such
is deemed useful. When the Inter-Loss-Period-Length between two bursty a consideration is deemed useful. When the Inter-Loss-Period-Length
loss periods is smaller, it could affect the loss concealing ability of between two bursty loss periods is smaller, it could affect the loss
multimedia codecs since there is relatively smaller history. When it is concealing ability of multimedia codecs since there is relatively
larger, an application may be able to rebuild its history which could smaller history. When it is larger, an application may be able to
dampen the effect of an impending loss (period). rebuild its history which could dampen the effect of an impending
loss (period).
5.5 Example 6.5. Examples
We continue with the same example as in Section 4.4.3. The three We continue with the same example as in Section 5.4.3. The three
statistics defined above will have the following values. statistics defined above will have the following values.
+ Let delta = 2. - Let delta = 2. In Type-P-One-Way-Loss-Distance-Stream
In Type-P-One-Way-Loss-Distance-Stream
{<0,0>,<0,1>,<0,0>,<0,0>,<3,1>,<0,0>,<2,1>,<0,0>,<2,1>,<1,1>}, there
are 3 loss distances that violate the delta of 2. Thus,
Type-P-One-Way-Loss-Noticeable-Rate = 3/5 {<0,0>,<0,1>,<0,0>,<0,0>,<3,1>,<0,0>,<2,1>,<0,0>,<2,1>,<1,1>},
((number of noticeable losses)/(number of total losses)) there are 3 loss distances that violate the delta of 2. Thus,
+ In Type-P-One-Way-Loss-Period-Stream Type-P-One-Way-Loss-Noticeable-Rate = 3/5 ((number of noticeable
{<0,0>,<1,1>,<0,0>,<0,0>,<2,1>,<0,0>,<3,1>,<0,0>,<4,1>,<4,1>}, the losses)/(number of total losses))
largest of the first entry in the sequence of <loss period,loss>
pairs is 4. Thus, - In Type-P-One-Way-Loss-Period-Stream
{<0,0>,<1,1>,<0,0>,<0,0>,<2,1>,<0,0>,<3,1>,<0,0>,<4,1>,<4,1>},
the largest of the first entry in the sequence of <loss
period,loss> pairs is 4. Thus,
Type-P-One-Way-Loss-Period-Total = 4 Type-P-One-Way-Loss-Period-Total = 4
+ In Type-P-One-Way-Loss-Period-Stream - In Type-P-One-Way-Loss-Period-Stream
{<0,0>,<1,1>,<0,0>,<0,0>,<2,1>,<0,0>,<3,1>,<0,0>,<4,1>,<4,1>}, the
lengths of individual loss periods are 1, 1, 1 and 2 respectively.
Thus,
Type-P-One-Way-Loss-Period-Lengths = {<1,1>,<2,1>,<3,1>,<4,2>} {<0,0>,<1,1>,<0,0>,<0,0>,<2,1>,<0,0>,<3,1>,<0,0>,<4,1>,<4,1>},
+ In Type-P-One-Way-Loss-Period-Stream the lengths of individual loss periods are 1, 1, 1 and 2
{<0,0>,<1,1>,<0,0>,<0,0>,<2,1>,<0,0>,<3,1>,<0,0>,<4,1>,<4,1>}, the respectively. Thus,
loss periods 1 and 2 are separated by 3 (5-2), loss periods 2 and 3
are separated by 2 (7-5), and 3 and 4 are separated by 2 (9-7).
Thus,
Type-P-One-Way-Inter-Loss-Period-Lengths = {<1,0>,<2,3>,<3,2>,<4,2>}
6. Security Considerations Type-P-One-Way-Loss-Period-Lengths =
Since this draft proposes sample metrics based on the base loss metric {<1,1>,<2,1>,<3,1>,<4,2>}
defined in [AKZ], it inherits the security considerations mentioned in
[AKZ].
7. Acknowledgements - In Type-P-One-Way-Loss-Period-Stream
Many thanks to Matt Zekauskas for the constructive feedback on the draft. {<0,0>,<1,1>,<0,0>,<0,0>,<2,1>,<0,0>,<3,1>,<0,0>,<4,1>,<4,1>},
Thanks to Guy Almes for encouraging the work, and Vern Paxson for the
comments during the IETF meetings. Thanks to Steve Glass for making the
presentation at the Oslo meeting.
8. References the loss periods 1 and 2 are separated by 3 (5-2), loss periods
2 and 3 are separated by 2 (7-5), and 3 and 4 are separated by 2
(9-7). Thus,
Type-P-One-Way-Inter-Loss-Period-Lengths =
[AKZ] G. Almes and S. Kalindindi and M. Zekauskas, "A One-way Packet {<1,0>,<2,3>,<3,2>,<4,2>}
Loss Metric for IPPM", RFC 2680, September 1999
[Bolot] J.-C. Bolot and A. vega Garcia, "The case for FEC-based 7. Security Considerations:
error control for Packet Audio in the Internet", ACM Multimedia
Since this draft proposes sample metrics based on the base loss
metric defined in [1], it inherits the security considerations
mentioned in [1].
Conducting Internet measurements raises both security and privacy
concerns. This document does not specify a particular implementation
of metrics, so it does not directly affect the security of the
Internet nor of applications which run on the Internet. However,
implementations of these metrics must be mindful of security and
privacy concerns.
The derived sample metrics in this document are based on the loss
metric defined in RFC-2680 [1], and thus they inherit the security
considerations of that document. The reader should consult [1] for a
more detailed treatment of security considerations.
Nevertheless, there are a few things to highlight. First,
the lambda specified in the Type-P-Loss-Distance-Stream and
Type-P-Loss-Period-Stream controls the rate at which test packets
are sent, and therefore if it is set inappropriately large could
perturb the network under test, cause congestion, or at worst be a
denial-of-service attack to the network under test.
Second, privacy of user data is not a concern, since the
underlying metric is intended to be implemented using test packets
that contain no user information. Even if packets contained user
information, the derived metrics do not release data sent by the
user. Third, the results could be perturbed by attempting to corrupt
or disrupt the underlying stream, for example adding extra packets
that look just like test packets.
In general, legitimate measurements must have their parameters
selected carefully in order to avoid interfering with normal traffic
in the network. Such measurements should also be authorized and
authenticated in some way so that attacks can be identified and
intercepted.
8. IANA Considerations
Since this document does not define a specific protocol, nor does
it define any well-known values, there are no IANA considerations for
this document.
9. Acknowledgements
Matt Zekauskas provided insightful feedback and the text for the
Security Considerations section. We sincerely thank him for his
painstaking review and for supporting this work along with Merike
Kaeo. Thanks to Guy Almes for encouraging the work, and Vern Paxson
for the comments during the IETF meetings. Thanks to Steve Glass for
making the presentation at the Oslo meeting.
References
[1] G. Almes and S. Kalindindi and M. Zekauskas, "A One-way Packet
Loss Metric for IPPM", RFC 2680, September 1999.
[2] J.-C. Bolot and A. vega Garcia, "The case for FEC-based error
control for Packet Audio in the Internet", ACM Multimedia
Systems, 1997. Systems, 1997.
[Borella] M. S. Borella, D. Swider, S. Uludag, and G. B. Brewster, [3] M. S. Borella, D. Swider, S. Uludag, and G. B. Brewster,
"Internet Packet Loss: Measurement and Implications for End-to-End "Internet Packet Loss: Measurement and Implications for
QoS," Proceedings, International Conference on Parallel Processing, End-to-End QoS," Proceedings, International Conference on
August 1998. Parallel Processing, August 1998.
[Handley] M. Handley, "An examination of MBONE performance", [4] S. Bradner, "Key words for use in RFCs to Indicate Requirement
Technical Report, USC/ISI, ISI/RR-97-450, July 1997 Levels," RFC 2119, Internet Engineering Task Force, March 1997.
[RK97] R. Koodli, "Scheduling Support for Multi-tier Quality of [5] M. Handley, "An examination of MBONE performance", Technical
Report, USC/ISI, ISI/RR-97-450, July 1997
[6] R. Koodli, "Scheduling Support for Multi-tier Quality of
Service in Continuous Media Applications", PhD dissertation, Service in Continuous Media Applications", PhD dissertation,
Electrical and Computer Engineering Department, University of Electrical and Computer Engineering Department, University of
Massachusetts, Amherst, MA 01003. Massachusetts, Amherst, MA 01003, September 1997.
[Padhye1] J. Padhye, V. Firoiu, J. Kurose and D. Towsley, "Modeling [7] J. Padhye, V. Firoiu, J. Kurose and D. Towsley, "Modeling TCP
TCP throughput: a simple model and its empirical validation", in throughput: a simple model and its empirical validation", in
Proceedings of SIGCOMM'98, 1998. Proceedings of SIGCOMM'98, 1998.
[Padhye2] J. Padhye, J. Kurose, D. Towsley and R. Koodli, "A [8] J. Padhye, J. Kurose, D. Towsley and R. Koodli, "A TCP-friendly
TCP-friendly rate adjustment protocol for continuous media flows rate adjustment protocol for continuous media flows over
over best-effort networks", short paper presentation in best-effort networks", short paper presentation in ACM
ACM SIGMETRICS'99. Available as Umass Computer Science tech report SIGMETRICS'99. Available as Umass Computer Science tech report
from ftp://gaia.cs.umass.edu/pub/Padhye98-tcp-friendly-TR.ps.gz from ftp://gaia.cs.umass.edu/pub/Padhye98-tcp-friendly-TR.ps.gz
[Paxson] V. Paxson, "End-to-end Internet packet dynamics", Computer [9] V. Paxson, "End-to-end Internet packet dynamics", IEEE/ACM
Communication review, Proceedings of ACM SIGCOMM'97 Conference, Transactions on Networking, 7(3), pages 277-292, June 1999.
Cannes, France, September 1997, 27(4), pages 139-152, October 1997
[frame-work] V. Paxson, G. Almes, J. Mahdavi, and M. Mathis, [10] V. Paxson, G. Almes, J. Mahdavi, and M. Mathis, "Framework for
"Framework for IP Performance Metrics", RFC 2330, May 1998. IP Performance Metrics", RFC 2330, May 1998.
[Sriram] K. Sriram and W. Whitt, "Characterizing superposition [11] K. Sriram and W. Whitt, "Characterizing superposition arrival
arrival processes in packet multiplexers for voice and data", IEEE processes in packet multiplexers for voice and data", IEEE
Journal on Selected Areas of Communication, September 1986, pages Journal on Selected Areas of Communication, pages 833-846,
833-846 September 1986,
[Yajnik] M. Yajnik, J. Kurose and D. Towsley, "Packet loss [12] M. Yajnik, J. Kurose and D. Towsley, "Packet loss correlation
correlation in the MBONE multicast network", Proceedings of IEEE in the MBONE multicast network", Proceedings of IEEE Global
Global Internet, London, UK, November 1996. Internet, London, UK, November 1996.
Author's Addresses Addresses
Rajeev Koodli Questions about this memo can be directed to the authors:
Nokia Research Center
313, Fairchild Drive
Mountain View, CA 94043
Phone: +1 650-625-2359
Email: rajeev.koodli@nokia.com
Rayadurgam Ravikanth Rajeev Koodli Rayadurgam Ravikanth
Axiowave Networks Inc. Communications Systems Lab Axiowave Networks Inc.
100 Nickerson Road Nokia Research Center 100 Nickerson Road
Marlborough, MA- 01752 313 Fairchild Drive Marlborough, MA- 01752
Email: rravikanth@axiowave.com Mountain View, California 94043 USA
USA Email: rravikanth@axiowave.com
Phone: +1-650 625-2359
EMail: rajeev.koodli@nokia.com
Fax: +1 650 625-2502
 End of changes. 

This html diff was produced by rfcdiff 1.23, available from http://www.levkowetz.com/ietf/tools/rfcdiff/