wdiff draft-ietf-ippm-delay-03.txt draft-ietf-ippm-delay-04.txt

Network Working Group G. Almes
Internet Draft S. Kalidindi
Expiration Date: December 1998 March 1999 M. Zekauskas
Advanced Network & Services
June
August 1998

A One-way Delay Metric for IPPM
<draft-ietf-ippm-delay-03.txt>
<draft-ietf-ippm-delay-04.txt>

1. Status of this Memo

This document is an Internet-Draft. Internet-Drafts are working
documents of the Internet Engineering Task Force (IETF), its areas,
and its working groups. Note that other groups may also distribute
working documents as Internet Drafts.

Internet-Drafts are draft documents valid for a maximum of six
months, and may be updated, replaced, or obsoleted by other documents
at any time. It is inappropriate to use Internet- Drafts as
reference material or to cite them other than as "work in progress."

To view the entire list of current Internet-Drafts, please check the
"1id-abstracts.txt" listing contained in the Internet-Drafts shadow
directories on ftp.is.co.za (Africa), nic.nordu.net (Northern
Europe), ftp.nis.garr.it (Southern Europe), munnari.oz.au (Pacific
Rim), ftp.ietf.org (US East Coast), or ftp.isi.edu (US West Coast).

This memo provides information for the Internet community. This memo
does not specify an Internet standard of any kind. Distribution of
this memo is unlimited.

2. Introduction

This memo defines a metric for one-way delay of packets across
Internet paths. It builds on notions introduced and discussed in the
IPPM Framework document, RFC 2223 2330 [1]; the reader is assumed to be
familiar with that document.

This memo is intended to be parallel in structure to a companion
document for Packet Loss ("A Packet Loss Metric for IPPM"
<draft-ietf-ippm-loss-02.txt>)
<draft-ietf-ippm-loss-04.txt>) [2].

The structure of the memo is as follows:

+ A 'singleton' analytic metric, called Type-P-One-way-Delay, will
be introduced to measure a single observation of one-way delay.

+ Using this singleton metric, a 'sample', called Type-P-One-way-
Delay-Poisson-Stream, will be introduced to measure a sequence of
singleton delays measured at times taken from a Poisson process.

+ Using this sample, several 'statistics' of the sample will be
defined and discussed.

This progression from singleton to sample to statistics, with clear
separation among them, is important.

Whenever a technical term from the IPPM Framework document is first
used in this memo, it will be tagged with a trailing asterisk. For
example, "term*" indicates that "term" is defined in the Framework.

2.1. Motivation:

One-way delay of a type-P Type-P* packet from a source host* to a
destination host is useful for several reasons:

+ Some applications do not perform well (or at all) if end-to-end
delay between hosts is large relative to some threshold value.

+ Erratic variation in delay makes it difficult (or impossible) to
support many real-time applications.

+ The larger the value of delay, the more difficult it is for
transport-layer protocols to sustain high bandwidths.

+ The minimum value of this metric provides an indication of the
delay due only to propagation and transmission delay.

+ The minimum value of this metric provides an indication of the
delay that will likely be experienced when the path* traversed is
lightly loaded.

+ Values of this metric above the minimum provide an indication of
the congestion present in the path.

It is outside the scope of this document to say precisely how delay
metrics would be applied to specific problems.

2.2. General Issues Regarding Time

Whenever a time (i.e., a moment in history) is mentioned here, it is
understood to be measured in seconds (and fractions) relative to UTC.

As described more fully in the Framework document, there are four
distinct, but related notions of clock uncertainty:

synchronization*

measures the extent to which two clocks agree on what time it
is. For example, the clock on one host might be 5.4 msec ahead
of the clock on a second host.

accuracy*

measures the extent to which a given clock agrees with UTC. For
example, the clock on a host might be 27.1 msec behind UTC.

resolution*

measures the precision of a given clock. For example, the clock
on an old Unix host might tick only once every 10 msec, and thus
have a resolution of only 10 msec.

skew*

measures the change of accuracy, or of synchronization, with
time. For example, the clock on a given host might gain 1.3
msec per hour and thus be 27.1 msec behind UTC at one time and
only 25.8 msec an hour later. In this case, we say that the
clock of the given host has a skew of 1.3 msec per hour relative
to UTC, and this threatens accuracy. We might also speak of the
skew of one clock relative to another clock, and this threatens
synchronization.

3. A Singleton Definition for One-way Delay

3.1. Metric Name:

Type-P-One-way-Delay

3.2. Metric Parameters:

+ Src, the IP address of a host

+ Dst, the IP address of a host

+ T, a time

3.3. Metric Units:

The value of a type-P-One-way-Delay Type-P-One-way-Delay is either a non-negative real
number
number, or an undefined (informally, infinite) number of seconds.

3.4. Definition:

For a non-negative real number dT, >>the *Type-P-One-way-Delay* from
Src to Dst at T is dT<< means that Src sent the first bit of a type-P Type-P
packet to Dst at wire-time* T and that Dst received the last bit of
that packet at wire-time T+dT.

>>The *Type-P-One-way-Delay* from Src to Dst at T is undefined
(informally, infinite)<< means that Src sent the first bit of a type- Type-
P packet to Dst at wire-time T and that Dst did not receive that
packet.

Suggestions for what to report along with metric values appear in
Section 3.8 after a discussion of the metric, methodologies for
measuring the metric, and error analysis.

3.5. Discussion:

Type-P-One-way-Delay is a relatively simple analytic metric, and one
that we believe will afford effective methods of measurement.

The following issues are likely to come up in practice:

+ Since delay values will often be as low as the 100 usec to 10 msec
range, it will be important for Src and Dst to synchronize very
closely. GPS systems afford one way to achieve synchronization to
within several 10s of usec. Ordinary application of NTP may allow
synchronization to within several msec, but this depends on the
stability and symmetry of delay properties among those NTP agents
used, and this delay is what we are trying to measure. A
combination of some GPS-based NTP servers and a conservatively
designed and deployed set of other NTP servers should yield good
results, but this is yet to be tested.

+ A given methodology will have to include a way to determine
whether a delay value is infinite or whether it is merely very
large (and the packet is yet to arrive at Dst). As noted by
Mahdavi and Paxson [4], simple upper bounds (such as the 255
seconds theoretical upper bound on the lifetimes of IP
packets [5]) could be used, but good engineering, including an
understanding of packet lifetimes, will be needed in practice.
{Comment: Note that, for many applications of these metrics, the
harm in treating a large delay as infinite might be zero or very
small. A TCP data packet, for example, that arrives only after
several multiples of the RTT may as well have been lost.}

+ The context in which If the metric packet is measured must be carefully
considered, and should always be reported duplicated along with metric
results.

As noted in the Framework document [1], path (or paths) so that
multiple non-corrupt copies arrive at the value of destination, then the metric
may depend on
packet is counted as received, and the type of IP packets used first copy to make arrive
determines the measurement,
or "type-P". The value of Type-P-One-way-Delay could change if packet's one-way delay.

+ If the packet is fragmented and if, for whatever reason,
reassembly does not occur, then the packet will be deemed lost.

3.6. Methodologies:

As with other Type-P-* metrics, the detailed methodology will depend
on the Type-P (e.g., protocol (UDP or TCP), number, UDP/TCP port number, size, or arrangement
precedence).

Generally, for
special treatment (e.g., IP precedence or RSVP) changes. The
exact Type-P used to make the measurements must be accurately
reported.

In addition, a given Type-P, the threshold (or methodology would proceed as
follows:

+ Arrange that Src and Dst are synchronized; that is, that they have
clocks that are very closely synchronized with each other and each
fairly close to distinguish) between
a large finite delay the actual time.

+ At the Src host, select Src and loss should be reported.

Finally, Dst IP addresses, and form a test
packet of Type-P with these addresses. Any 'padding' portion of
the path traversed by packet needed only to make the test packet a given size should
be reported, if
possible. In general it is impractical filled with randomized bits to know the precise path avoid a
given packet takes through situation in which the network. The precise path may
measured delay is lower than it would otherwise be
known for certain Type-P on short or stable paths. If Type-P
includes due to
compression techniques along the record route (or loose-source route) option path.

+ At the Dst host, arrange to receive the packet.

+ At the Src host, place a timestamp in the IP
header, prepared Type-P packet,
and send it towards Dst.

+ If the path is short enough, and all routers* on packet arrives within a reasonable period of time, take a
timestamp as soon as possible upon the path
support record (or loose-source) route, then receipt of the path will packet. By
subtracting the two timestamps, an estimate of one-way delay can
be
precisely recorded. This is impractical because computed. Error analysis of a given implementation of the route
method must be
short enough, many routers do not support (or are not configured
for) record route, and use take into account the closeness of this feature would often
artificially worsen synchronization
between Src and Dst. If the performance observed by removing delay between Src's timestamp and the
actual sending of the packet from common-case processing. However, partial information is still valuable context. For example, if a host can choose
between two links* (and hence two separate routes from src to
dst), known, then the initial link used is valuable context. {Comment:
For example, with Merit's NetNow setup, a Src on one NAP can reach
a Dst on another NAP estimate could be
adjusted by either subtracting this amount; uncertainty in this value
must be taken into account in error analysis. Similarly, if the
delay between the actual receipt of several different backbone
networks.}

The above list the packet and Dst's timestamp
is not exhaustive; any additional information that known, then the estimate could be useful adjusted by subtracting this
amount; uncertainty in interpreting applications of the metrics should this value must be reported. taken into account in
error analysis. See the next section, "Errors and Uncertainties",
for a more detailed discussion.

+ If the packet is duplicated along fails to arrive within a reasonable period of time,
the path (or paths) so one-way delay is taken to be undefined (informally, infinite).
Note that
multiple non-corrupt copies arrive at the destination, then threshold of 'reasonable' is a parameter of the
methodology.

Issues such as the packet is counted format, the means by which Dst knows when
to expect the test packet, and the means by which Src and Dst are
synchronized are outside the scope of this document. {Comment: We
plan to document elsewhere our own work in describing such more
detailed implementation techniques and we encourage others to as received,
well.}

3.7. Errors and Uncertainties:

The description of any specific measurement method should include an
accounting and analysis of various sources of error or uncertainty.
The Framework document provides general guidence on this point, but
we note here the following specifics related to delay metrics:

+ Errors or uncertainties due to uncertainties in the clocks of the
Src and Dst hosts.

+ Errors or uncertainties due to the difference between 'wire time'
and 'host time'.

In addition, the loss threshold may affect the results. Each of
these are discussed in more detail below, along with a section
("Calibration") on accounting for these errors and uncertainties.

3.7.1. Errors or uncertainties related to Clocks

The uncertainty in a measurement of one-way delay is related, in
part, to uncertainties in the clocks of the Src and Dst hosts. In
the following, we refer to the clock used to measure when the packet
was sent from Src as the source clock, we refer to the clock used to
measure when the packet was received by Dst as the dest clock, we
refer to the observed time when the packet was sent by the source
clock as Tsource, and the observed time when the packet was received
by the dest clock as Tdest. Alluding to the notions of
synchronization, accuracy, resolution, and skew mentioned in the
Introduction, we note the following:

+ Any error in the synchronization between the source clock and the
dest clock will contribute to error in the delay measurement. We
say that the source clock and the dest clock have a
synchronization error of Tsynch if the source clock is Tsynch
ahead of the dest clock. Thus, if we know the value of Tsynch
exactly, we could correct for clock synchronization by adding
Tsynch to the uncorrected value of Tdest-Tsource.

+ The accuracy of a clock is important only in identifying the time
at which a given delay was measured. Accuracy, per se, has no
importance to the accuracy of the measurement of delay. When
computing delays, we are interested only in the differences
between clock values, not the values themselves.

+ The resolution of a clock adds to uncertainty about any time
measured with it. Thus, if the source clock has a resolution of
10 msec, then this adds 10 msec of uncertainty to any time value
measured with it. We will denote the resolution of the source
clock and the dest clock as Rsource and Rdest, respectively.

+ The skew of a clock is not so much an additional issue as it is a
realization of the fact that Tsynch is itself a function of time.
Thus, if we attempt to measure or to bound Tsynch, this needs to
be done periodically. Over some periods of time, this function
can be approximated as a linear function plus some higher order
terms; in these cases, one option is to use knowledge of the
linear component to correct the clock. Using this correction, the
residual Tsynch is made smaller, but remains a source of
uncertainty that must be accounted for. We use the function
Esynch(t) to denote an upper bound on the uncertainty in
synchronization. Thus, |Tsynch(t)| <= Esynch(t).

Taking these items together, we note that naive computation Tdest-
Tsource will be off by Tsynch(t) +/- (|Rsource|+|Rdest|). Using the
notion of Esynch(t), we note that these clock-related problems
introduce a total uncertainty of Esynch(t)+|Rsource|+|Rdest|. This
estimate of total clock-related uncertainty should be included in the first copy
error/uncertainty analysis of any measurement implementation.

3.7.2. Errors or uncertainties related to arrive
determines the packet's Wire-time vs Host-time

As we have defined one-way delay.

+ If delay, we would like to measure the packet is fragmented and if, for whatever reason,
reassembly does not occur, then time
between when the test packet will be deemed lost.

3.6. Methodologies:

As with other Type-P-* metrics, the detailed methodology will depend
on the Type-P (e.g., protocol number, UDP/TCP port number, size,
precedence).

Generally, for a given Type-P, leaves the methodology would proceed as
follows:

+ Arrange that network interface of Src and Dst are synchronized; that is, that they have
clocks that are very closely synchronized with each other
when it (completely) arrives at the network interface of Dst, and each
fairly close we
refer to this as 'wire time'. If the actual time.

+ At the Src host, select timings are themselves
performed by software on Src and Dst IP addresses, and form a test
packet of Type-P with these addresses. Any 'padding' portion of
the packet needed Dst, however, then this software can
only directly measure the time between when Src grabs a timestamp
just prior to make sending the test packet and when Dst grabs a given size should
be filled with randomized bits to avoid a situation in which timestamp
just after having received the
measured delay is lower than it would otherwise be due test packet, and we refer to
compression techniques along this as
'host time'.

To the path.

+ At extent that the Dst host, arrange difference between wire time and host time is
accurately known, this knowledge can be used to receive the packet.

+ At the Src host, place a timestamp in the prepared Type-P packet, correct for host time
measurements and send it towards Dst.

+ If the packet arrives within a reasonable period of time, take a
timestamp as soon as possible upon corrected value more accurately estimates the receipt of
desired (wire time) metric.

To the packet. By
subtracting extent, however, that the two timestamps, an estimate of one-way delay can difference between wire time and
host time is uncertain, this uncertainty must be computed. Error accounted for in an
analysis of a given implementation of measurement method. We denote by Hsource an
upper bound on the
method must take into account uncertainty in the closeness of synchronization difference between Src wire time
and Dst. If host time on the delay between Src's timestamp Src host, and similarly define Hdest for the
actual sending of the packet is known, Dst
host. We then the note that these problems introduce a total uncertainty
of Hsource+Hdest. This estimate could be
adjusted by subtracting this amount; of total wire-vs-host uncertainty in this value
must
should be taken into account included in error analysis. Similarly, if the
delay between the actual receipt error/uncertainty analysis of any
measurement implementation.

3.7.3. Calibration

Generally, the packet and Dst's timestamp
is known, then measured values can be decomposed as follows:

measured value = true value + systematic error + random error

If the estimate could systematic error (the constant bias in measured values) can be adjusted by subtracting this
amount; uncertainty
determined, it can be compensated for in this the reported results.

reported value must = measured value - systematic error

therefore

reported value = true value + random error
The goal of calibration is to determine the systematic and random
error in as much detail as possible. At a minimum, a bound ("e")
should be taken into account found such that the reported value is in
error analysis.

+ If the packet fails range (true
value - e) to arrive within a reasonable period (true value + e) at least 95 percent of time, the one-way delay time. We
call "e" the error bar for the measurements. {Comment: 95 percent
was chosen because (1) some confidence level is taken desirable to be undefined (informally, infinite).
Note able
to remove outliers which will be found in measuring any physical
property; (2) a particular confidence level should be specified so
that the threshold results of 'reasonable' here is independent implementations can be compared; and
(3) even with a parameter of prototype user-level implementation, 95% was loose
enough to exclude outliers.}

From the
methodology.

Issues such as discussion in the packet format, previous two sections, the means error in
measurements could be bounded by which Dst knows when
to expect determining all the test packet, individual
uncertainties, and adding them together to form
Esynch(t) + |Rsource| + |Rdest| + Hsource + Hdest.
However, reasonable bounds on both the means clock-related uncertainty
captured by which Src and Dst are
synchronized are outside the scope of this document. {Comment: We
plan to document elsewhere our own work in describing such more
detailed implementation techniques and we encourage others to as
well.}

3.7. Errors first three terms and Uncertainties:

The description of any specific measurement method the host-related uncertainty
captured by the last two terms should include an
accounting be possible by careful design
techniques and analysis of various sources of error/uncertainty. The
Framework document provides general guidence on this point, but we
note here calibrating the following specifics related to delay metrics:

+ Errors/uncertainties due to uncertainties instruments using a known, isolated,
network in a lab.

For example, the clocks of the Src
and Dst hosts.

+ Errors/uncertainties due to clock-related uncertainties are greatly reduced
through the difference between 'wire time' and
'host time'.

Each use of these are discussed in more detail below.

3.7.1. Errors/uncertainties related to Clocks

The uncertainty in a measurement GPS time source. The sum of one-way delay Esynch(t) +
|Rsource| + |Rdest| is related, in
part, to uncertainties in small, and is also bounded for the clocks duration of
the Src and Dst hosts. In measurement because of the following, we refer to global time source.

The host-related uncertainties, Hsource + Hdest, could be bounded by
connecting two instruments back-to-back with a high-speed serial link
or isolated LAN (depending on the clock used to measure when intended network connection for
actual measurement), and performing repeated measurements. In this
case, unlike measuring live networks, repeated measurements are
measuring the packet
was sent from Src as same wire time. (When measuring live networks, the source clock, we refer to
wire time is what you are measuring, and varies with the clock used to
measure when load
encountered on the packet was received path traversed by Dst as the dest clock, we
refer to test packets.)

If the observed test packets are small, such a network connection has a
minimal wire time when the packet was sent that may be approximated by zero. The measured
delay therefore contains only systematic and random error in the source
clock as Tsource,
instrumentation. The "average value" of repeated measurements is the
systematic error, and the observed time when variation is the packet was received
by random error.

One way to compute the dest clock as Tdest. Alluding systematic error, and the random error to a
95% confidence is to repeat the notions experiment many times - at least
hundreds of
synchronization, accuracy, resolution, tests. The systematic error would then be the median,
and skew mentioned in likely the
Introduction, we note mode (the most frequently occuring value). {Comment:
It's likely the following:

+ Any systematic error in is represented by the synchronization between minimum value
(which is also the source clock median and the
dest clock mode); with unloaded instruments on
a single test path all the random error will contribute tend to be increased
time due to host processing. The only error in the resulting an a delay measurement. We
say that
less than the source clock systematic error would be due to clock-related
uncertainties (resolution and relative skew).} The random error could
then be found by removing the dest clock have a
synchronization systematic error of Tsynch if from the source clock is Tsynch
ahead of measured
values. The 95% confidence interval would be the dest clock. Thus, if we know range from the value 2nd
percentile to the 97th percentile of Tsynch
exactly, we these deviations from the true
value. The error bar "e" could correct for clock synchronization by adding
Tsynch then be taken to be the uncorrected largest
absolute value of Tdest-Tsource.

+ The accuracy these two numbers, plus the clock-related
uncertainty. If all of a clock is important only in identifying the time
at which a given delay was measured. Accuracy, per se, has no
importance to deviations are positive, then the accuracy 95%
confidence interval is simply the 95th percentile, and that value
should be used instead of the measurement larger of delay. This the 2nd and 97th percentiles.
{Comment: as described, this bound is
because, when computing delays, we relatively loose since the
uncertainties are interested only in added, and the
differences between clock values.

+ The resolution absolute value of the largest
deviation is used. As long as the resulting value is not a clock adds to uncertainty about any time
significant fraction of the measured with it. Thus, if values, it is a reasonable
bound. If the source clock has resulting value is a resolution significant fraction of
10 msec, the
measured values, then more exact methods will be needed to compute an
error bar.}

Note that random error is a function of measurement load. For
example, if many paths will be measured by one instrument, this adds 10 msec might
increase interrupts, process scheduling, and disk I/O (for example,
recording the measurements), all of uncertainty to any time value which may increase the random
error in measured singletons. Therefore, in addition to minimal load
measurements to find the systematic error, calibration measurements
should be performed with it. We the same measurement load that the
instruments will see in the field.

In addition to calibrating the instruments for finite one-way delay,
two checks should be made to ensure that packets reported as losses
were really lost. First, the threshold for loss should be verified.
In particular, ensure the "reasonable" threshold is reasonable: that
it is very unlikely a packet will denote the resolution of arrive after the source
clock threshold value,
and therefore the dest clock as Rsource and Rdest, respectively.

+ The skew number of a clock is not so much packets lost over an additional issue as it interval is a
realization of not
sensitive to the fact error bound on measurements. Second, consider the
probability that Tsynch is itself a function of time.
Thus, if we attempt packet arrives at the network interface, but is
lost due to measure congestion on that interface or to bound Tsynch, this needs to
be done periodically. Over some periods of time, this function
can be approximated as a linear function plus some higher order
terms; other resource
exhaustion (e.g. buffers) in these cases, one option is to use knowledge of the
linear component to correct instrument.

3.8. Reporting the clock. Using this correction, metric:

The calibration and context in which the
residual Tsynch metric is made smaller, but remains a source of
uncertainty that measured must be accounted for.
carefully considered, and should always be reported along with metric
results. We use the function
Esynch(t) to denote an upper bound on the uncertainty in
synchronization. Thus, |Tsynch(t)| <= Esynch(t).

Taking these now present four items together, we note that naive computation Tdest-
Tsource will be off by Tsynch(t) +/- (|Rsource|+|Rdest|). Using to consider: the
notion Type-P of Esynch(t), we note that these clock-related problems
introduce a total uncertainty test
packets, the threshold of Esynch(t)+|Rsource|+|Rdest|. infinite delay (if any), error calibration,
and the path traversed by the test packets. This
estimate of total clock-related uncertainty should list is not
exhaustive; any additional information that could be included useful in
interpreting applications of the
error/uncertainty analysis of any measurement implementation.

3.7.2. Errors/uncertainties related to Wire-time vs Host-time

As we've defined one-way delay, we'd like to measure the time between
when the test packet leaves metrics should also be reported.

3.8.1. Type-P

As noted in the network interface of Src and when it
(completely) arrives at Framework document [1], the network interface value of Dst, and we refer to
this as 'wire time'. If the timings are themselves performed by
software metric may
depend on Src and Dst, however, then this software can only
directly measure the time between when Src grabs a timestamp just
prior type of IP packets used to sending make the test packet and when Dst grabs a timestamp just
after having received measurement, or
"type-P". The value of Type-P-One-way-Delay could change if the test packet, and we refer
protocol (UDP or TCP), port number, size, or arrangement for special
treatment (e.g., IP precedence or RSVP) changes. The exact Type-P
used to this as 'host
time'.

To make the extent that measurements must be accurately reported.

3.8.2. Loss threshold

In addition, the difference threshold (or methodology to distinguish) between wire time a
large finite delay and host time is
accurately known, this knowledge loss should be reported.

3.8.3. Calibration results

+ If the systematic error can be used to correct for host time
measurements and determined, it should be removed
from the corrected measured values.

+ Report an error bar, e, such that the true value more accurately estimates is the
desired (wire time) metric.

To reported
value plus or minus e, with 95% confidence.

+ If possible, report the extent, however, probability that the difference between wire time and
host time a test packet with finite
delay is uncertain, this uncertainty must reported as lost due to resource exhaustion on the
measurement instrument.

3.8.4. Path

Finally, the path traversed by the packet should be accounted for in an
analysis of reported, if
possible. In general it is impractical to know the precise path a
given measurement method. We denote by Hsource an
upper bound packet takes through the network. The precise path may be
known for certain Type-P on short or stable paths. If Type-P
includes the uncertainty record route (or loose-source route) option in the difference between wire time IP
header, and host time on the Src host, path is short enough, and similarly define Hdest for all routers* on the Dst
host. We path
support record (or loose-source) route, then note that these problems introduce a total uncertainty
of Hsource+Hdest. the path will be
precisely recorded. This estimate of total wire-vs-host uncertainty
should is impractical because the route must be included in
short enough, many routers do not support (or are not configured for)
record route, and use of this feature would often artificially worsen
the error/uncertainty analysis performance observed by removing the packet from common-case
processing. However, partial information is still valuable context.
For example, if a host can choose between two links* (and hence two
separate routes from src to dst), then the initial link used is
valuable context. {Comment: For example, with Merit's NetNow setup,
a Src on one NAP can reach a Dst on another NAP by either of any
measurement implementation. several
different backbone networks.}

4. A Definition for Samples of One-way Delay

Given the singleton metric Type-P-One-way-Delay, we now define one
particular sample of such singletons. The idea of the sample is to
select a particular binding of the parameters Src, Dst, and Type-P,
then define a sample of values of parameter T. The means for
defining the values of T is to select a beginning time T0, a final
time Tf, and an average rate lambda, then define a pseudo-random
Poisson arrival process of rate lambda, whose values fall between T0
and Tf. The time interval between successive values of T will then
average 1/lambda.

4.1. Metric Name:

Type-P-One-way-Delay-Poisson-Stream

4.2. Metric Parameters:

+ Src, the IP address of a host

+ Dst, the IP address of a host

+ T0, a time

+ Tf, a time

+ lambda, a rate in reciprocal seconds

4.3. Metric Units:

A sequence of pairs; the elements of each pair are:

+ T, a time, and

+ dT, either a non-negative real number or an undefined number of
seconds.

The values of T in the sequence are monotonic increasing. Note that
T would be a valid parameter to Type-P-One-way-Delay, and that dT
would be a valid value of Type-P-One-way-Delay.

4.4. Definition:

Given T0, Tf, and lambda, we compute a pseudo-random Poisson process
beginning at or before T0, with average arrival rate lambda, and
ending at or after Tf. Those time values greater than or equal to T0
and less than or equal to Tf are then selected. At each of the times
in this process, we obtain the value of Type-P-One-way-Delay at this
time. The value of the sample is the sequence made up of the
resulting <time, delay> pairs. If there are no such pairs, the
sequence is of length zero and the sample is said to be empty.

4.5. Discussion:

Note first that, since a pseudo-random number sequence is employed,
the sequence of times, and hence the value of the sample, is not
fully specified. Pseudo-random number generators of good quality
will be needed to achieve the desired qualities.

The sample is defined in terms of a Poisson process both to avoid the
effects of self-synchronization and also capture a sample that is
statistically as unbiased as possible. {Comment: there is, of
course, no claim that real Internet traffic arrives according to a
Poisson arrival process.}

All the singleton Type-P-One-way-Delay metrics in the sequence will
have the same values of Src, Dst, and Type-P.

Note also that, given one sample that runs from T0 to Tf, and given
new time values T0' and Tf' such that T0 <= T0' <= Tf' <= Tf, the
subsequence of the given sample whose time values fall between T0'
and Tf' are also a valid Type-P-One-way-Delay-Poisson-Stream sample.

4.6. Methodologies:

The methodologies follow directly from:

+ the selection of specific times, using the specified Poisson
arrival process, and

+ the methodologies discussion already given for the singleton Type-
P-One-way-Delay metric.

Care must, of course, be given to correctly handle out-of-order
arrival of test packets; it is possible that the Src could send one
test packet at TS[i], then send a second one (later) at TS[i+1],
while the Dst could receive the second test packet at TR[i+1], and
then receive the first one (later) at TR[i].

4.7. Errors and Uncertainties:

In addition to sources of errors and uncertainties associated with
methods employed to measure the singleton values that make up the
sample, care must be given to analyze the accuracy of the Poisson
arrival process of the wire-time of the sending of the test packets.
Problems with this process could be caused by either of several things,
including problems with the pseudo-random number techniques used to
generate the Poisson arrival process, or with jitter in the value of
Hsource (mentioned above as uncertainty in the singleton delay
metric). The Framework document shows how to use an the Anderson-
Darling test to verify the accuracy of the Poisson process.

4.8. Reporting the metric:

You should report the calibration and context for this. the underlying
singletons along with the stream. (See "Reporting the metric" for
Type-P-One-way-Delay.)

5. Some Statistics Definitions for One-way Delay

Given the sample metric Type-P-One-way-Delay-Poisson-Stream, we now
offer several statistics of that sample. These statistics are
offered mostly to be illustrative of what could be done.

5.1. Type-P-One-way-Delay-Percentile

Given a Type-P-One-way-Delay-Poisson-Stream and a percent X between
0% and 100%, the Xth percentile of all the dT values in the Stream.
In computing this percentile, undefined values are treated as
infinitely large. Note that this means that the percentile could
thus be undefined (informally, infinite). In addition, the Type-P-
One-way-Delay-Percentile is undefined if the sample is empty.

Example: suppose we take a sample and the results are:
Stream1 = <
<T1, 100 msec>
<T2, 110 msec>
<T3, undefined>
<T4, 90 msec>
<T5, 500 msec>
>
Then the 50th percentile would be 110 msec, since 90 msec and 100
msec are smaller and 110 msec and 'undefined' are larger.

Note that if the probability that a finite packet is reported as lost
is significant, then a high percentile (90th or 95th) might be
reported as infinite instead of finite.

5.2. Type-P-One-way-Delay-Median

Given a Type-P-One-way-Delay-Poisson-Stream, the median of all the dT
values in the Stream. In computing the median, undefined values are
treated as infinitely large.

As noted in the Framework document, the median differs from the 50th
percentile only when the sample contains an even number of values, in
which case the mean of the two central values is used.

Example: suppose we take a sample and the results are:
Stream2 = <
<T1, 100 msec>
<T2, 110 msec>
<T3, undefined>
<T4, 90 msec>
>
Then the median would be 105 msec, the mean of 100 msec and 110 msec,
the two central values.

5.3. Type-P-One-way-Delay-Minumum

Given a Type-P-One-way-Delay-Poisson-Stream, the minimum of all the
dT values in the Stream. In computing this, undefined values are
treated as infinitely large. Note that this means that the minimum
could thus be undefined (informally, infinite) if all the dT values
are undefined. In addition, the Type-P-One-way-Delay-Minimum is
undefined if the sample is empty.

In the above example, the minimum would be 90 msec.

5.4. Type-P-One-way-Delay-Inverse-Percentile

Given a Type-P-One-way-Delay-Poisson-Stream and a non-negative time
duration threshold, the fraction of all the dT values in the Stream
less than or equal to the threshold. The result could be as low as
0% (if all the dT values exceed threshold) or as high as 100%.

In the above example, the Inverse-Percentile of 103 msec would be
50%.

6. Security Considerations

Conducting Internet measurements raises both security and privacy
concerns. This memo does not specify an implementation of the
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.

There are two types of security concerns: potential harm caused by
the measurements, and potential harm to the measurements. The
measurements could cause harm because they are active, and inject
packets into the network. The measurement parameters must be
carefully selected so that the measurements inject trivial amounts of
additional traffic into the networks they measure. If they inject
"too much" traffic, they can skew the results of the measurement, and
in extreme cases cause congestion and denial of service.

The measurements themselves could be harmed by routers giving
measurement traffic a different priority than "normal" traffic, or by
an attacker injecting artificial measurement traffic. If routers can
recognize measurement traffic and treat it separately, the
measurements will not reflect actual user traffic. If an attacker
injects artificial traffic that is accepted as legitimate, the loss
rate will be artificially lowered. Therefore, the measurement
methodologies should include appropriate techniques to reduce the
probability measurement traffic can be distinguished from "normal"
traffic. Authentication techniques, such as digital signatures, may
be used where appropriate to guard against injected traffic attacks.

The privacy concerns of network measurement are limited by the active
measurements described in this memo. Unlike passive measurements,
there can be no release of existing user data.

7. Acknowledgements

Special thanks are due to Vern Paxson of Lawrence Berkeley Labs for
his helpful comments on issues of clock uncertainty and statistics.
Thanks also to Will Leland, Sean Shapira Shapira, and to Roland Wittig for
several useful suggestions.

8. References

[1] V. Paxson, G. Almes, J. Mahdavi, and M. Mathis, "Framework for
IP Performance Metrics", RFC 2330, May 1998.

[2] G. Almes, S. Kalidindi, and M. Zekauskas, "A One-way Delay Packet Loss Metric
for IPPM", Internet-Draft <draft-ietf-ippm-delay-02.txt>,
June <draft-ietf-ippm-loss-04.txt>, August
1998.

[3] D. Mills, "Network Time Protocol (v3)", RFC 1305, April 1992.

[4] J. Mahdavi and V. Paxson, "Connectivity", Work in Progress,
November 1997. "IPPM Metrics for Measuring
Connectivity", Internet-Draft <draft-ietf-ippm-
connectivity-02.txt>, August 1998.

[5] J. Postel, "Internet Protocol", RFC 791, September 1981.

9. Authors' Addresses
Guy Almes
Advanced Network & Services, Inc.
200 Business Park Drive
Armonk, NY 10504
USA

Phone: +1 914 765 1120
EMail: almes@advanced.org

Sunil Kalidindi
Advanced Network & Services, Inc.
200 Business Park Drive
Armonk, NY 10504
USA

Phone: +1 914 765 1128
EMail: kalidindi@advanced.org

Matthew J. Zekauskas
Advanced Network & Services, Inc.
200 Buisiness Park Drive
Armonk, NY 10504
USA

Phone: +1 914 765 1112
EMail: matt@advanced.org

Expiration date: December, 1998 March, 1999