idnits 2.17.1 draft-ietf-ippm-loss-episode-metrics-03.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- No issues found here. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year == The document seems to lack the recommended RFC 2119 boilerplate, even if it appears to use RFC 2119 keywords -- however, there's a paragraph with a matching beginning. Boilerplate error? (The document does seem to have the reference to RFC 2119 which the ID-Checklist requires). == The document seems to contain a disclaimer for pre-RFC5378 work, but was first submitted on or after 10 November 2008. The disclaimer is usually necessary only for documents that revise or obsolete older RFCs, and that take significant amounts of text from those RFCs. If you can contact all authors of the source material and they are willing to grant the BCP78 rights to the IETF Trust, you can and should remove the disclaimer. Otherwise, the disclaimer is needed and you can ignore this comment. (See the Legal Provisions document at https://trustee.ietf.org/license-info for more information.) -- The document date (October 27, 2011) is 4558 days in the past. Is this intentional? Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) -- Looks like a reference, but probably isn't: '0' on line 672 -- Looks like a reference, but probably isn't: '1' on line 672 ** Obsolete normative reference: RFC 2680 (Obsoleted by RFC 7680) Summary: 1 error (**), 0 flaws (~~), 3 warnings (==), 3 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group N. Duffield 3 Internet-Draft AT&T Labs-Research 4 Intended status: Standards Track A. Morton 5 Expires: April 29, 2012 AT&T Labs 6 J. Sommers 7 Colgate University 8 October 27, 2011 10 Loss Episode Metrics for IPPM 11 draft-ietf-ippm-loss-episode-metrics-03 13 Abstract 15 The IETF has developed a one way packet loss metric that measures the 16 loss rate on a Poisson probe stream between two hosts. However, the 17 impact of packet loss on applications is in general sensitive not 18 just to the average loss rate, but also to the way in which packet 19 losses are distributed in loss episodes (i.e., maximal sets of 20 consecutively lost probe packets). This draft defines one-way packet 21 loss episode metrics, specifically the frequency and average duration 22 of loss episodes, and a probing methodology under which the loss 23 episode metrics are to be measured. 25 Requirements Language 27 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 28 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 29 document are to be interpreted as described in RFC 2119 [RFC2119] 31 Status of this Memo 33 This Internet-Draft is submitted in full conformance with the 34 provisions of BCP 78 and BCP 79. 36 Internet-Drafts are working documents of the Internet Engineering 37 Task Force (IETF). Note that other groups may also distribute 38 working documents as Internet-Drafts. The list of current Internet- 39 Drafts is at http://datatracker.ietf.org/drafts/current/. 41 Internet-Drafts are draft documents valid for a maximum of six months 42 and may be updated, replaced, or obsoleted by other documents at any 43 time. It is inappropriate to use Internet-Drafts as reference 44 material or to cite them other than as "work in progress." 46 This Internet-Draft will expire on April 29, 2012. 48 Copyright Notice 49 Copyright (c) 2011 IETF Trust and the persons identified as the 50 document authors. All rights reserved. 52 This document is subject to BCP 78 and the IETF Trust's Legal 53 Provisions Relating to IETF Documents 54 (http://trustee.ietf.org/license-info) in effect on the date of 55 publication of this document. Please review these documents 56 carefully, as they describe your rights and restrictions with respect 57 to this document. Code Components extracted from this document must 58 include Simplified BSD License text as described in Section 4.e of 59 the Trust Legal Provisions and are provided without warranty as 60 described in the Simplified BSD License. 62 This document may contain material from IETF Documents or IETF 63 Contributions published or made publicly available before November 64 10, 2008. The person(s) controlling the copyright in some of this 65 material may not have granted the IETF Trust the right to allow 66 modifications of such material outside the IETF Standards Process. 67 Without obtaining an adequate license from the person(s) controlling 68 the copyright in such materials, this document may not be modified 69 outside the IETF Standards Process, and derivative works of it may 70 not be created outside the IETF Standards Process, except to format 71 it for publication as an RFC or to translate it into languages other 72 than English. 74 Table of Contents 76 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 5 77 1.1. Background and Motivation . . . . . . . . . . . . . . . . 5 78 1.2. Loss Episode Metrics and Bi-Packet Probes . . . . . . . . 6 79 1.3. Outline and Contents . . . . . . . . . . . . . . . . . . . 7 80 2. Singleton Definition for Type-P-One-way Bi-Packet Loss . . . . 8 81 2.1. Metric Name . . . . . . . . . . . . . . . . . . . . . . . 8 82 2.2. Metric Parameters . . . . . . . . . . . . . . . . . . . . 8 83 2.3. Metric Units . . . . . . . . . . . . . . . . . . . . . . . 8 84 2.4. Metric Definition . . . . . . . . . . . . . . . . . . . . 8 85 2.5. Discussion . . . . . . . . . . . . . . . . . . . . . . . . 9 86 2.6. Methodologies . . . . . . . . . . . . . . . . . . . . . . 9 87 2.7. Errors and Uncertainties . . . . . . . . . . . . . . . . . 9 88 2.8. Reporting the Metric . . . . . . . . . . . . . . . . . . . 9 89 3. General Definition of samples for 90 Type-P-One-way-Bi-Packet-Loss . . . . . . . . . . . . . . . . 9 91 3.1. Metric Name . . . . . . . . . . . . . . . . . . . . . . . 10 92 3.2. Metric Parameters . . . . . . . . . . . . . . . . . . . . 10 93 3.3. Metric Units . . . . . . . . . . . . . . . . . . . . . . . 10 94 3.4. Metric Definition . . . . . . . . . . . . . . . . . . . . 10 95 3.5. Discussion . . . . . . . . . . . . . . . . . . . . . . . . 10 96 3.6. Methodologies . . . . . . . . . . . . . . . . . . . . . . 10 97 3.7. Errors and Uncertainties . . . . . . . . . . . . . . . . . 11 98 3.8. Reporting the Metric . . . . . . . . . . . . . . . . . . . 11 99 4. An active probing methodology for Bi-Packet Loss . . . . . . . 11 100 4.1. Metric Name . . . . . . . . . . . . . . . . . . . . . . . 11 101 4.2. Metric Parameters . . . . . . . . . . . . . . . . . . . . 11 102 4.3. Metric Units . . . . . . . . . . . . . . . . . . . . . . . 12 103 4.4. Metric Definition . . . . . . . . . . . . . . . . . . . . 12 104 4.5. Discussion . . . . . . . . . . . . . . . . . . . . . . . . 12 105 4.6. Methodologies . . . . . . . . . . . . . . . . . . . . . . 12 106 4.7. Errors and Uncertainties . . . . . . . . . . . . . . . . . 13 107 4.8. Reporting the Metric . . . . . . . . . . . . . . . . . . . 13 108 5. Loss Episode Proto-Metrics . . . . . . . . . . . . . . . . . . 13 109 5.1. Loss-Pair-Counts . . . . . . . . . . . . . . . . . . . . . 13 110 5.2. Bi-Packet-Loss-Ratio . . . . . . . . . . . . . . . . . . . 14 111 5.3. Bi-Packet-Loss-Episode-Duration-Number . . . . . . . . . . 14 112 5.4. Bi-Packet-Loss-Episode-Frequency-Number . . . . . . . . . 14 113 6. Loss Episode Metrics derived from Bi-Packet Loss Probing . . . 14 114 6.1. Geometric Stream: Loss Ratio . . . . . . . . . . . . . . . 15 115 6.1.1. Metric Name . . . . . . . . . . . . . . . . . . . . . 15 116 6.1.2. Metric Parameters . . . . . . . . . . . . . . . . . . 15 117 6.1.3. Metric Units . . . . . . . . . . . . . . . . . . . . . 16 118 6.1.4. Metric Definition . . . . . . . . . . . . . . . . . . 16 119 6.1.5. Discussion . . . . . . . . . . . . . . . . . . . . . . 16 120 6.1.6. Methodologies . . . . . . . . . . . . . . . . . . . . 16 121 6.1.7. Errors and Uncertainties . . . . . . . . . . . . . . . 16 122 6.1.8. Reporting the Metric . . . . . . . . . . . . . . . . . 16 123 6.2. Geometric Steam: Loss Episode Duration . . . . . . . . . . 16 124 6.2.1. Metric Name . . . . . . . . . . . . . . . . . . . . . 16 125 6.2.2. Metric Parameters . . . . . . . . . . . . . . . . . . 16 126 6.2.3. Metric Units . . . . . . . . . . . . . . . . . . . . . 17 127 6.2.4. Metric Definition . . . . . . . . . . . . . . . . . . 17 128 6.2.5. Discussion . . . . . . . . . . . . . . . . . . . . . . 17 129 6.2.6. Methodologies . . . . . . . . . . . . . . . . . . . . 17 130 6.2.7. Errors and Uncertainties . . . . . . . . . . . . . . . 17 131 6.2.8. Reporting the Metric . . . . . . . . . . . . . . . . . 18 132 6.3. Geometric Stream: Loss Episode Frequency . . . . . . . . . 18 133 6.3.1. Metric Name . . . . . . . . . . . . . . . . . . . . . 18 134 6.3.2. Metric Parameters . . . . . . . . . . . . . . . . . . 18 135 6.3.3. Metric Units . . . . . . . . . . . . . . . . . . . . . 18 136 6.3.4. Metric Definition . . . . . . . . . . . . . . . . . . 18 137 6.3.5. Discussion . . . . . . . . . . . . . . . . . . . . . . 19 138 6.3.6. Methodologies . . . . . . . . . . . . . . . . . . . . 19 139 6.3.7. Errors and Uncertainties . . . . . . . . . . . . . . . 19 140 6.3.8. Reporting the Metric . . . . . . . . . . . . . . . . . 19 141 7. Applicability of Loss Episode Metrics . . . . . . . . . . . . 19 142 7.1. Relation to Gilbert Model . . . . . . . . . . . . . . . . 19 143 8. IPR Considerations . . . . . . . . . . . . . . . . . . . . . . 20 144 9. Security Considerations . . . . . . . . . . . . . . . . . . . 20 145 10. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 21 146 11. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 21 147 12. References . . . . . . . . . . . . . . . . . . . . . . . . . . 21 148 12.1. Normative References . . . . . . . . . . . . . . . . . . . 21 149 12.2. Informative References . . . . . . . . . . . . . . . . . . 21 150 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 21 152 1. Introduction 154 1.1. Background and Motivation 156 Packet loss in the Internet is a complex phenomenon due to the bursty 157 nature of traffic and congestion processes, influenced by both end- 158 users and applications, and the operation of transport protocols such 159 as TCP. For these reasons, the simplest model of packet loss--the 160 single parameter Bernoulli (independent) loss model--does not 161 represent the complexity of packet loss over periods of time. 162 Correspondingly, a single loss metric--the average packet loss ratio 163 over some period of time--arising, e.g., from a stream of Poisson 164 probes as in [RFC2680] is not sufficient to determine the effect of 165 packet loss on traffic in general. 167 Moving beyond single parameter loss models, Markovian and Markov- 168 modulated loss models involving transitions between a good and bad 169 state, each with an associated loss rate, have been proposed by 170 Gilbert and more generally by Elliot. In principle, Markovian models 171 can be formulated over state spaces involving patterns of loss of any 172 desired number of packets. However further increase in the size of 173 the state space makes such models cumbersome both for parameter 174 estimation (accuracy decreases) and prediction in practice (due to 175 computational complexity and sensitivity to parameter inaccuracy). 176 In general, the relevance and importance of particular models can 177 change in time, e.g. in response to the advent of new applications 178 and services. For this reason we are drawn to empirical metrics that 179 do not depend on a particular model for their interpretation. 181 An empirical measure of packet loss complexity, the index of 182 dispersion of counts (IDC), comprise, for each t >0, the ratio v(t) \ 183 a(t) of the variance v(t) and average a(t) of the number of losses 184 over successive measurement windows of a duration t. However, a full 185 characterization of packet loss over time requires specification of 186 the IDC for each window size t>0. 188 In the standards arena, loss pattern sample metrics are defined in 189 [RFC3357]. Following the Gilbert-Elliot model, burst metrics 190 specific for VoIP that characterize complete episodes of lost, 191 transmitted and discarded packets are defined in [RFC3611] 193 All these considerations motivate formulating empirical metrics of 194 one-way packet loss that provide the simplest generalization of the 195 successful [RFC2680] that can capture deviations from independent 196 packet loss in a robust model-independent manner, and, to define 197 efficient measurement methodologies for these metrics. 199 1.2. Loss Episode Metrics and Bi-Packet Probes 201 The losses experienced by the packet stream can be viewed as 202 occurring in loss episodes, i.e., maximal set of consecutively lost 203 packets. This memo describes one-way loss episode metrics: their 204 frequency and average duration. Although the average loss ratio can 205 be expressed in terms of these quantities, they go further in 206 characterizing the statistics of the patterns of packet loss within 207 the stream of probes. This is useful information in understanding 208 the effect of packet losses on application performance, since 209 different applications can have different sensitivities to patterns 210 of loss, being sensitive not only to the long term average loss rate, 211 but how losses are distributed in time. As an example: MPEG video 212 traffic may be sensitive to loss involving the I-frame in a group of 213 pictures, but further losses within an episode of sufficiently short 214 duration have no further impact; the damage is already done. 216 The loss episode metrics presented here have the following useful 217 properties: 219 1. the metrics are empirical and do not depend on an underlying 220 model; e.g., the loss process is not assumed to be Markovian. On 221 the other hand, it turns out that the metrics of this memo can be 222 related to the special case of the Gilbert Model parameters; see 223 Section 7. 225 2. the metric units can be directly compared with applications or 226 user requirements or tolerance for network loss performance, in 227 the frequency and duration of loss episodes, as well as the usual 228 packet loss ratio, which can be recovered from the loss episode 229 metrics upon dividing the average loss episode duration by the 230 loss episode frequency. 232 3. the metrics provide the smallest possible increment in complexity 233 beyond, but in the spirit of, the IPPM average packet loss ratio 234 metrics [RFC2680] i.e., moving from a single metric (average 235 packet loss ratio) to a pair of metrics (loss episode frequency 236 and average loss episode duration). 238 The draft also describes a probing methodology under which loss 239 episode metrics are to be measured. The methodology comprises 240 sending probe packets in pairs, where packets within each probe pair 241 have a fixed separation, and the time between pairs takes the form of 242 a geometric distributed number multiplied by the same separation. 243 This can be regarded a generalization of Poisson probing where the 244 probes are pairs rather than single packets as in [RFC2680], and also 245 of geometric probing described in [RFC2330]. However, it should be 246 distinguished from back to back packet pairs whose change in 247 separation on traversing a link is used to probe bandwidth. In this 248 draft, the separation between the packets in a pair is the temporal 249 resolution at which different loss episodes are to be distinguished. 250 One key feature of this methodology is its efficiency: it estimates 251 the average length of loss episodes without directly measuring the 252 complete episodes themselves. Instead, this information is encoded 253 in the observed relative frequencies of the 4 possible outcomes 254 arising from the loss or successful transmission of each of the two 255 packets of the probe pairs. This is distinct from the approach of 256 [RFC3611] that reports on directly measured episodes. 258 The metrics defined in this memo are "derived metrics", according to 259 Section 6.1 of [RFC2330] the IPPM framework. They are based on the 260 singleton loss metric defined in Section 2 of [RFC2680] . 262 1.3. Outline and Contents 264 o Section 2 defines the fundamental singleton metric for the 265 possible outcomes of a probe pair: Type-P-One-way-Bi-Packet-Loss. 267 o Section 3 defines sample sets of this metric derived from a 268 general probe stream: Type-P-One-way-Bi-Packet-Loss-Stream. 270 o Section 4 defines the prime example of the Bi-Packet-Loss-Stream 271 metrics, specifically Type-P-One-way-Bi-Packet-Loss-Geometric- 272 Stream arising from the geometric stream of packet-pair probes 273 that was described informally in Section 1. 275 o Section 5 defines Loss episode proto-metrics that summarize the 276 outcomes from a stream metrics as an intermediate step to forming 277 the loss episode metrics; they need not be reported in general. 279 o Section 6 defines the final loss episode metrics that are the 280 focus of this memo, the new metrics 282 * Type-P-One-way-Bi-Packet-Loss-Geometric-Stream-Episode- 283 Duration, the average duration, in seconds, of a loss episode 285 * Type-P-One-way-Bi-Packet-Loss-Geometric-Stream-Episode- 286 Frequency, the average frequency, per second, at which loss 287 episodes start. 289 * Type-P-One-way-Bi-Packet-Loss-Geometric-Stream-Ratio, which is 290 the average packet loss ratio metric arising from the geometric 291 stream probing methodology 293 o Section 7 details applications and relations to existing loss 294 models. 296 2. Singleton Definition for Type-P-One-way Bi-Packet Loss 298 2.1. Metric Name 300 Type-P-One-way-Bi-Packet-Loss 302 2.2. Metric Parameters 304 o Src, the IP address of a source host 306 o Dst, the IP address of a destination host 308 o T1, a sending time of the first packet 310 o T2, a sending time of the second packet, with T2>T1 312 o F, a selection function defining unambiguously the two packets 313 from the stream selected for the metric. 315 o P, the specification of the packet type, over and above the source 316 and destination addresses 318 2.3. Metric Units 320 A Loss Pair is pair (l1, l2) where each of l1 and l2 is a binary 321 value 0 or 1, where 0 signifies successful transmission of a packet 322 and 1 signifies loss. 324 The metric unit of Type-P-One-way-Bi-Packet-Loss is a Loss Pair. 326 2.4. Metric Definition 328 1. "The Type-P-One-way-Bi-Packet-Loss with parameters (Src, Dst, T1, 329 T2, F, P) is (1,1)" means that Src sent the first bit of a Type-P 330 packet to Dst at wire-time T1 and the first bit of a Type-P 331 packet to Dst a wire-time T2>T1, and that neither packet was 332 received at Dst. 334 2. The Type-P-One-way-Bi-Packet-Loss with parameters (Src, Dst, T1, 335 T2, F, P) is (1,0)" means that Src sent the first bit of a Type-P 336 packet to Dst at wire-time T1 and the first bit of a Type-P 337 packet to Dst a wire-time T2>T1, and that the first packet was 338 not received at Dst, and the second packet was received at Dst 340 3. The Type-P-One-way-Bi-Packet-Loss with parameters (Src, Dst, T1, 341 T2, F, P) is (0,1)" means that Src sent the first bit of a Type-P 342 packet to Dst at wire-time T1 and the first bit of a Type-P 343 packet to Dst a wire-time T2>T1, and that the first packet was 344 received at Dst, and the second packet was not received at Dst 346 4. The Type-P-One-way-Bi-Packet-Loss with parameters (Src, Dst, T1, 347 T2, F, P) is (0,0)" means that Src sent the first bit of a Type-P 348 packet to Dst at wire-time T1 and the first bit of a Type-P 349 packet to Dst a wire-time T2>T1, and that both packet were 350 received at Dst. 352 2.5. Discussion 354 The purpose of the selection function is to specify exactly which 355 packets are to be used for measurement. The notion is taken from 356 Section 2.5 of [RFC3393], where examples are discussed. 358 2.6. Methodologies 360 The methodologies related to the Type-P-One-way-Packet-Loss metric in 361 Section 2.6 of [RFC2680] are similar for the Type-P-One-way-Bi- 362 Packet-Loss metric described above. In particular, the methodologies 363 described in RFC 2680 apply to both packets of the pair. 365 2.7. Errors and Uncertainties 367 Sources of error for the Type-P-One-way-Packet-Loss metric in Section 368 2.7 of [RFC2680] apply to each packet of the pair for the Type-P-One- 369 way-Bi-Packet-Loss metric. 371 2.8. Reporting the Metric 373 Refer to Section 2.8 of [RFC2680]. 375 3. General Definition of samples for Type-P-One-way-Bi-Packet-Loss 377 Given the singleton metric for Type-P-One-way-Bi-Packet-Loss, we now 378 define examples of samples of singletons. The basic idea is as 379 follows. We first specify a set of times T1 < T2 <...1 594 o 0 if N(0,1) + N(1,0) + N(1,1) = 0 (no probe packets lost) 596 o Undefined if N(0,1) + N(1,0) + N(0,0) = 0 (all probe packets lost) 598 Note N(0,1) + N(1,0) is zero if there are no transitions between loss 599 and no-loss outcomes. 601 5.4. Bi-Packet-Loss-Episode-Frequency-Number 603 The Bi-Packet-Loss-Episode-Frequency-Number associated with a set of 604 n loss pairs L1,,,,Ln is defined in terms of their Loss-Pair-Counts 605 as Bi-Packet-Loss-Ratio / Bi-Packet-Loss-Episode-Duration-Number, 606 when this can be defined, specifically, it is: 608 o (N(1,0)+N(1,1)) * (N(0,1)+N(1,0)) / (2*N(1,1)+N(0,1)+N(1,0) ) / n 609 if N(0,1)+N(0,1) > 0 611 o 0 if N(0,1)+N(1,0) +N(1,1) = 0 (no probe packets lost) 613 o 1 if N(0,1) +N(1,0) +N(0,0) = 0 (all probe packets lost) 615 6. Loss Episode Metrics derived from Bi-Packet Loss Probing 617 Metrics for the time frequency and time duration of loss episodes are 618 now defined as functions of set of n loss pairs L1,....,Ln. Although 619 a loss episode is defined as a maximal set of successive lost 620 packets, the loss episode metrics are not defined directly in terms 621 of the sequential patterns of packet loss exhibited by loss pairs. 622 This is because samples, including Type-P-One-way-Bi-Packet-Loss- 623 Geometric-Stream, generally do not report all lost packets in each 624 episode. Instead, the metrics are defined as functions of the Loss- 625 Pair-Counts of the sample, for reasons that are now described. 627 Consider an idealized Type-P-One-way-Bi-Packet-Loss-Geometric-Stream 628 sample in which the launch probability q =1. It is shown in [SBDR08] 629 that the average number of packets in a loss episode of this ideal 630 sample is exactly the Bi-Packet-Loss-Episode-Duration derived from 631 its set of loss pairs. Note this computation makes no reference to 632 the position of lost packet in the sequence of probes. 634 A general Type-P-One-way-Bi-Packet-Loss-Geometric-Stream sample with 635 launch probability q < 1, independently samples, with probability q, 636 each loss pair of an idealized sample. On average, the Loss-Pair- 637 Counts (if normalized by the total number of pairs) will be the same 638 as in the idealized sample. The loss episode metrics in the general 639 case are thus estimators of those for the idealized case; the 640 statistical properties of this estimation, including a derivation of 641 the estimation variance, is provided in [SBDR08]. 643 6.1. Geometric Stream: Loss Ratio 645 6.1.1. Metric Name 647 Type-P-One-way-Bi-Packet-Loss-Geometric-Stream-Ratio 649 6.1.2. Metric Parameters 651 o Src, the IP address of a source host 653 o Dst, the IP address of a destination host 655 o T0, the randomly selected starting time [RFC3432] for periodic 656 launch opportunities 658 o d, the time spacing between potential launch times, Ti and Ti+1 660 o n, a count of potential measurement instants 662 o q, a launch probability 664 o F, a selection function defining unambiguously the two packets 665 from the stream selected for the metric. 667 o P, the specification of the packet type, over and above the source 668 and destination address 670 6.1.3. Metric Units 672 A number in the interval [0,1] 674 6.1.4. Metric Definition 676 The result obtained by computing the Bi-Packet-Loss-Ratio over a 677 Type-P-One-way-Bi-Packet-Loss-Geometric-Stream sample with the metric 678 parameters. 680 6.1.5. Discussion 682 Type-P-One-way-Bi-Packet-Loss-Geometric-Stream-Ratio estimates the 683 fraction of packets lost from the geometric stream of Bi-Packet 684 probes. 686 6.1.6. Methodologies 688 Refer to Section 4.6 690 6.1.7. Errors and Uncertainties 692 Because Type-P-One-way-Bi-Packet-Loss-Geometric-Stream is sampled in 693 general (when the launch probability q <1) the metrics described in 694 this Section can be regarded as statistical estimators of the 695 corresponding idealized version corresponding to q = 1. Estimation 696 variance as it applies to Type-P-One-way-Bi-Packet-Loss-Geometric- 697 Stream-Loss-Ratio is described in [SBDR08]. 699 For other issues refer to Section 4.7 701 6.1.8. Reporting the Metric 703 Refer to Section 4.8 705 6.2. Geometric Steam: Loss Episode Duration 707 6.2.1. Metric Name 709 Type-P-One-way-Bi-Packet-Loss-Geometric-Stream-Episode-Duration 711 6.2.2. Metric Parameters 712 o Src, the IP address of a source host 714 o Dst, the IP address of a destination host 716 o T0, the randomly selected starting time [RFC3432] for periodic 717 launch opportunities 719 o d, the time spacing between potential launch times, Ti and Ti+1 721 o n, a count of potential measurement instants 723 o q, a launch probability 725 o F, a selection function defining unambiguously the two packets 726 from the stream selected for the metric. 728 o P, the specification of the packet type, over and above the source 729 and destination address 731 6.2.3. Metric Units 733 A non-negative number of seconds. 735 6.2.4. Metric Definition 737 The result obtained by computing the Bi-Packet-Loss-Episode-Duration- 738 Number over a Type-P-One-way-Bi-Packet-Loss-Geometric-Stream sample 739 with the metric parameters, then multiplying the result by the launch 740 spacing parameter d. 742 6.2.5. Discussion 744 Type-P-One-way-Bi-Packet-Loss-Geometric-Stream-Episode-Duration 745 estimates the average duration of a loss episode, measured in 746 seconds. The duration measured in packets is obtained by dividing 747 the metric value by the packet launch spacing parameter d. 749 6.2.6. Methodologies 751 Refer to Section 4.6 753 6.2.7. Errors and Uncertainties 755 Because Type-P-One-way-Bi-Packet-Loss-Geometric-Stream is sampled in 756 general (when the launch probability q <1) the metrics described in 757 this Section can be regarded as statistical estimators of the 758 corresponding idealized version corresponding to q = 1. Estimation 759 variance as it applies to Type-P-One-way-Bi-Packet-Loss-Geometric- 760 Stream-Episode-Duration is described in [SBDR08]. 762 For other issues refer to Section 4.7 764 6.2.8. Reporting the Metric 766 Refer to Section 4.8 768 6.3. Geometric Stream: Loss Episode Frequency 770 6.3.1. Metric Name 772 Type-P-One-way-Bi-Packet-Loss-Geometric-Stream-Episode-Frequency 774 6.3.2. Metric Parameters 776 o Src, the IP address of a source host 778 o Dst, the IP address of a destination host 780 o T0, the randomly selected starting time [RFC3432] for periodic 781 launch opportunities 783 o d, the time spacing between potential launch times, Ti and Ti+1 785 o n, a count of potential measurement instants 787 o q, a launch probability 789 o F, a selection function defining unambiguously the two packets 790 from the stream selected for the metric. 792 o P, the specification of the packet type, over and above the source 793 and destination address 795 6.3.3. Metric Units 797 A positive number. 799 6.3.4. Metric Definition 801 The result obtained by computing the Bi-Packet-Loss-Episode-Frequency 802 over a Type-P-One-way-Bi-Packet-Loss-Geometric-Stream sample with the 803 metric parameters, then dividing the result by the launch spacing 804 parameter d. 806 6.3.5. Discussion 808 Type-P-One-way-Bi-Packet-Loss-Geometric-Stream-Episode-Frequency 809 estimates the average frequency per unit time with which loss 810 episodes start (or finish). The frequency relative to the count of 811 potential probe launches is obtained by multiplying the metric value 812 by the packet launch spacing parameter d. 814 6.3.6. Methodologies 816 Refer to Section 4.6 818 6.3.7. Errors and Uncertainties 820 Because Type-P-One-way-Bi-Packet-Loss-Geometric-Stream is sampled in 821 general (when the launch probability q <1) the metrics described in 822 this Section can be regarded as statistical estimators of the 823 corresponding idealized version corresponding to q = 1. Estimation 824 variance as it applies to Type-P-One-way-Bi-Packet-Loss-Geometric- 825 Stream-Episode-Frequency is described in [SBDR08]. 827 For other issues refer to Section 4.7 829 6.3.8. Reporting the Metric 831 Refer to Section 4.8 833 7. Applicability of Loss Episode Metrics 835 7.1. Relation to Gilbert Model 837 The general Gilbert-Elliot model is a discrete time Markov chain over 838 two states, Good (g) and Bad (b), each with its own independent 839 packet loss rate. In the simplest case, the Good loss rate is 0 840 while the Bad loss rate is 1. Correspondingly, there are two 841 independent parameters, the Markov transition probabilities P(g|b) = 842 1- P(b|b) and P(b|g) = 1- P(g|g), where P(i|j) is the probability to 843 transition from state j and step n to state i at step n+1. With 844 these parameters, the fraction of steps spent in the bad state is 845 P(b|g)/(P(b|g) + P(g|b)) while the average duration of a sojourn in 846 the bad state is 1/P(g|b) steps. 848 Now identify the steps of the Markov chain with the possible sending 849 times of packets for a Type-P-One-way-Bi-Packet-Loss-Geometric-Stream 850 with launch spacing d. Suppose the loss episode metrics Type-P-One- 851 way-Bi-Packet-Loss-Geometric-Stream-Ratio and ype-P-One-way-Bi- 852 Packet-Loss-Geometric-Stream-Episode-Duration take the values r and m 853 respectively. Then from the discussion in Section 6.2.5 the 854 following can be equated: 856 r = P(b|g)/(P(b|g) + P(g|b)) and m/d = 1/P(g|b). 858 These relationships can be inverted in order to recover the Gilbert 859 model parameters: 861 P(g|b) = d/m and P(b|g)=d/m/(1/r - 1) 863 8. IPR Considerations 865 An IPR disclosure concerning some of the material covered in this 866 draft has been made to the IETF: see 867 https://datatracker.ietf.org/ipr/1354/ 869 9. Security Considerations 871 Conducting Internet measurements raises both security and privacy 872 concerns. This memo does not specify an implementation of the 873 metrics, so it does not directly affect the security of the Internet 874 nor of applications which run on the Internet. 875 However,implementations of these metrics must be mindful of security 876 and privacy concerns. 878 There are two types of security concerns: potential harm caused by 879 the measurements, and potential harm to the measurements. The 880 measurements could cause harm because they are active, and inject 881 packets into the network. The measurement parameters MUST be 882 carefully selected so that the measurements inject trivial amounts of 883 additional traffic into the networks they measure. If they inject 884 "too much" traffic, they can skew the results of the measurement, and 885 in extreme cases cause congestion and denial of service. The 886 measurements themselves could be harmed by routers giving measurement 887 traffic a different priority than "normal" traffic, or by an attacker 888 injecting artificial measurement traffic. If routers can recognize 889 measurement traffic and treat it separately, the measurements may not 890 reflect actual user traffic. If an attacker injects artificial 891 traffic that is accepted as legitimate, the loss rate will be 892 artificially lowered. Therefore, the measurement methodologies 893 SHOULD include appropriate techniques to reduce the probability that 894 measurement traffic can be distinguished from "normal" traffic. 895 Authentication techniques, such as digital signatures, may be used 896 where appropriate to guard against injected traffic attacks. The 897 privacy concerns of network measurement are limited by the active 898 measurements described in this memo: they involve no release of user 899 data. 901 10. IANA Considerations 903 11. Acknowledgements 905 12. References 907 12.1. Normative References 909 [RFC2680] Almes, G., Kalidindi, S., and M. Zekauskas, "A One-way 910 Packet Loss Metric for IPPM", RFC 2680, September 1999. 912 [RFC3393] Demichelis, C. and P. Chimento, "IP Packet Delay Variation 913 Metric for IP Performance Metrics (IPPM)", RFC 3393, 914 November 2002. 916 [RFC3611] Friedman, T., Caceres, R., and A. Clark, "RTP Control 917 Protocol Extended Reports (RTCP XR)", RFC 3611, 918 November 2003. 920 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 921 Requirement Levels", BCP 14, RFC 2119, March 1997. 923 [RFC3432] Raisanen, V., Grotefeld, G., and A. Morton, "Network 924 performance measurement with periodic streams", RFC 3432, 925 November 2002. 927 12.2. Informative References 929 [RFC2330] Paxson, V., Almes, G., Mahdavi, J., and M. Mathis, 930 "Framework for IP Performance Metrics", RFC 2330, 931 May 1998. 933 [RFC3357] Koodli, R. and R. Ravikanth, "One-way Loss Pattern Sample 934 Metrics", RFC 3357, August 2002. 936 [SBDR08] IEEE/ACM Transactions on Networking, 16(2): 307-320, "A 937 Geometric Approach to Improving Active Packet Loss 938 Measurement", 2008. 940 Authors' Addresses 942 Nick Duffield 943 AT&T Labs-Research 944 180 Park Avenue 945 Florham Park, NJ 07932 946 USA 948 Phone: +1 973 360 8726 949 Fax: +1 973 360 8871 950 Email: duffield@research.att.com 951 URI: http://www.research.att.com/people/Duffield_Nicholas_G 953 Al Morton 954 AT&T Labs 955 200 Laurel Avenue South 956 Middletown,, NJ 07748 957 USA 959 Phone: +1 732 420 1571 960 Fax: +1 732 368 1192 961 Email: acmorton@att.com 962 URI: http://home.comcast.net/~acmacm/ 964 Joel Sommers 965 Colgate University 966 304 McGregory Hall 967 Hamilton, NY 13346 968 USA 970 Phone: +1 315 228 7587 971 Fax: 972 Email: jsommers@colgate.edu 973 URI: http://cs.colgate.edu/faculty/jsommers