Traffic Management for the Available Bit Rate (ABR) Service in ...
Traffic Management for the Available Bit Rate (ABR) Service in ... Traffic Management for the Available Bit Rate (ABR) Service in ...
Option # VC Rate Input Rate Link Allocated Method Measurement Congestion Rate point E ect Updated at A From FRM1 per-VC prev loop only FRM1 only B At per-class Q per-class both loops FRM1 only C From FRM1 per-VC both loops FRM1 only D At per-class Q per-class both loops FRM1 and BRM2 E From FRM1 per-VC both loops BRM2 only F At per-class Q per-class both loops BRM2 only Table 9.2: Summary of viable VS/VD design alternatives 9.5 Performance Evaluation of VS/VD Design Options 9.5.1 Metrics We use four metrics to evaluate the performance of these alternatives: Response Time: is the time taken to reach near optimal behavior on startup. Convergence Time: is the time for rate oscillations to decrease (time to reach the steady state). Throughput: Total data transferred per unit time. Maximum Queue: The maximum queue before convergence. The di erence between response time and convergence time is illustrated in Fig- ure 9.15. The following sections present simulation results with respect to the above metrics. Note that we have used greedy (in nite) tra c sources in our simulations. Wehave studied the algorithmic enhancements in non-VS/VD switches for non-greedy 357
sources in chapter 6. We expect the best implementation option (see below) to work well and produce consistent results when such (bursty) tra c is used. Response Time Figure 9.15: Response time vs Convergence time Without VS/VD all response times are close to the round-trip delay. With VS/VD, the response times are close to the feedback delay from the bottleneck. Since VS/VD reduces the response time during the rst round trip, it is good for long delay paths. The quick response time (10 ms in the parking lot con guration which has a 30 ms round trip time) is shown in Figure 9.13. Response time is also important for bursty tra c like TCP le transfer over ATM which \starts up" at the beginning of every active period (when the TCP window increases) after the corresponding idle period (see chapter 7). Throughput The number of cells received at the destination is a measure of the throughput achieved. These values are listed in Table 9.3. The top row is a list of the con guration codes (these codes are explained in Table 9.2. The nal column lists the throughput values for the case when a non-VS/VD switch is used. The 2 source+VBR and the parking lot con gurations have been introduced in earlier section. 358
- Page 333 and 334: hand, the frequency of the VBR is h
- Page 335 and 336: like ERICA+ which uses the queueing
- Page 337 and 338: minimum fairshare is low. This may
- Page 339 and 340: 8.17 E ect of Long-Range Dependent
- Page 341 and 342: terminology) are called \Presentati
- Page 343 and 344: The key point is that the MPEG-2 ra
- Page 345 and 346: the MPEG-2 Transport Stream, and st
- Page 347 and 348: 8.17.4 Observations on the Long-Ran
- Page 349 and 350: For the video sources, we choose me
- Page 351 and 352: Video Sources ABR Metrics # Mean St
- Page 353 and 354: However, with modi cations to ERICA
- Page 355 and 356: Video Sources ABR Metrics # Avg Src
- Page 357 and 358: Video Sources ABR Metrics # Avg Src
- Page 359 and 360: On the other hand, if the applicati
- Page 361 and 362: of managing bu ers, queueing, sched
- Page 363 and 364: hence control the total load on the
- Page 365 and 366: call such a switch a \VS/VD switch"
- Page 367 and 368: Figure 9.2: Per-class queues in a n
- Page 369 and 370: which arises is where the rate calc
- Page 371 and 372: 9.2 The ERICA Switch Scheme: Renota
- Page 373 and 374: The unknowns in the above equations
- Page 375 and 376: Figure 9.9: Two methods to measure
- Page 377 and 378: 9.4 VS/VD Switch Design Options 9.4
- Page 379 and 380: # VC Rate VC Input Rate Input Rate
- Page 381 and 382: 8 uses source rate measurement, we
- Page 383: The allocated rate update and the e
- Page 387 and 388: con guration mentioned in the table
- Page 389 and 390: can be very di erent for di erent V
- Page 391 and 392: CHAPTER 10 IMPLEMENTATION ISSUES At
- Page 393 and 394: With an enhanced UBR service, appli
- Page 395 and 396: 2. Some switch schemes have a proce
- Page 397 and 398: 4. Large legacy switches have a pro
- Page 399 and 400: section 9. Further, WAN switches wo
- Page 401 and 402: CHAPTER 11 SUMMARY AND FUTURE WORK
- Page 403 and 404: good transient performance. Since r
- Page 405 and 406: variant background tra c conditions
- Page 407 and 408: APPENDIX A SOURCE, DESTINATION AND
- Page 409 and 410: 7. After following behaviors #5 and
- Page 411 and 412: set the QL and SN elds to zero, pre
- Page 413 and 414: d) VS/VD Control: The switch may se
- Page 415 and 416: 5. Setting of other parameters at V
- Page 417 and 418: 4. The averaging interval timer exp
- Page 419 and 420: 1. Initialization: Target Cell Rate
- Page 421 and 422: THEN IF (OCR In Cell Fair Share Rat
- Page 423 and 424: APPENDIX C ERICA SWITCH ALGORITHM:
- Page 425 and 426: Number Active VCs In Last Interval
- Page 427 and 428: IF (NOT(Averaging VCs Option)) THEN
- Page 429 and 430: IF (Load Factor = In nity) THEN Loa
- Page 431 and 432: ; (Contribution[VC] = Decay Factor)
- Page 433 and 434: Name Explanation Flow Chart (FC) or
sources <strong>in</strong> chapter 6. We expect <strong>the</strong> best implementation option (see below) to work<br />
well and produce consistent results when such (bursty) tra c is used.<br />
Response Time<br />
Figure 9.15: Response time vs Convergence time<br />
Without VS/VD all response times are close to <strong>the</strong> round-trip delay. With<br />
VS/VD, <strong>the</strong> response times are close to <strong>the</strong> feedback delay from <strong>the</strong> bottleneck. S<strong>in</strong>ce<br />
VS/VD reduces <strong>the</strong> response time dur<strong>in</strong>g <strong>the</strong> rst round trip, it is good <strong>for</strong> long delay<br />
paths. The quick response time (10 ms <strong>in</strong> <strong>the</strong> park<strong>in</strong>g lot con guration which has a<br />
30 ms round trip time) is shown <strong>in</strong> Figure 9.13.<br />
Response time is also important <strong>for</strong> bursty tra c like TCP le transfer over ATM<br />
which \starts up" at <strong>the</strong> beg<strong>in</strong>n<strong>in</strong>g of every active period (when <strong>the</strong> TCP w<strong>in</strong>dow<br />
<strong>in</strong>creases) after <strong>the</strong> correspond<strong>in</strong>g idle period (see chapter 7).<br />
Throughput<br />
The number of cells received at <strong>the</strong> dest<strong>in</strong>ation is a measure of <strong>the</strong> throughput<br />
achieved. These values are listed <strong>in</strong> Table 9.3. The top row is a list of <strong>the</strong> con guration<br />
codes (<strong>the</strong>se codes are expla<strong>in</strong>ed <strong>in</strong> Table 9.2. The nal column lists <strong>the</strong> throughput<br />
values <strong>for</strong> <strong>the</strong> case when a non-VS/VD switch is used. The 2 source+VBR and <strong>the</strong><br />
park<strong>in</strong>g lot con gurations have been <strong>in</strong>troduced <strong>in</strong> earlier section.<br />
358