Comtech advanced backhaul_solutions-rio_nov_2012
0 likes•161 views
This document provides best practices for designing advanced backhaul solutions over satellite links. It discusses the importance of understanding traffic patterns and characteristics, preparing proper link budgets, and using tools like compression, adaptive coding and modulation, and carrier-in-carrier technology to optimize bandwidth usage in a statistical manner. It also covers considerations for different traffic types like synchronous data, packet-based links, and voice traffic with regards to latency, jitter, and traffic modeling.
Comtech advanced backhaul_solutions-rio_nov_2012
- 1. Advanced Backhaul Solutions: November 2012 © Copyright 2012 Comtech EF Data Corporation
- 2. Best Practices Solution Design • Before you even run your first link budget – Understanding your data and the tools that can be used with your type of data – Why the characteristics of a link are important to understand • Preparing a proper link budget – Amplifier size, linearity, and backoff – Satellite operation, attenuation, IBO vs. OBO – Inclined orbit ramifications • Understanding traffic patterns – Difference between lossless and non-lossless compression – Understanding the key benefits between Static and Statistical tools – What are the ramifications of these tools on my infrastructure / IT • Lessons learned 2
- 3. Characteristics of the Traffic • Circuit based services (Synchronous Data) – T1, E1,T3, E3, STS-1, STM-1 and OC-3 Consider underlying protocol requirements Typically symmetrical trunking type applications Clock and Clock Reference Issues, GPS clocking Frame and Superframe bound Fixed DR / Fixed Capacity Latency and Jitter Issues • Packet based links – 100/1000 Ethernet Links, Frame based Many types of Ethernet / IP traffic – Video over IP – Voice / Pseudowire over IP – Data – Mix 3
- 4. Voice Traffic • Latency and Jitter considerations – Quality impacted by high latency and jitter – Latency impacted by sizing of Jitter queues • Typically small transport packets cause high overhead • Traffic dimensioned using Erlang models, statistical estimates of Busy Hour traffic • No retransmission of lost transport packets • No queuing if congestion occurs, needs congestion managements techniques 4
- 5. Bandwidth on Demand for GSM and Erlang • Customer with 48 sites deployed across 3 highway spans provided detailed Erlang information per site. • Memotec analysis showed that network design based on pt-to- mpt topology was optimal (forward 2.8Mbps, returns 6.8Mbps). 5 0,00 2,00 4,00 6,00 8,00 10,00 12,00 14,00 16,00 18,00 20,00 00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:00 09:00 10:00 11:00 12:00 13:00 14:00 15:00 16:00 17:00 18:00 19:00 20:00 21:00 22:00 23:00 BTS 1 AVE BTS 10 AVE BTS 11 AVE BTS 12 AVE BTS 13 AVE BTS 15 AVE BTS 16 AVE BTS 18 AVE BTS 19 AVE Note: despite TDMA vendor claims, all sites reach peak at same time defeating argument of bandwidth sharing across IP/TDMA VSAT networks. Only SCPCs can offer low latency and jitter while providing optimal modulation and coding (lowest cost bit/hz). Note: Peak utilization for 16hrs per day Comtech EF Data Confidential
- 6. Characteristics of Link • Is there existing infrastructure – Limitations on Satellite Dish size / location – Limitations on existing Amplification Systems • Is the link Symmetric or Asymmetric – In terms of user traffic – In terms of ground resources • What kind of satellite is being considered – Cross Strapped vs. bent pipe – Frequency Band – Beam (Global, Hemi, Spot) – Inclined Orbit – Fixed Services Satellite (FSS), High speed Trunking Satellite (HTS) 6
- 7. Tools to reduce BW usage • Difference between Static and Statistical savings – Static advantage is a calculable and definitive advantage for the life of the circuit Improvements in FEC Carrier in Carrier Technology – Statistical advantages rely on probability, advantage is averaged over time but is rarely exact at any instant AUPC and CnC-APC Compression and Optimization Adaptive Coding and Modulation (ACM) 7
- 8. Static Advantages FEC and Coding Gain, Carrier in Carrier Technology 8
- 9. FEC and Coding Gain Design Considerations 9
- 10. Improvements In FEC (Forward Error Correction) • Shannon-Hartley Bound: There is a maximum bound on the amount of error free data that can be transmitted though a noisy transmission medium 10 Coding Type Vs. Shannon Bound Sequential or Viterbi + 4-8 dB Turbo Product Codes + 2-3 dB Versa FEC + 1-2 dB DVB-S2 LDPC+BCH + 0.7 – 1.5 dB
- 11. Advances in Coding Gain on: Fixed BW Maximize Throughput 11 Viterbi + RS TPC Coding Versa FEC DVB-S2 Fixed C/N 6.0 dB 6.0 dB 6.0 dB 6.0 dB Best Modulation / Coding for Es/No QPSK ½ QPSK ¾ QPSK .803 QPSK 5/6 (.827) Spectral Efficiency 0.92 bits / Hz 1.5 bits / Hz 1.61 bits / Hz 1.65 bits / Hz BW = SR 5.0 MHz 5.0 MHz 5.0 MHz 5.0 MHz User Data Rate 4.6 Mbps 7.5 Mbps 8.1 Mbps 8.3 Mbps
- 12. Carrier in Carrier Design Considerations 12
- 13. DoubleTalk® Carrier-in-Carrier® • Based on patented “Adaptive Cancellation”, Carrier- in-Carrier (CnC) allows carriers in a Duplex satellite link to occupy the same transponder space 13 Without DoubleTalk Carrier-in-Carrier With DoubleTalk Carrier-in-Carrier Carrier-in-Carrier is a Registered Trademark of Comtech EF Data DoubleTalk is a Registered Trademark of Applied Signal Technology, Inc.
- 14. Multi Dimensional Optimization • Combined with proper Modulation and FEC, Carrier-in- Carrier allows for multi dimensional optimization – Reducing OPEX Occupied Bandwidth & Transponder Power – Reducing CAPEX BUC/HPA Size and/or Antenna Size – Increasing throughput – Increasing link availability – Or a combination to meet different objectives 14
- 15. Extreme Spectral Efficiency • Carrier-in-Carrier can be used to achieve extreme spectral utilization that would not otherwise be possible • Best DVB-S2 Spec Efficiency = 4.45 b/Hz – Asia: 16APSK ¾ FEC --- 90Mbps DPLX in 15.5MHz Spectral Eff = 5.78 b/Hz = 64QAM – US: 32APSK ¾ FEC --- 217Mbps DPLX in 30MHz Spectral Eff = 7.24 b/Hz = 256QAM – THESE ARE CABLE MODEM FIGURES 15
- 16. Statistical Advantages AUPC, Compression, Optimization, ACM
- 17. AUPC and CnC-APC Design Considerations 17
- 18. AUPC vs CnC-APC • AUPC (Automatic Uplink Power Control) – Modem technology Historically simplistic method Change in uplink power due to Es/No fade at demod Does not differentiate between uplink or downlink fade • CnC-APC – Modem and CnC combined technology Non simplistic method Can discern between uplink and downlink fade Improves effective link margin Improves availability in CnC links 18
- 19. AUPC • Compensates for ANY fade condition by increasing TX power 19
- 20. CnC-APC • Modems on a CnC link to automatically compensate for rain loss while maintaining a fixed PEB • CnC modems share link margin between each other 20
- 22. Compression and Optimization Advantages Technology Business Benefits Quantifiable Benefits RAN Optimization/Multiplexing • Relieves congestion • Makes room for 3G • 50% (average) backhaul bandwidth reduction IP Optimization & Acceleration/Caching • Enhances user experience • 50% (average) bandwidth, and Latency mitigation IP Header & Payload Compression • Reduces OPEX • 30% (average) payload , 60% (average) header bandwidth reduction Ultra-Low Overhead Protocols • Reduces OPEX • 60% (average) overhead reduction Superior Modulation, ACM • Increases service availability • Enhances user experience • Reduces CAPEX • 50% (average) throughput gain of • Use of a smaller BUC/HPA and/or antenna Advanced Forward Error Correction and Protocols • Increases service availability • Enhances user experience • Reduces CAPEX • 30% (average) more throughput for same transponder • Use of a smaller BUC/HPA and/or antenna Reduce Costs, Enhance User Experience 22
- 23. Compression • Can my data be compressed? – Synchronous Data – Packet Based Data • Does my transport interface allow for compression • Lossy Compression vs. Lossless Compression – Lossy compression can be used on voice, video and multimedia – Lossless compression is the only compression suggested for data • How much can I expect my traffic to compress? – Good Question 23
- 24. Lossy Compression 24 Uncompressed Lossy Compression
- 25. Lossless Compression (Data) 25 • Pattern recognition: 20 bytes of data is represented by 12 bytes 40% Savings
- 26. Adaptive Coding and Modulation Design Considerations
- 27. Adaptive Coding & Modulation (ACM) • Adaptive Coding & Modulation (ACM) converts the link margin into increased capacity – average throughput gain of 100% (or more) is possible, compared to traditional CCM – Most of the year, the link operates at significantly increased throughput – For the worst few hours of the year, the link may be available with lower throughput 27
- 28. ACM Concept is Simple • (ACM) allows for automatic change in Modulation and Coding in response to current link conditions • Symbol Rate and Transmit Power are fixed and the data rate changes as the modulation and code rate are changed – ACM carriers use fixed bandwidth and power on the transponder – The receiving modem in a link, provides signal quality updates to the transmit modem – The transmit modem changes the Modulation and/or Code Rate to the most spectrally efficient Mod/Cod for the current conditions 28
- 29. Adaptive Coding and Modulation (ACM) • What type of advantage is ACM? – ACM is a Statistical, non-static throughput advantage – The advantage due to ACM will dynamically change – Throughput can not be guaranteed but is predictable – Will work on cross-strapped transponders – All margin (antenna pointing, incl orbit, rain fade, link budget margin, etc.) can be turned into higher capacity throughput • What are the restrictions? – Must be a closed loop system (feedback to uplink modulator) – Must have sufficient system link margin to be of value – Must be a packet based or scaleable link (not synchronous)
- 30. Overall Link Margin • ACM’s value is in the conversion of Link Margin • Greater the Link Margin the greater the value of ACM • Most Link Margin boils down to two main factors: – Link availability (99.8% vs. 99.6%) – Effects of Rain Fade (Ku-Band vs. C-Band) 30
- 31. Ku-Band Link Margin (Germany – Nigeria) 0 5 10 15 20 25 30 9999,199,299,399,499,599,699,799,899,9100 Margin(dB) Annual Availibility Figure 1: Ku-Band Link Margin (dB) Margin (dB) 2.5 dB Margin equates to +/- 0.2% Availability 31
- 32. C-Band Link Margin (Italy – China) 0 0,5 1 1,5 2 2,5 3 3,5 4 4,5 9999,199,299,399,499,599,699,799,899,9100 Margin(dB) Annual Availibility Figure 2: C-Band Link Margin (dB) Margin (dB) 0.35 dB Margin equates to +/- 0.2% Availability 32
- 34. How Does C/N Convert to User DR 0,0000 0,5000 1,0000 1,5000 2,0000 2,5000 3,0000 3,5000 4,0000 4,5000 5,0000 -5 0 5 10 15 20 SpectralEfficiency Es/No (Ideal) DVB-S2 Es/No Performance at Quasi Error Free PER=10-7 64K Blocks w/ Pilots S2-QPSK S2-8PSK S2-16APSK S2-32APSK 34 5.5 dB Margin 1 dB of converted margin = 10-15% DR Increase 5.5dB of converted margin = 55% -> 83% DR Increase
- 35. ACM in Operation 35 0 10 20 30 40 50 60 70 Time 3:10 3:30 3:50 4:10 4:30 4:50 5:10 5:30 5:50 6:10 6:30 6:50 7:10 7:30 7:50 8:10 8:30 8:50 9:10 9:30 9:50 10:… 10:… Mbps ACM Operation and User Throughput User DR Avail DR Rain fade: 2.2 dB 16-APSK 5/6 @ 64 Mbps To 16-APSK 2/3 @ 51 Mbps Deep Rain fade: 6.9 dB 16-APSK 5/6 @ 64 Mbps To QPSK 4/5 @ 31 Mbps
- 36. Inclined Orbit Issues • Inclined Orbit satellites create unwanted effects on the ground. Creates downlink beam movement that appears to the ground station as a sinusoidal EIRP change that will grow as the inclination grows. 36
- 37. Inclined Orbit Effect on Es/No • EIRP at Beam Contours will Fluxuate • This creates Es/No changes as seen by ground • How to compensate for Es/No changes – Design system for worst case Es/No – Use tools such as ACM to adjust data throughput 37
- 38. Proper Pt-Pt ACM Implementation 38 Monitor WAN Capacity Ingest Data WAN OP Prioritize Data based on filter rules Drain Data Based on drain rules
- 39. Monitor WAN Capacity 39 FX QOS/PEP CDM-750 Feedback TX Capacity • Constant monitor of the CDM-750 TX WAN capacity ensures QOS / WAN OP are enforced and maximize throughput
- 40. Voice Voice 40 Proper Pt-Pt ACM Implementation Ingest Data FX QOS/PEP Signalling Voice Video HTML Signalling Voice Video HTML • Data Ingest of all traffic destined for the WAN All Traffic Ingested
- 41. Proper Pt-Pt ACM Implementation Prioritize Data • Various rules and filters can be used • All traffic is associated with a priority classification 41 Voice Voice FX QOS/PEP Signalling Video HTML Signalling Voice Video HTML Rules / Filters Protocol IP Subnet VLAN ID DiffServ Destination Port Priority Classification 1 2 3 4
- 42. Proper Pt-Pt ACM Implementation Data Drain • Drain rules are applied to minimize latency and ensure high priority traffic maintains CIR 42 Voice Voice Signalling Video HTML Signalling Voice Video HTML Priority Classification 1 2 3 4 Drain Options Strict Priority Fair Weighted Min / Max Signalling Signalling Voice Voice Video Drain Algorithm
- 43. Carrier ID Reduction in Satellite Carrier Interference
- 44. Carrier ID to Reduce Interference • By 2009, Satellite operators and service providers recognized that the time they were spending on interference mitigation was increasing • The SUIRG had been created a few years earlier, and had proposed the use of the Network Information Table (NIT in the MPEG-2 profile) as the location for Carrier ID • The SUIRG started searching for techniques to address installed legacy equipment for VSAT and Data modems • They created requirements for this Carrier ID; 44Comtech EF Data Proprietary
- 45. The SUIRG Carrier ID Requirement • The Carrier ID must be read in the clear, by a properly configured Carrier ID receiver, even if the referenced carrier is encrypted. • The Carrier ID must be transmitted in an industry accepted format, so that the number and type of Carrier ID receivers are kept to a minimum. • The Carrier ID insertion must have a minimal effect on the data carrier overhead, efficiency, Es/No, phase noise and other carrier quality measurements. 45Comtech EF Data Proprietary
- 46. Comtech EF Data Created the MetaCarrier™ for Carrier ID • MetaCarrier means that we have a separate spread spectrum carrier that contains information that is used to describe another single carrier • The MetaCarrier is embedded within the carrier, without adding appreciable noise to the carrier, and is completely waveform agnostic • This version works for static carriers, video and SCPC • During 2010 this had been demonstrated to Intelsat, SES, Turner, and CBS with a month long national feed demo • The WBU-ISOG, sIRG, RFI-EUI and GVF started supporting and promoting it
- 48. • During the 2011 IBC the sIRG held a meeting with modulator manufacturers with the intention of submitting it to the DVB for standardization • In December 2011 the first DVB Commercial Module meeting for Carrier ID took place with the DVB and DVB members • The commercial requirements were adapted by the DVB Commercial Module on Feb 15 2012 • The standard is based on our original design along with complimentary input from others (noteably Newtec) • Technical details you may not care about such as scrambler method, power levels, transmission sequencing etc. have slightly changed since our original design • The Technical Module is underway and will submit it’s draft for comment at the next general TM meeting in January 2013 Carrier ID Progress within the DVB
- 49. Carrier ID Progress Outside of the DVB • Longer term operation underway during and after this past Summer Olympics – MetaCarrier Carrier ID embedders installed on Intelsat North America, SES Europe and Eutelsat Carrier ID test carriers • Detectors for the MetaCarrier Carrier IDs are installed at Carrier Monitoring System manufacturers for the development of their specific monitor and control interfaces – Crystal Solutions – Sat Corp (Monics) – Siemens AG • The FCC has issued a notice of proposed rulemaking for part 25 (satellite operation) that recommends adaptation of either the NIT or a spread spectrum technique; comments from industry are due by December 24th
- 50. Carrier ID Implementation Requirements • For implementation in a modulator – The modulator must have enough FPGA resource to support this additional feature – This may require a trade-off by the manufacturer to drop legacy features that the market no longer requests – Comtech modulators will support the DVB Carrier ID starting three months after the DVB standard is issued Comtech EF Data Proprietary & 50
- 51. African Design Success GSM over Satellite
- 52. Success Story • Africa – Providing service across 14 African countries including Chad, Ghana, Niger, Burkina Faso, Sierra Leone, Nigeria, DRC, Congo Brazzaville, Zambia, Madagascar, and Tanzania – Bundle CDM625 CnC + CXU Abis Optimizer solution CDM625 with CnC and VersaFEC for satellite efficiency CXU for GSM 2G Abis layer efficiency – Deployed as point-to-point Abis links (BTS) backhaul – Standardized solution with eye toward OPEX savings. DS0 Aggregation -D&I- and 2:1 optimization SAT modemSAT modem GSM BSC GSM
- 53. Success Story • Amongst fierce competitive and price pressures where every tactic is important, the measurable impacts are therefore in the following areas: – OPEX decrease in satellite transponder lease – Minimum US$600K OPEX savings per year just on upgrading the existing links (new deployments accounted separately) – Increase in number of customers in rural areas due to extended coverage, high quality of service, affordable services – Decrease in end user pricing 53
- 54. Success Story • Airtel added 8.9 million customers during the year (March 2012 estimates). The growth of 20% in the customer base has translated into higher consumption of minutes on the network, thereby driving robust revenue growth. And, the satellite backhaul strategy has been a key element in supporting the growth. 54
- 56. Comtech EF Data 2114 West 7th Street Tempe, AZ 85281 USA Tel +1.480.333.2200 FAX +1.480.333.2540 [email protected] www.comtechefdata.com