Presentation on theme: "Power saving in (X)GPONs"— Presentation transcript:
1Power saving in (X)GPONs Joint ITU/IEEE Workshop on Ethernet - Emerging Applications and Technologies (Geneva, Switzerland, 22 September2012)Power saving in (X)GPONsFrank EffenbergerRapporteur Q2/15VP Access R&D, Futurewei
2IntroductionPower saving in the ITUSupplement G.45Actual power consumption
3Initial activitiesFirst contributions were made by semiconductor vendors, to consider signaling methodsThe Q2/15 group thought these were prematureThe requirements for power saving were not clearThe impact on existing systems was not clearThere was a concern it would degrade the user experienceIt was agreed that a survey would be made of the operators to learn their requirements
4Survey gathered information on Power saving survey #1Survey gathered information onBasic power supply designsWho pays, Who changes the battery?Overall requirements and interest in power savingThe most telling result was: Which is a higher priority, service availability or power savings?The answer was overwhelmingly “Service quality is much more important that saving power”Apparently, the “Green Revolution” had not yet happened
5Power saving surveys #2 and #3 Survey #2 focused on the case of power failureWhat UNI’s are in common use?How do they get powered down if the main power fails?Who can control this powering down process?Survey #3 focused on regulations about lifelineAre their regulations that force the maintenance of service during a power failure?How do those regulations vary from service to service (POTS vs. video vs. Internet)?
6Outcome of the surveysIt was clear that there was interest in power savingFor the most part, power saving was seen as a way to lengthen the life of the battery during an outageOperators were not willing to compromise much in the name of power savingCan’t cost any moreServices can’t be effectedThis set the stage for the G.sup45 document
7IntroductionPower saving in the ITUSupplement G.sup45Access power consumption
8Outline of G.sup45RequirementsClassification of techniquesSignaling of ONU operationsComparative analysisConclusions
9Requirements Surveys were used as a primary requirement EU CoC Power saving mode should be triggered by power failure, and NOT low traffic or unused portsEU CoCThe current state of these power targets was reviewedThe “low power” mode was noted to be only half of full power modeThere are two main requirementsTo maintain service during a power failureTo save power at all timesNo operator consensus on the balance between these two requirements
11Power Shedding When triggered, the ONU turns off the UNI’s Turn off in this context means fully powering down the circuit (not just deactivating service)Each UNI type can have a different shut-down period (e.g., video can turn off after 30 minutes, but POTS is maintained for several hours)When to trigger is a questionDuring power failure, some UNI’s can be turned off safely (e.g., video, because the TV sets will not have power in most cases)During normal times, it is difficult to judge if UNI is busyThis is the least service effecting methodONU maintains contact with OLT at all times
12DozingWhen triggered, the ONU should stop transmitting in the upstream, even if it is given BW allocationsThis allows transmitter circuitry to go into low power modeLower than normal “off” in between burstsMay take longer to recover (10’s of milliseconds)Trigger would be the inactivity of the ONUDifficulty is that data services (and VoIP) tend to “chatter” all the timeDownstream receiver and signal chain remains onONU can be signaled by the OLTIncoming calls can be received without delayThis impacts services slightlyOutgoing communications might suffer a delay, as normal bidirectional communication is reestablished
13Fast SleepWhen triggered, the ONU shuts off entire PON interface for a short period of timeONU periodically wakes up to see if OLT has anything to sayDuring the shutdown, the ONU could conceivably have nearly zero power drain (only the wake-up timer would be powered)Key issue is how fast can you wake up the opticsNormal transceiver designs are not optimized to turn on fastHowever, proper optimization could get times down to ~1 msSide note: Recent result have shown circa 60% reductionsThis method can have relatively low service impactONU maintains contact with OLT (albeit transiently)Interactions with higher layer protocols must be considered
14Deep Sleep When triggered, the ONU completely shuts off Services are definitely impacted, no apologies for thatPower drain is zero, or nearly soChallenge: How to wake up?Snow White method: A prince (the user) kisses the deep sleeper (presses a button on the ONU)Rip Van Winkle method: Deep sleeper wakes up after a preset time, and sees if anything has changedThis method only appropriate for long outagesIt seems that the usual obligations are excused if power is out for a long time, and users and regulators understand that
15Signaling of ONU operations Dying gasp: Enhancing the existing messageFor G-PON, not accepted, because it changes the TC-layerPLOAM-based: Signaling for fast sleep methodFor G-PON, not accepted , because it changes the TC-layerOMCI-based: Configuration of power featuresFor G-PON, OMCI additions have been madeExtended Power shedding: Detailed controlFor G-PON, fine-grain control of shedding has be standardizedImplicit signaling: OLT suppresses alarmsNo standards impact, so OLT vendors are free to implementSecurity aspect: Impostor attackWhen the ONU is asleep, impostor can more easily jump in
16Key findings of this evaluation Comparative analysisModel of ONU power consumption is given, and used to evaluate the savings for each type power savingThis model is only an example, based on a particular ONU design and circuit power values (these change over time)Key findings of this evaluationPower shedding accomplishes a lot (70%) of power savingThe other methods have increasing implementation difficulty and declining efficacy of power saving
17Conclusions of G.sup45 Power saving is an important topic Main object is to improve handling of power failuresRecommendations to improve power usageContinuous improvement of design (ASIC, optics, power conv. Etc.)Power shedding should be supported and activatedDozing can be implemented with little cost“Aggressive” sleeping modes are of lowest priorityFinal note: G-PON saves power in ICT field and other industries, so some credit should be given for that
18IntroductionPower saving in the ITUSupplement G.45Access power consumption
19System architecture of a VDSL system DSLAMCPELTSWWANLTLTCPESWWANLTLTCPETypical VDSL linecard consumptiontoday is 2W per line (i.e., per user)Typical VDSL HG CPE consumption is 10W per user
20System architecture of G-PON system OLTONULTSWWANLTLTONUSWWANLTLTONUTypical OLT linecard consumptiontoday is 7W per PON port28 users/PON = 250mW/user!)Typical GPON HG CPE consumption is 10W per user
21Central office / Node “power crunch” Central office dissipation is dictated by NEBSTypical US number: 2000W per bay, 3 racks per bayTypical DSLAM has 16x24 lines = 768W per rackThis barely fits in the 2000W numberTypical OLT has 16x8 PONs = 896 W per rackHave to leave 1 rack-space empty!Is PON hitting the “crunch”? NO!One OLT serves 3584 users, while a DSLAM serves only 384 usersWe need 9 times fewer OLTs than DSLAMs
22Trend in broadband access CO equipment The power per chassis is increasing marginallyPerhaps a 30% increase generation-over-generationThe capability per chassis is increasing incrediblyAggregate bandwidth increases 4~10x per generationUsers per chassis increased ~10x from copper to fiberTotal access power per user is already decliningDriven by the acceptance of fiber accessPower density is increasingSuggests a rethinking of the CO power design guidelinesPerhaps even a redesign of the cooling method entirelyDiffused air cooling (typical in today’s CO) is inappropriate for intense point heat loads
23The CPE power issue - Functional blocks MemPOTSInterf (x2)WANinterfaceMACEthernetInterf (x2)Typical Single family home gateway CPE consumption is 10W
25Observations on baseline consumption Power consumption is reasonably balanced amongst functions – there is not one “bad actor”The majority of power (60%) lies in functions that are not particularly related to PONYou find them in any access systemMany are legacy dictates (ringing a bell)They are designed for reliability and performanceE.g. Power converters consuming 20% of the power… why? To handle the stress environment that Telco requirements give usFlexible hardware (e.g., CP instead of ASIC) is usedThe flexibility is a meta requirement of the ever changing marketBut, this is never the most power efficient way to build equipmentIf the true power cost of all the requirements was rationalized, just imagine what we might save!
26Power saving technologies Most natural path is intrinsic improvementCurrent designs were not designed with power as a key requirementTime to market, performance, and simplicity were always more important to the designerExample: burst mode laser driverThe average ONU duty cycle is ~3% (32 ONUs per PON)But, the typical laser driver consumes current 100% of the timeWhy? Because it was easier that wayThis is straightforward to fixThe designers only need to be guided that power consumption is an important goal that has valueThis process is underway already!
27“Always on” means “always polluting” Recall the original telephone networkYou only used power when off hook – very efficient, and natural behavior to the userData separated the “session” from userIdeally, users interact with their computer, and the computer establishes the (logical) sessions automaticallyUser involvement in session control (dial up) was slow and painfulThis quickly drove the “always on” modelPower consumption was not considered!
28Sleep modes for access equipment Protocols for sleeping and dozing are standardized in the XG-PON systemONU state diagramOLT state diagram
29Future possibilitiesOLT power consumption could be reduced in future PON systemsOLT “shedding”: If a port is not used, it should be powered downAs deep into the card as possibleOLT “sleeping”: If a TWDM-PON is underused, reduce the active wavesONUs would be concentrated onto fewer channelsThis could improve the load-dynamic power consumption of the CO
30Getting a good night’s sleep Standardization is only the beginningThe hardware must be designed to use itOptoelectronics must have fast turn-on/offLogic devices must support the protocolsSwitching must recognize that link is transientThe operators must be motivated to use itOperators respond to competitors and usersIn a choice between performance and power-saving, which wins?Example: ADSL has power saving for some years now – almost never used
31ConclusionsCurrent access power consumption is trending in the right direction, considering the incredible BW improvementsThe CO-side solution is in our hands: deploy PON, and you cut your CO power by an order of magnitudeThe CPE-side is much larger problemLegacy interface requirements are an issueIf we could only redesign POTS…If only Telco’s could agree on a service profile and stick to it…Power saving modes have good potentialChanging “always on” into “always available”Already standardized – we just have to do it