EP2883318A1 - TIME DIVISION DUPLEXING FOR EPoC - Google Patents

Abstract

A method, system, and computer program for implementing TDD in an EPoC network. Multiple PHY-layers for coax or HFC segments operate cooperatively to establish a constant delay for an upstream path, and a constant delay for a downstream path, resulting in a constant RTT rangeable by either: an EPoC CLT (116); or an EPON OLT with an EPoC OCU. The delay for at least one PHY-layer is not established as constant, but suffers variable delays (100), while at least one other PHY-layer establishes delays which are Complementary (102) to said variable delays.

Description

TIME DIVISION DUPLEXING FOR EPoC RELATED APPLICATIONS

[0001] The present application claims priority to U.S. Provisional Applications Nos.

61 /681 ,808, filed August 10, 2012, and 61 /692,610, filed August 23, 2012, assigned to assignee hereof the specification of which is incorporated herein by reference herein. [0002] The present application is related to co-pending U.S. Patent Application, Ser. No.

13/890,115, filed on May 8, 2013, assigned to the assignee hereof and expressly incorporated by reference herein. FIELD

[0003] This disclosure is related to a communication system and more particularly to

extending Ethernet Passive Optical Networks (EPON) Protocol over Coax based access networks. BACKGROUND INFORMATION EPON

[0004] EPON is an IEEE 802.3 protocol specification enabling Ethernet Passive Optical

Networks. Passive Optical Networks (PONs) use an Optical Distribution Network (ODN) generally using passive fiber-optic cables and passive optical splitters forming a point-to-multipoint topology. EPON is often deployed by Operator/Service Providers (OSPs) as an Access Network, to provide high-speed access to the internet backbone and Business Services to medium-to-large businesses seeking strict Quality of Service (QoS) Service Level Agreement (SLA) contracts including low-latency, low-jitter, and guaranteed throughput. Typically there is an Optical Line Terminal (OLT) at the headend (e.g., located in an OSP’s central office site), and there is an Optical Network Unit (ONU) at each of one or more Customer Premise Equipment (CPE) endpoint sites. The service group for an EPON OLT often comprises up to 16~32 ONUs. The headend OLT can send messages Downstream (DS) over the ODN point-to-multipoint, and the ONUs at the CPE endpoints can send messages to the OLT multipoint-to-point over the ODN. The OLT produces downstream messages in the form of serial binary bitstreams that are converted to optical signals (e.g., OOK On-Off-Keying pulses produced by so-called‘digital’ laser) onto a fiber- optic cable and into the ODN to reach each ONU at the CPE endpoints. The ODN generally comprises passive optical components, so substantially the same optical signals reach all of the ONUs. However, due to ODN topology (e.g., lengths of fiber and location of splitters), there are generally differences in propagation times among all the branches in the ODN, often resulting in differing arrival times and differing arrival amplitudes of the optical signal among all the ONUs. [0005] The OLT produces the downstream serial bitstream at some constant EPON data-rate, such as 1 Gbps or 10Gbps. If there are no messages to send downstream, then the OLT will transmit IDLE characters between data traffic. Thus, EPON downstream traffic is a continuous bitstream at some constant EPON data-rate. [0006] Upstream (US) transmissions are formed by ONUs as a serial binary bitstream, but are generally not continuous, so upstream traffic from a plurality of ONUs is coordinated by the OLT in order to ensure that non-continuous so-called burst transmissions from various ONUs do not collide (overlap in time) and that the OLT will observe an orderly sequential arrival of burst transmissions from different ONUs in a predictable order and at predictable times (within some tolerance of time-jitter). This approach is often called TDMA time-division multiple access.

[0007] There are three versions of EPON currently specified:

1Gbps symmetric (formerly amendment 802.3ah);

10Gbps symmetric (amendment 802.3av-2009); and Asymmetric 1Gbps upstream, 10Gbps downstream (802.3av-2009). [0008] Upstream (US) traffic generally uses the same wavelength for both 1 Gbps and

10Gbps data-rates. Downstream (DS) traffic generally uses different optical wavelengths for 1 Gbps and 10Gbps data-rates. It can be deduced that there is interest in supporting both symmetric and asymmetric upstream/downstream data-rates. [0009] Since EPON’s upstream traffic and downstream traffic use different wavelengths, bitstreams can be transmitted over the ODN in both directions simultaneously and independently (i.e., full duplex). This particular duplexing strategy is called Wavelength Division Duplex (WDD), or more generally, Frequency Division Duplex (FDD). The OLT has exclusive use and access to the downstream wavelength(s), and the OLT can coordinate/schedule use of the upstream wavelength independently from the downstream. [0010] OLTs use EPON’s Multipoint Control Protocol (MPCP) to coordinate and schedule the TDMA upstream bursts. The MPCP protocol relies on constant Round-Trip Time (RTT) as observed and measured by the OLT. The OLT may measure a different RTT for each ONU, but that RTT must remain more or less constant (within some tolerance). MPCP messages include timestamps to facilitate OLT’s measurement of RTT. Each ONU maintains its own MPCP Clock by setting its clock counter value to that of the OLT’s timestamp embedded in downstream MPCP messages received from the OLT. Since fibers to each ONU may have varying length, the MPCP Clocks among different ONUs are not necessarily synchronized. The RTT comprises a downstream trip plus an upstream trip, which may be different (e.g., different wavelengths may propagate at different velocities on a fiber). The OLT will observe/measure RTTs, but may also know (e.g., be configured for) or assume some fractional split (e.g., 50%: 50%) of the RTT into separate downstream and upstream link delays. [0011] ONUs hold traffic destined for the OLT in various queues often associated with particular Service Flows (e.g., an ordered sequence of Ethernet Frames with similar classification), and identified by Logical Link Identifiers (LLIDs) assigned by the OLT. ONUs report the status (e.g., fullness) of their various upstream queues in the form of a MPCP REPORT message. The OLT receives such REPORTs from the ONUs, then the OLT’s MAC Control Client (aka Scheduler) schedules upstream traffic from the various queues of various ONUs, then issues TDMA grants to particular ONUs in the form of MPCP GATE messages. All upstream traffic is scheduled or granted in this fashion, even REPORT messages must be granted via a GATE message in the downstream. GATE messages grant a startTime and a length. When an ONU’s MPCP Clock reaches the GATE-specified startTime, the ONU transmits upstream at the constant EPON data-rate, from the GATE-specified LLID queue, and for a duration equal to the GATE-specified length. The GATE-specified grant yields an upstream transmission of some integer number of Layer 2 payload bytes (the exact number of bytes is known to both ONU transmitter and OLT receiver), which usually corresponds to some integer number of variably-sized Ethernet Frames. [0012] The OLT’s scheduler arranges the grants, ensuring the OLT will observe an orderly sequential arrival of burst transmissions from a plurality of ONUs, arriving in a predictable order and at predictable times (within some tolerance of time-jitter). The OLT’s scheduler understands that grants will depend on the RTT for each particular ONU. For example, the OLT could transmit downstream two GATE messages with identical startTime and identical short grant length, destined for two different ONUs, one with 1 km effective fiber length, and the other with 20km effective fiber length; understanding that the consequent upstream transmissions will not overlap/collide with each other, due to their differing RTTs (i.e., the upstream transmission from the more distant ONU will arrive after that from the nearby ONU). [0013] In summary, EPON protocols were designed around assumptions based on FDD simultaneous US and DS optical fiber transmission:

a) serial bitstream emissions from Layer 2 submitted into Layer 1 of a transmitting device;

b) serial bitstream undergoing constant processing delay through Layer 1 of the transmitter (Tx);

c) serial bitstream undergoing constant propagation delay through the ODN;

d) serial bitstream undergoing constant processing delay through Layer 1 of a receiving device;

e) received serial bitstream being submitted to Layer 2 of a receiver (Rx);

f) resulting in a constant US link delay for a given serial bit, from ONU Tx Layer 2 to OLT Rx Layer 2;

g) resulting in a constant DS link delay for a given serial bit, from OLT Tx Layer 2 to ONU Rx Layer 2;

h) US + DS link delays summing to a constant RTT, bit-for-bit, as observed/measured by the OLT.

[0014] There are other PON specifications, such as APON, BPON, and GPON, which share many of the same characteristics as EPON so this disclosure applies to them as well. HFC Access Networks

[0015] A publicly-available overview of hybrid fiber and coaxial (HFC) Cable Systems (e.g., slides 5 & 6) can be found at:

http://www.ieee802.org/3/epoc/public/mar12/schmitt_01 _0312.pdf. HFC Cable Access Networks are typically deployed by multiple system operators (MSOs), which are OSPs that operate multiple HFC cable systems. They are used to provide subscribers access to a variety of services, such as pay television (TV), video on demand (VoD), voice over internet protocol (VoIP) telephony, residential cable modem internet service, and small-medium business (SMB) Business Class Internet service. These various services have been designed, and the plants engineered, to support simultaneous coexistence on the shared HFC medium. The point-to- multipoint topology deployed varies according to the size and footprint of the service group of CPEs, and how distant they may be from the headend (or Hub). For example, in China, the service group is often a multiple dwelling unit (MDU) with dense concentration of the CPEs in the service group, and relatively short distance to the headend often located in the basement (e.g., Fiber-to-the-Basement (FTTB)). For example, in North America, the service group may be larger and more dispersed (e.g., spanning suburban neighborhoods), and the headend might be remotely located (e.g., tens of miles away). [0016] CPE endpoints are connected via coax (coaxial cable), and the coax plant is driven by one or more Radio Frequency (RF) amplifiers passing a variety of modulation techniques depending on the particular service and its assigned spectral occupation in the RF band (typically within 5~1002MHz). Smaller plants can be serviced by coax alone, so the headend can interface the coax plant directly. Remote headends can drive the HFC via fiber, with Fiber Nodes deployed at various locations in the middle of the network to convert to/from fiber and coax. These‘analog’ fiber plants in HFC networks are typically driven by‘analog’ lasers, modulating the amplitude of the optical signal in direct correspondence to an RF signal waveform (i.e., amplitude modulation (AM)). Fiber Nodes perform a relatively direct media conversion:

DS: from RF-modulated optical signal on fiber to RF electrical signal on coax to the CPEs;

US: from RF electrical signal on coax to RF-modulated optical signal on fiber to the headend;

where imperfect Optical-to-Electrical (OE) and Electrical-to-Optical (EO) conversions, along with AM transmission over fiber, contributes impairments to the RF signal fidelity (e.g., degradation of Signal-to-Noise Ratio (SNR)). [0017] The topology of the coax plant is a cascade of various active and passive components, such as amplifiers, rigid trunk-line coax, feeder-line coax, multitaps, drop-line coax (to individual customer premises), and RF splitters. Cascade lengths vary from:

a) Node+0’ cascades: with zero active components (e.g., no in- line amplifiers) after the Fiber Node (if any), meaning the coax plant contains only passive elements (e.g., taps or splitters). Node+0 plants are quite common in China. They are less common among North America MSOs, but remain a goal for the future evolution of their HFCs.

b) Node+1: with one active amplifier after the Fiber Node (if any);

c) Node+2: with two active amplifiers after the Fiber node (if any);

d) Node+N: with N amplifiers (e.g., Node+5 cascades are common among North American MSOs’ HFCs). [0018] Many HFC plants have been deployed with FDD operation within certain frequency bands, using diplex filters installed throughout the HFC infrastructure (e.g., within various RF amplifiers). This FDD infrastructure was often deployed decades ago, before the advent of widespread internet use, and MSOs now find their existing split locations to restrict future use cases. In particular, MSOs are studying the possibility of moving the split location to allocate additional spectrum for the upstream channel. Moving the split is an expensive and labor-intensive upgrade that may require thousands of truckrolls to deploy (and with consequent service disruptions), so MSOs try to anticipate the evolution of future usage. Predicting the future presents its own risks if the MSOs guess wrong, but this is the predicament that MSOs find themselves in having FDD HFCs already deployed. [0019] The coax plants of HFC networks in North America are often operated as FDD within

US spectral allocations (typically 5~42MHz) and DS spectral allocations (typically from 54MHz up to 750, 860 or 1002MHz as examples), with an allowance for a so- called‘Split’ or guard band (typically 42~54MHz) where FDD diplexing filters are used to isolate the simultaneous US & DS transmissions from each other. Coax plants of HFC networks outside North America might be operated with a different FDD split location in the spectrum. An example of a FDD service is Data Over Cable System Interface Specification (DOCSIS), wherein cable modem service may occupy one or more single-carrier‘QAM’ channels occupying 6MHz of spectrum in the DS band, and one or more QAM channels in the US band. DOCSIS headend equipment is known as a Cable Modem Termination System (CMTS). DOCSIS CPEs include Cable Modems, Residential Gateways and Set-Top Boxes. [0020] As subscribers consume more and more throughput capacity in both upstream and downstream directions, MSOs have lashed more and more fiber overlaying the existing coax infrastructure in order to locate additional Fiber Nodes deeper into the cascade. This has the effect of segmenting the cascade, thereby reducing the service group size such that each subscriber competes with fewer neighbors for shared coax resources, resulting in greater throughput capacity available to CPEs. DOCSIS revisions, such as version 3.1 , continue to improve capacity to address the seemingly inevitable migration to‘All-IP’ (Internet Protocol packetized) delivery, including video.

EPoC: EPON Protocol over Coax

[0021] MSOs currently must deploy fiber to the premises to support EPON for high-end

Business Services subscribers. This often involves digging trenches or other significant cable-laying expenses, even if those customer premises are already passed by the coax plant of a MSO’s HFC network. The MSO may already offer Business Class Internet (DOCSIS) services over the existing HFC plant, but some subscribers will require strict QoS performance (such as that described by the Metro Ethernet Forum specification MEF-23.1) SLAs that may require EPON to satisfy.

Consequently, MSOs desire an invention that would reduce expenses by enabling deployment of EPON-class QoS to subscribers without having to deploy fiber to the premises, but instead utilizing the existing HFC plant, or the coax portion of the HFC plant. In addition, EPON OLTs are significantly less expensive than DOCSIS CMTSs, which can further reduce MSO expenses. Thus, EPoC represents a desire for MSOs to have a lower-cost option of using the existing HFC medium for EPON- like services. [0022] MSOs also desire that EPoC devices be manageable in some similar way as they manage EPON (e.g., DPoE DOCSIS Provisioning of EPON specification from CableLabs). Consequently, there is a desire to maintain most/all of EPON’s layers and sublayers above Layer 1 PHYsical layer. Most particularly, the IEEE EPoC effort seeks to preserve unchanged EPON’s Ethernet Medium Access Control (MAC) Sublayer within Layer 2, and to make only‘minimal augmentation’ of other sublayers in Layer 2 (e.g., in the MPCP sublayer) and higher layers (such as Operations, Administration, and Management (OAM)), by confining most of the new RF coax protocols to a Layer 1 PHY specification. MSOs believe that end-to-end management of EPoC devices will be easier to accomplish if a single EPON MAC domain can span from OLT to EPoC CPEs. Consequently, there is a desire to make operation of EPoC CPEs transparent to the OLT. Since EPON protocols were designed around an FDD medium, and because North American MSOs have already deployed FDD HFCs, EPoC intends to support FDD over coax. EPoC Architecture

[0023] EPoC CPEs, which connect directly to the coax plant 20, are called coax networking units (CNUs) 10, and are desired to resemble ONUs 12 at Layer 2 and above, as illustrated in Figs. 1 and 2. An un-augmented or minimally augmented EPON OLT 14 connects to fiber plant 16 at the headend. In one embodiment, an optical-coax unit (OCU) 18, aka FCU fiber-coax unit can be located somewhere in the middle that performs bidirectional conversions from EPON’s‘digital’ fiber 16 to RF coax 20. OCU 18 and its conversions are desired to be transparent to OLT 14 so that the OLT can remain un-augmented or minimally augmented. In the presently claimed invention, an OCU may filter-out DS payloads (based on LLID or some other criteria) that are not intended for CNUs residing on the coax that the OCU services. In other words, the digital fiber may carry payloads intended for ONUs, or intended for CNUs belonging to some other OCU, and it is desirable for OCUs to filter-out these payloads out before relaying DS traffic onto the RF coax in order to avoid unnecessary traffic from consuming coax resources. SUMMARY

[0024] The presently claimed invention provides solutions to the problems raised above.

EPoC specifically contemplates a new coax line terminal (CLT) 22 device that would resemble an OLT, but instead interface via RF signals, either to the‘analog’ fiber 24 at the headend of an HFC, or directly to the headend of an all-coax plant, as shown in Figs. 3 and 4. [0025] Preserving the EPON MAC sublayer at both endpoints implies PHY-layer processing and transport of the serial bitstream with constant RTT, corresponding to the sum of the downstream and upstream link delays:

in the downstream, measured from emission by the CLT/OLT MAC sublayer, to submission to the EPON MAC sublayer in the CNU; and,

in the upstream: measured from emission by the EPON MAC sublayer in the CNU, to submission to the CLT/OLT MAC sublayer. [0026] In EPoC, there may be alternative ways of measuring DS and US (and hence RTT) delays, such as measuring the difference in arrival times of Ethernet frames, instead of the usual bit-for-bit delay of the serial bitstream. [0027] An FDD mode of operation for EPoC seems certain. In the FDD mode of operation, downstream traffic gets converted relatively directly by the OCU from WDD/FDD over digital fiber into FDD over RF coax. Such relatively direct conversion by the OCU is also known as Media Conversion (aka PHY-level Repeater), since there is little complication beyond straightforward conversion from fiber medium to coax medium. Similarly, upstream burst traffic from CNUs gets converted by the OCU from FDD on coax to WDD/FDD on digital fiber. In the FDD mode of operation, the OCU performs media conversions for both downstream and upstream traffic simultaneously, by using to two different RF channels over coax. Such PHY-layer Media Conversion can be accomplished with constant processing delay to satisfy EPON protocols’ reliance on constant RTT. [0028] However, many MSOs desire an additional TDD mode of operation for EPoC. Such a TDD mode seems quite challenging to specify because the EPON protocols that MSOs wish to preserve were specifically designed only for FDD’s simultaneously available full-duplex US & DS channels. EPON protocols were not designed for alternative duplexing strategies, such as Time-Division Duplex (TDD), where a single wavelength or RF spectral channel-width would be used, alternating-in-time between upstream and downstream (half duplex). TDD’s single half-duplex channel alternates between US and DS traffic, which implies the DS link would be unavailable during US traffic, resulting in fluctuating delays for DS traffic while waiting for the DS phase of the TDD Cycle and vice versa, resulting in fluctuating delays for US traffic while waiting for the US phase of the TDD Cycle. Further complicating the challenge of TDD operation are EPON constraints outlined above such as maintaining constant RTT, and the desire to preserve unchanged the MAC sublayer. [0029] Despite these severe challenges, MSOs nevertheless wish to consider such a mode due to TDD’s increased flexibility (compared to FDD) for adapting to the evolution of future US and DS traffic patterns. One benefit of TDD is that the symmetry or asymmetry of the US and DS capacities is a relatively simple (and possibly realtime) adjustment of the duty-cycle phasing of the TDD Cycle. Use of TDD in the Access Network would have enabled a more flexible way for MSOs to easily, quickly, and inexpensively adjust the relative throughput capacity of the upstream and downstream directions within a single spectral allocation, whereas FDD requires paired spectral allocations established by inflexible diplex filters distributed throughout the coax cascade. For a given total aggregate spectral allocation, TDD’s single spectral allocation could be made as wide as the sum of FDD’s paired allocations, enabling TDD’s burst datarate capability in either direction being approximately double that of FDD in either direction (for symmetric US and DS FDD allocations). Use of TDD in the Access Network would have enabled fewer or no splits in some coax plants. [0030] The following presents a simplified summary of one or more embodiments in order to provide a basic understanding of some aspects of such embodiments. This summary is not an extensive overview of the one or more embodiments, and is intended to neither identify key or critical elements of the embodiments nor delineate the scope of such embodiments. Its sole purpose is to present some concepts of the described embodiments in a simplified form as a prelude to the more detailed description that is presented later. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The disclosed method and apparatus, in accordance with one or more various embodiments, is described with reference to the following figures. The drawings are provided for purposes of illustration only and merely depict examples of some embodiments of the disclosed method and apparatus. These drawings are provided to facilitate the reader’s understanding of the disclosed method and apparatus. They should not be considered to limit the breadth, scope, or applicability of the claimed invention. It should be noted that for clarity and ease of illustration these drawings are not necessarily made to scale. [0032] Fig. 1 is an illustration of an OLT to ONU fiber connection and an OCU

conversion from a fiber to a coax for CNUs. [0033] Fig. 2 illustrates the OCU conversion of Fig. 1. [0034] Fig. 3 illustrates a new CLT that resembles the combination of an OLT plus an

OCU that interfaces to an analog fiber. [0035] Fig. 4 illustrates a CLT that connects directly to a coax. [0036] Fig. 5 illustrates a DFT/iDFT embodiment that automatically establishes the complementarity to the variable delay, wherein at least one PHY-layer delay is variable, while at least one other PHY-layer establishes a delay which is complementary to the variable delay. [0037] Fig. 6 illustrates an alternative embodiment of rate-adaptation by instead postponing or delaying the US fiber transmission until sufficiently enough of the US burst has been received at the OCU via coax, and buffered in memory, until the corresponding US fiber transmission can be made at the full fiber datarate. [0038] The figures are not intended to be exhaustive or to limit the claimed invention to the precise form disclosed. It should be understood that the disclosed method and apparatus can be practiced with modification and alteration, and that the invention should be limited only by the claims and the equivalents thereof. DETAILED DESCRIPTION [0039] Most state-of-the-art PHY-layers have no need, and make no attempt, to establish a constant DS delay time, constant US delay time, or constant RTT. These characteristics are largely unique to EPON networks. Current state of the art methods for guaranteeing these characteristics in EPON involve each PHY-layer establishing a constant delay time. Consequently, the EPON RTT includes a sequence of additive delay times, each of which may be different, but each held constant, summing to some constant RTT. Applying these same methods to EPoC would be appropriate for a FDD mode of operation, where the US delay would include:

a) fixed delay through the CNU PHY onto coax;

b) a fixed propagation delay over coax; and

c) followed by another fixed delay through the CLT PHY. [0040] Correspondingly, the DS delay would comprise:

a) fixed delay through the CLT PHY onto coax;

b) a fixed propagation delay over coax; and

c) followed by another fixed delay through the CNU PHY. [0041] These fixed delay times for US and DS would sum to produce a fixed RTT, as required for EPoC. [0042] However, these same methods would be problematic for a TDD mode of operation in EPoC that satisfies MSOs’ desire to re-use existing EPON OLTs with little or no augmentation. In this case, an EPON OLT would be largely unaware of the coax segments, or that the coax is operated as TDD, having new PHY-layers for EPoC hiding the complexity of TDD and its delays from the OLT, yet maintaining some constant RTT for each CNU. Each CNU can have a different RTT, but each of those RTTs must remain constant. TDD operation generally involves having some traffic suffer a wait until an appropriate phase of the TDD Cycle is available for transmission. Since traffic ingress generally arrives or otherwise become available asynchronously, at any time, its delay suffered waiting for a TDD Cycle phase is not constant, but variable. Such variable delays are the antithesis of EPON, and the current state of the art for EPoC, where a transmitting PHY establishes a fixed delay onto coax, and a separate fixed delay is established by a receiving PHY. [0043] The DS and US channels each are comprised of several links and paths. For example, the DS channel may comprise all or some of: the OLT’s Tx PHY, the digital fiber ODN, the OCU’s fiber Rx PHY (similar to an ONU’s Rx PHY), the OCU’s coax Tx PHY, the coax cable plant, and the CNU’s coax Rx PHY. Conversely, the US channel may comprise the corresponding path links but in reverse direction: CNU’s coax Tx PHY, the coax cable plant, the OCU’s coax Rx PHY, the OCU’s fiber Tx PHY (similar to an ONU’s Tx PHY), the digital fiber ODN, and the OLT’s Rx PHY. For EPoC

configurations using a CLT (instead of an OLT and one or more OCUs) the DS channel may comprise: the CLT’s coax Tx PHY, the coax cable plant, and the CNU’s coax Rx PHY, as well as the converse US channel comprising: the CNU’s coax Tx PHY, the coax cable plant, and the CLT’s coax Rx PHY. [0044] Fig. 5 shows an embodiment wherein at least one such PHY-layer delay is not established as constant, but allowed to remain variable 100, while at least one other PHY-layer establishes a delay which is complementary 102 to variable delay 100. The embodiment of variable delay 100 on at least one path link allows the best modes of communication technology to be used in the new coax PHY for EPoC, with complementary delay 102 enabling support for variable delay 100, while maintaining a constant US delay, a constant DS delay, and a constant RTT. Best modes of communication technology would mean, for example, use of modern PHY-layer techniques yielding high spectral efficiency, high throughput, high QoS, high robustness, low latency, or low-cost. [0045] An example of complementary delay 102 can be: EPON was designed around serial bitstreams and fixed-delay processing of those serial bitstreams.

However, many modern PHY-layers, such as those commonly used for TDD over coax cable, use various block-processing steps. Block processing 104 or 133 involves more or less simultaneous processing of blocks of information bits, often hundreds or thousands of bits at a time. The concept of EPON’s fixed bit-for-bit PHY-layer delay vanishes because hundreds or thousands of bits first need to be accumulated, then processed together as a block substantially simultaneously, then the resultant output bits are produced together as a block, all at essentially the same time. For example, RF- modulated multi-carrier transmissions 110, such as OFDM or OFDMA, may involve hundreds or thousands of subcarriers, each carrying roughly 8-to-12 bits per subcarrier per symbol, and each being transmitted concurrently.

Block processing 104 or 133 is used in OFDM, such as the Discrete Fourier Transform (DFT) 106, and its inverse DFT (iDFT) 108. Bits 112 from a serial bitstream 114 are accumulated in preparation for block processing 133. Some of those bits arrive early and must wait for subsequent remaining bits to arrive before block processing 133 can begin. This comprises variable delay 100, since the early-arriving bits must wait for a longer delay period than the remaining bits in the block that arrive subsequently. Fig. 5 depicts this example, including accumulation of bits 112 from a serial bitstream 114

(comprising a variable delay 100) in a CNU transmitter 114, block-processing 133 of de-serialized bits 112, the transmission via multi-carriers onto coax 110, the reception of multi-carriers at an OCU or CLT 116 on the coax plant, block processing 104 (e.g., DFT 106) of received bits 112’, comprising a complementary delay 102 established at a receiver during re-serialization. [0046] As shown in Fig. 5, transmitting PHY 114 does NOT establish a constant delay, but suffers a variable delay 100 for each serial bit. Receiving PHY 116 also does NOT establish a constant delay, but comprises a complementary delay 102 for each bit 112’ (i.e., a delay that is complementary to the variable delay). That is, when variable delay 100 suffered by a bit 112 is short, complementary delay 102 for bit 112’ is correspondingly long. Conversely, when variable delay 100 is long, complementary delay 102 for that bit 112’ is correspondingly short. In this example, variable delay 100 is depicted as the vertical shift-register 118 in transmitting PHY 114 with the serial bitstream clocked-in at the bottom of shift-register 118. In this example, complementary delay 102 is depicted as vertical shift-register 120 in receiving PHY 116 with the serial bitstream clocked-out from the top of shift-register 120. Both registers shift vertically, in the same direction, and at the same serial bitrate, so it is apparent from the depiction that those bits 112 that suffer a longer delay shifting through transmitting PHY 114 will incur a correspondingly shorter complementary delay shifting through receiving PHY 116, and vice versa. According to this particular example of the presently-described embodiment, it is this combination of variable delay 100 in transmitting PHY 114, plus a complementary delay 102 in receiving PHY 116, which together comprise a constant delay over a given coax path link for each bit:

Variable Delay + Complementary Delay = Constant Delay [0047] Indeed, this same expression can be rearranged to describe what

Complementary Delay must be established in order to exactly compensate for whatever variable delay has been suffered:

Complementary Delay = Constant Delay– Variable Delay [0048] Modern PHY-layers, such as those commonly used over coax, specify other best mode block-processing steps, such as Forward Error Correction (FEC), Scramblers, Interleavers, Ciphers, and/or other processes, that share similar characteristics to that of the example above. These characteristics include: An accumulation of bits, followed by block processing before transmission, with individual bits no longer necessarily having a unique transmission times like they would on EPON digital fiber. Some of these PHY-layer block- processes change the amount of data bits, such as FEC encoding which generates additional parity bits to be transmitted, thereby generating additional variability of delay. According to the presently-described embodiment, the transmitting PHY-layer need not equalize these variable delays, and can instead freely allow them to incur, because a receiving PHY-layer will establish a complementary delay. [0049] The TDD mode of operation is analogous to the example described above.

Ingress traffic may arrive at most any time, and perhaps thousands of bits of such information must in general be accumulated while waiting for an appropriate period within an appropriate phase of a TDD Cycle to become available for transmission. Some of those bits arrive early and other subsequent bits arrive later so this difference in waiting time comprises a variable delay. Rather than establishing fixed delays through each of the transmitting and receiving PHYs, the presently-described embodiments teach that the transmitting PHY can transmit this ingress traffic onto coax with variable delay, with the receiving PHY establishing a complementary delay. The resulting sum of variable delay (for a transmitting PHY) plus

complementary delay (for a receiving PHY) results in a constant delay. The upstream may comprise a pair of such compensating PHYs: a transmitting PHY in the CNU, and a receiving PHY in the CLT or OCU; while the downstream may comprise a pair of such compensating PHYs: a transmitting PHY in the CLT or OCU, and a receiving PHY in the CNU. The sum of a constant delay for the US, plus a constant delay for DS, comprises a constant RTT. [0050] There are still other PHY-layer delays that must be suffered that are variable in duration. For example, there may be a need to perform rate-adaptation in an OCU between the EPON fiber datarate, and whatever datarate can be achieved over the coax segments. MSOs anticipate that digital fiber datarates will commonly exceed the datarates that would be available over coax, given scarce limitations of spectrum allocation on coax. In the upstream, this rate- adaptation could perhaps be accommodated by slowing the digital fiber datarate to roughly match that available over coax, but this results in undesirably poor efficiency on digital fiber. An alternative embodiment comprises rate-adaptation by instead postponing or delaying the US fiber transmission until sufficiently enough of the US burst has been received at the OCU via coax, and buffered in memory, until the corresponding US fiber transmission can be made at the full fiber datarate. This postponement or delay waiting for sufficiently enough of the US burst, is shown in Fig. 6, where: MinDelay 120 > (#Bytes)÷(CoaxDataRate) 122– (#Bytes)÷(EPON FiberDataRate) 124. [0051] That is, the OCU accumulates bits from the US burst for a time period at least as long as calculated MinDelay time 120. As can be observed in these expressions, MinDelay time 120 is a variable delay, which depends on the length of each burst (#Bytes 126) as well as the coax 132 and digital fiber datarates 130. The length of US bursts (#Bytes 126) in-turn depends on the duration of GATE grants, which in EPON can be quite long (e.g., maximum GATE duration of 1.05 milliseconds at 10Gbps digital fiber datarate, or roughly one million Bytes). Rearranging the above expression yields:

MinDelay 120 > GATEmax × {(EPON FiberDataRate 130)÷(CoaxDataRate 132)– 1}, which describes MinDelay times that could easily exceed EPoC’s target latency and QoS requirements from MSOs. In a preferred embodiment, the OLT scheduler limits the duration of its GATE lengths in order to limit the variability of MinDelay duration, and thereby improve latency and QoS. [0052] In the DS direction, when the coax datarate 132 is greater than the digital fiber datarate 130, the embodiment for rate-adaptation closely corresponds to that described for the upstream, comprising a variable delay time in the OCU before launching the DS transmission onto coax described by: MinDelay 120 > (#Bytes 126) ÷ (EPON FiberDataRate 130)– (#Bytes 126) ÷ (CoaxDataRate 132), with the receiving PHY in the CNU establishing the complementary delay. [0053] In a preferred embodiment, the OLT scheduler shapes its DS traffic destined for CNUs serviced by an OCU, avoiding long uninterrupted DS fiber transmissions, in order to limit the variability of the OCU’s MinDelay duration. Shaping controls the duration and occurrence rate of traffic, thereby limiting the variability of the #Bytes parameter, and hence the variability of the OCU’s MinDelay duration. This in-turn reduces the variability of the complementary delay established by the CNU’s receiving PHY, thereby improving EPoC’s DS latency and QoS. The OLT could implement such traffic shaping by scheduling so-called Pause or Idle symbols to interrupt DS fiber transmissions destined for CNUs serviced by an OCU, while maintaining the constant serial bitrate at the EPON MAC sub-layers. [0054] In some embodiments, the complementarity between two PHY-layers: one incurring a variable delay; the other a complementary delay; can be established by design of the two PHY-layers according to the presently- described embodiments. The DFT/iDFT example described in Fig. 5 is one such example, where the design, as depicted, automatically establishes the complementarity. In other cases, it may not be possible to cost-effectively establish such automatic complementarity. In an alternative embodiment, the transmitting PHY-layer includes in its transmission some indication comprising information to facilitate a receiving PHY-layer in establishing a complementary delay. For example, the transmitting PHY-layer includes some indication of how much variable delay or Lateness was incurred waiting for a transmission to be launched onto the coax. The receiving PHY-layer uses said indication to help calculate how much complementary delay it will establish in order to equalize to some constant delay. In another example, the transmitting PHY-layer includes some indication for each of one or more Ethernet Frames aggregated or encapsulated in a transmission. In another example, the transmitting PHY, based on whatever variable delay it has suffered, pre-calculates the corresponding complementary delay that will be required at the receiving PHY, and transmits some corresponding indication. This pre-calculation by the transmitting PHY has the benefit of simplifying the receiving PHY-layer by relieving it from having to calculate a complementary delay. Another benefit of this embodiment is that the value of the complementary delay is calculated by the same PHY in which the variable delay was incurred, using the same clock (e.g., the transmitter’s MPCP clock) to measure both, thereby alleviating any clocking errors that might otherwise occur if two different clocks with differing counts or drift-rates were used. Such indications could be included in the transmissions in various ways, such as inside the data payload itself, in a header field, in a tag, via framing bits, via pilot tones, via timestamp(s), via preamble, via side-channel, via control- channel, via out-of-band channel, or any similar signaling mechanism known to those skilled in the art. [0055] The US path, or the DS path, may comprise multiple segments, such as digital fiber, analog fiber, coax, or some other transmission medium, or various combinations of one or more of these, each with a transmitting PHY-layer and a receiving PHY-layer, often resulting in a plurality of PHY-layers for each path. Someone skilled in the art will recognize that variable delays may be suffered in one or more PHY-layers, and that complementary delays may be established or distributed in one or more PHY-layers, and that the presently- described invention teaches how to accommodate such embodiments, resulting in constant delay for any of: the US path, the DS path, or the RTT. [0056] The presently-described invention teaches about complementary delays

established in at least one other PHY-layer. Someone skilled in the art will recognize that these complementary delays, as described herein, can be equivalently established outside of a PHY-layer. For example, with an OLT- OCU-CNU arrangement, such as depicted in Fig. 1, the CNU’s transmitting PHY-layer might suffer some variable delay, and the receiving PHY-layer on the OCU could establish the complementary delay before relaying the receptions onto digital fiber destined for the OLT. It would be possible to equivalently establish the complementary delay in some other layer or sub- layer in the OCU before the reception gets relayed onto digital fiber. An OCU, for example, might include some additional processing of US traffic that occurs after its coax PHY-layer receives a CNU’s transmission, but before the processed traffic is relayed onto digital fiber destined for the OLT. In this example, it would be possible for these additional processing steps to establish the complementary delay, as described herein. Someone skilled in the art will recognize that such examples are equivalent embodiments of this disclosure. [0057] While various embodiments of the disclosed method and apparatus have been described above, it should be understood that they have been presented by way of example only, and should not limit the claimed invention. Likewise, the various diagrams may depict an example architectural or other configuration for the disclosed method and apparatus. This is done to aid in understanding the features and functionality that can be included in the disclosed method and apparatus. The claimed invention is not restricted to the illustrated example architectures or configurations, rather the desired features can be implemented using a variety of alternative architectures and configurations. Indeed, it will be apparent to one of skill in the art how alternative functional, logical or physical partitioning and configurations can be implemented to implement the desired features of the disclosed method and apparatus. Also, a multitude of different constituent module names other than those depicted herein can be applied to the various partitions. Additionally, with regard to flow diagrams, operational descriptions and method claims, the order in which the steps are presented herein shall not mandate that various embodiments be implemented to perform the recited functionality in the same order unless the context dictates otherwise. [0058] Although the disclosed method and apparatus is described above in terms of various exemplary embodiments and implementations, it should be understood that the various features, aspects and functionality described in one or more of the individual embodiments are not limited in their applicability to the particular embodiment with which they are described. Thus, the breadth and scope of the claimed invention should not be limited by any of the above- described exemplary embodiments. [0059] Terms and phrases used in this document, and variations thereof, unless

otherwise expressly stated, should be construed as open ended as opposed to limiting. As examples of the foregoing: the term“including” should be read as meaning“including, without limitation” or the like; the term“example” is used to provide exemplary instances of the item in discussion, not an exhaustive or limiting list thereof; the terms“a” or“an” should be read as meaning“at least one,”“one or more” or the like; and adjectives such as “conventional,”“traditional,”“normal,”“standard,”“known” and terms of similar meaning should not be construed as limiting the item described to a given time period or to an item available as of a given time, but instead should be read to encompass conventional, traditional, normal, or standard technologies that may be available or known now or at any time in the future. Likewise, where this document refers to technologies that would be apparent or known to one of ordinary skill in the art, such technologies encompass those apparent or known to the skilled artisan now or at any time in the future. [0060] A group of items linked with the conjunction“and” should not be read as requiring that each and every one of those items be present in the grouping, but rather should be read as“and/or” unless expressly stated otherwise.

Similarly, a group of items linked with the conjunction“or” should not be read as requiring mutual exclusivity among that group, but rather should also be read as“and/or” unless expressly stated otherwise. Furthermore, although items, elements or components of the disclosed method and apparatus may be described or claimed in the singular, the plural is contemplated to be within the scope thereof unless limitation to the singular is explicitly stated. [0061] The presence of broadening words and phrases such as“one or more,”“at least,”“but not limited to” or other like phrases in some instances shall not be read to mean that the narrower case is intended or required in instances where such broadening phrases may be absent. The use of the term“module” does not imply that the components or functionality described or claimed as part of the module are all configured in a common package. Indeed, any or all of the various components of a module, whether control logic or other components, can be combined in a single package or separately maintained and can further be distributed in multiple groupings or packages or across multiple locations. [0062] Additionally, the various embodiments set forth herein are described in terms of exemplary block diagrams, flow charts and other illustrations. As will become apparent to one of ordinary skill in the art after reading this document, the illustrated embodiments and their various alternatives can be implemented without confinement to the illustrated examples. For example, block diagrams and their accompanying description should not be construed as mandating a particular architecture or configuration.

Claims

CLAIMS 1. A method for implementing Time-Division Duplex (TDD) in an EPoC Ethernet Passive Optical Network (EPON) Protocol over Coax network, the method comprising the steps of:

a) incurring variable delays (100) in at least one first PHY-layer; and b) establishing Complementary Delays (102) in at least one second PHY-layer.

2. A method for implementing Time-Division Duplex (TDD) in an EPoC Ethernet Passive Optical Network (EPON) Protocol over Coax network, the method comprising the steps of:

a) incurring variable delays (100) in at least one PHY-layer; and

b) establishing Complementary Delays (102) in at least one processing layer.

3. A system for implementing Time-Division Duplex (TDD) in an EPoC Ethernet Passive Optical Network (EPON) Protocol over Coax network comprising: means for incurring variable delays (100) in at least one first PHY-layer; and means for establishing Complementary Delays (102) in at least one second PHY- layer.

4. A system for implementing Time-Division Duplex (TDD) in an EPoC Ethernet Passive Optical Network (EPON) Protocol over Coax network comprising:

means for incurring variable delays (100) in at least one PHY-layer; and

means for establishing Complementary Delays (102) in at least one processing layer.

5. A non-transitory computer-executable storage medium comprising program instructions which are computer-executable to implement Time-Division Duplex (TDD) in an EPoC Ethernet Passive Optical Network (EPON) Protocol over Coax network, comprising: program instructions that cause variable delays (100) to be incurred in at least one first PHY-layer; and

program instructions that cause Complementary Delays (102) to be established in at least one second PHY-layer.

6. A non-transitory computer-executable storage medium comprising program instructions which are computer-executable to implement Time-Division Duplex (TDD) in an EPoC Ethernet Passive Optical Network (EPON) Protocol over Coax network, comprising: program instructions that cause variable delays (100) to be incurred in at least one PHY-layer; and

program instructions that cause Complementary Delays (102) to be established in at least one processing layer.

7. The method, system or storage medium according to any one of claims 1-6, wherein the step of, means or program instructions, wherein the at least one first and at least one second PHY-layers comprise an upstream path and/or wherein the at least one first and at least one second PHY-layers comprise a downstream path and/or further comprising transmitting onto or receiving from a coax or a hybrid fiber and coaxial system (HFC) by the at least one first and at least one second PHY-layers and/or comprising at least one transmitting PHY-layer (114) and at least one receiving PHY-layer (116) and/or wherein the at least one PHY-layer for incurring the variable delays (100) and the at least one PHY-layer for establishing the Complimentary delays (102) comprise a constant delay and/or wherein the at least one PHY-layer for incurring the variable delays (100) and the at least one PHY- layer for establishing the Complimentary delays (102) comprise block-processing (108, 133) and complementary shift-registers (118, 120) and/or wherein the at least one PHY-layer for incurring the variable delays (100) comprise waiting for an appropriate period within an upstream phase of a TDD Cycle to become available for transmission and/or wherein the at least one PHY-layer for incurring the variable delays (100) comprise waiting for an appropriate period within a downstream phase of a TDD Cycle to become available for transmission and/or wherein the at least one PHY-layer for incurring the variable delays (100) comprise waiting for a sufficient amount of a transmission to be received for rate- adaptation and/or wherein the at least one PHY-layer for incurring the variable delays (100) comprise indicating or an indicator for information to facilitate at least one other PHY-layer in establishing the Complementary Delays (102) and/or wherein the at least one PHY-layer for establishing the Complimentary Delays (102) comprise substantially equivalent processing by at least one other layer.

8. The method, system or program instructions of claim 7 further comprising traffic shaping or limiting traffic durations by a scheduler to reduce the variable delays (100) for waiting for a sufficient amount of a transmission to be received before rate-adaptation.

9. The method, system or program instructions of claim 7 wherein indicating or an indicator comprises providing how much variable delay (100) or Lateness was incurred waiting for a transmission to be launched.

10. The method, system or program instructions of claim 7 wherein the step of indicating or indicator comprises establishing how much Complementary Delay (102) is needed based on the variable delays incurred waiting for a transmission to be launched and/or wherein indicating or the indicator comprises including information for each of one or more Ethernet Frames in the transmission and/or wherein indicating or indicator comprises indicating in a transmission via a member from the group consisting of inside the data payload itself, in a header field, in a tag, framing bits, pilot tones, timestamps, preamble, side- channel, control-channel, and out-of-band channel.