| Figure 1–1 | Hybrid Fiber–Coax Architecture. |
|---|
| Figure 1–2 | Fiber–to–the–Home (FTTH) deployment scenarios. |
|---|
| Figure 2–1 | Light propagation in fiber |
|---|
| Figure 2–2 | Comparison of Multi–mode and single–mode fiber |
|---|
| Figure 2–3 | 8x8 couplers created from multiple 2x2 couplers. |
|---|
| Figure 2–4 | PON topologies. |
|---|
| Figure 2–5 | PON using a single fiber. |
|---|
| Figure 2–6 | Illustration of near–far problem in a TDM PON: a snapshot of power levels received from four ONUs. |
|---|
| Figure 3–1 | Comparison of Ethernet framing overhead and ATM cell tax. |
|---|
| Figure 4–1 | Relationship of IEEE 802.3 layering model to Open Systems Interconnection reference model. |
|---|
| Figure 5–1 | Downstream transmission in EPON. |
|---|
| Figure 5–2 | Upstream transmission in EPON. |
|---|
| Figure 5–3 | Processes and agents involved in bandwidth assignment. |
|---|
| Figure 5–4 | Sequential and pipelined timeslot assignments. |
|---|
| Figure 5–5 | In "Just–in–time" granting schemes, data collisions are possible due to delayed GATE message. |
|---|
| Figure 5–6 | Decoupled GATE arrival time and timeslot start time. |
|---|
| Figure 5–7 | Processes and agents involved in auto–discovery. |
|---|
| Figure 5–8 | Relationship of discovery slot and discovery window. |
|---|
| Figure 5–9 | Applying random delay during discovery process to avoid persistent collisions. |
|---|
| Figure 5–10 | Discovery attempt with and without REGISTER_REQ collision. |
|---|
| Figure 5–11 | Round–trip time measurement. |
|---|
| Figure 5–12 | Pre–calculation of arrival time when the OLT and ONUs use different GATE timestamp reference points. |
|---|
| Figure 5–13 | Pre–calculation of arrival time when the OLT and ONUs use different REGISTER_REQ timestamp reference points. |
|---|
| Figure 6–1 | Point–to–point virtual topology emulation. |
|---|
| Figure 6–2 | Bridging between ONU 1 and ONU 2 using the point–to–point emulation. |
|---|
| Figure 6–3 | Shared–medium emulation. |
|---|
| Figure 6–4 | Combined point–to–point and shared–medium emulation mode. |
|---|
| Figure 6–5 | Format of frame preamble in EPON. |
|---|
| Figure 6–6 | Position of SLD field depending on odd/even byte alignment. |
|---|
| Figure 6–7 | Shift register generating CRC–8. |
|---|
| Figure 6–8 | Function calculating CRC8 value. |
|---|
| Figure 7–1 | Timing diagram of Data Detector function. |
|---|
| Figure 7–2 | Illustration of partial laser shut down during ONUs transmission. |
|---|
| Figure 7–3 | Data Detector state diagram (reprinted from IEEE Std802.3ah with permission from IEEE).. |
|---|
| Figure 8–1 | Format of generic MPCP frame. |
|---|
| Figure 8–2 | Format of REPORT frame. |
|---|
| Figure 8–3 | Example of queues composition and reported values. |
|---|
| Figure 8–4 | REPORT frame formats with different numbers of reported queues and queue sets. |
|---|
| Figure 8–5 | Format of GATE message: (a) discovery GATE and (b) normal GATE. |
|---|
| Figure 8–6 | Grant structure. |
|---|
| Figure 8–7 | Format of REGISTER_REQ frame. |
|---|
| Figure 8–8 | Format of REGISTER frame. |
|---|
| Figure 8–9 | Format of REGISTER_ACK frame. |
|---|
| Figure 8–10 | (a) OLT Control Parser state diagram and (b) PARSE TIMESTAMP state in ONU Control Parser (reprinted from IEEE Std802.3ah with permission from IEEE). |
|---|
| Figure 8–11 | Accumulation of delay variabilities in MAC and PHY sub–layers. |
|---|
| Figure 8–12 | ONU Control Multiplexer state diagram (reprinted from IEEE Std802.3ah with permission from IEEE). |
|---|
| Figure 8–13 | Relationship of multiple MPCP instances and Multi–Point Transmission Control. |
|---|
| Figure 8–14 | Multi–Point Transmission Control state diagram (reprinted from IEEE Std802.3ah with permission from IEEE). |
|---|
| Figure 8–15 | OLT Control Multiplexer state diagram (reprinted from IEEE Std802.3ah with permission from IEEE). |
|---|
| Figure 8–16 | Gate Generation state diagram (reprinted from IEEE Std802.3ah with permission from IEEE). |
|---|
| Figure 8–17 | ONU Gate Reception state diagram (reprinted from IEEE Std802.3ah with permission from IEEE). |
|---|
| Figure 8–18 | ONU Gate Activation state diagram (reprinted from IEEE Std802.3ah with permission from IEEE). |
|---|
| Figure 8–19 | Examples of (a) back–to–back grant and (b) hidden grant. |
|---|
| Figure 8–20 | REPORT Generation state diagram (reprinted from IEEE Std802.3ah with permission from IEEE). |
|---|
| Figure 8–21 | OLT Report Reception state diagram (reprinted from IEEE Std802.3ah with permission from IEEE). |
|---|
| Figure 8–22 | Auto–discovery message exchange. |
|---|
| Figure 8–23 | OLT Discovery Gate Generation (reprinted from IEEE Std802.3ah with permission from IEEE). |
|---|
| Figure 8–24 | OLT Request Reception state diagram (reprinted from IEEE Std802.3ah with permission from IEEE). |
|---|
| Figure 8–25 | OLT Register Generation state diagram (reprinted from IEEE Std802.3ah with permission from IEEE). |
|---|
| Figure 8–26 | OLT Final Registration state diagram (reprinted from IEEE Std802.3ah with permission from IEEE). |
|---|
| Figure 8–27 | ONU Discovery state diagram (reprinted from IEEE Std802.3ah with permission from IEEE). |
|---|
| Figure 9–1 | Structure of FEC–coded frame. |
|---|
| Figure 9–2 | Hamming distance between /T_FEC_O/ and /T_FEC_E/ delimiters. |
|---|
| Figure 9–3 | Example of bit stream matching both /T_FEC_O/ and /T_FEC_E/ delimiters. |
|---|
| Figure 9–4 | FEC encoder block diagram. |
|---|
| Figure 9–5 | FEC decoder block diagram. |
|---|
| Figure 10–1 | Implementation of Counter Mode. |
|---|
| Figure 10–2 | Relation of cipher counter to MPCP counter. |
|---|
| Figure 10–3 | Relationship between cipher counter sequence, block counter sequence, and cipher input values. |
|---|
| Figure 10–4 | 128–bit cipher input block is produced by concatenation of cipher counter and block counter. |
|---|
| Figure 10–5 | Alignment of cipher counter with an upstream burst. |
|---|
| Figure 11–1 | Redundant PON topologies. |
|---|
| Figure 12–1 | Guard band components. |
|---|
| Figure 12–2 | Structure of a FEC–coded frame. |
|---|
| Figure 13–1 | Graphical representation of pair–wise collision probability. |
|---|
| Figure 13–2 | Average number of successful registrations in one attempt. |
|---|
| Figure 13–3 | Efficiency of the discovery slot size. |
|---|
| Figure 13–4 | Optimal discovery slot size as a function of the number of contending ONUs. |
|---|
| Figure 14–1 | Simulation model of EPON segment with SSA. |
|---|
| Figure 14–2 | Average packet delay as a function of ONU's offered load. |
|---|
| Figure 14–3 | Average queue size as a function of ONU's offered load. |
|---|
| Figure 14–4 | Packet loss ratio as a function of ONU's offered load. |
|---|
| Figure 14–5 | Illustration of packet scheduling. |
|---|
| Figure 14–6 | Average link utilization (FIFO vs. First Fit). |
|---|
| Figure 14–7 | Algorithm for connection–based packet reordering. |
|---|
| Figure 15–1 | Operation of DBA agent at the OLT. |
|---|
| Figure 15–2 | Time diagram of the limited service in IPACT algorithm. |
|---|
| Figure 15–3 | Components of packet delay at the ONU. |
|---|
| Figure 15–4 | Average packet delay. |
|---|
| Figure 15–5 | Average cycle times for various service disciplines. |
|---|
| Figure 15–6 | Average packet delay as a function of effective network load and ONU offered load. |
|---|
| Figure 15–7 | Packet–loss ratio as a function of effective network load and ONU offered load. |
|---|
| Figure 16–1 | Intra–ONU and inter–ONU scheduling. |
|---|
| Figure 16–2 | Packet delay for limited/priority service. |
|---|
| Figure 16–3 | Tandem queue at an ONU. |
|---|
| Figure 16–4 | Packet delay for limited/tandem scheme. |
|---|
| Figure 16–5 | Packet delay for CBR–credit/priority scheme. |
|---|
| Figure 16–6 | Calculation of the credit interval. |
|---|
| Figure 16–7 | Histogram of unused slot remainder and most frequent preemption combinations. |
|---|
| Figure 16–8 | Average slot remainder. |
|---|
| Figure 16–9 | Average cycle time: (a) absolute value and (b) normalized to limited/priority service. |
|---|
| Figure 16–10 | Bandwidth utilization. |
|---|
| Figure 17–1 | Direct (single–level) scheduling in EPON. |
|---|
| Figure 17–2 | Hierarchical scheduling in EPON. |
|---|
| Figure 17–3 | (a) Sibling–fair scheduling based on cumulative group weight Σφ
(b) Sibling–fair scheduling based on cumulative group work Σq;
(c) Cousin–fair scheduling. |
|---|
| Figure 18–1 | Construction of service envelopes. |
|---|
| Figure 18–2 | PROCESS_GRANT procedure calculates start times for all children of node i. |
|---|
| Figure 18–3 | Granting phase for the scheduling hierarchy shown in Figure $X$(c). |
|---|
| Figure 18–4 | Approximation of SE function in Min–Error approach. |
|---|
| Figure 18–5 | Approximating the service envelope E using the Min–Points approach. |
|---|
| Figure 18–6 | The CONSTRUCT_APPROX_LINEAR algorithm |
|---|
| Figure 18–7 | TEST_TANGENT(point, n) function |
|---|
| Figure 18–8 | The CONSTRUCT_APPROX_BINARY algorithm. |
|---|
| Figure 18–9 | MIN_POINT_APPROX() procedure. |
|---|
| Figure 18–10 | Approximation error for Min–Error and Min–Points schemes. |
|---|
| Figure 18–11 | Collecting all requests before scheduling grants. |
|---|
| Figure 18–12 | Scheduling two groups of nodes separately. |
|---|
| Figure 18–13 | Size of unused remainder in an ONU. |
|---|
| Figure 18–14 | Experimental EPON model. |
|---|
| Figure 18–15 | Throughput of test queues under different ambient loads. |
|---|
| Figure 18–16 | Ratio of throughputs for queues located in different ONUs. |
|---|
| Figure 18–17 | Maximum difference in bandwidth allocated to queue #4 in ONUs A and B. |
|---|
| Figure A–1 | Scaling behavior of LRD and SRD traffic flows. |
|---|
| Figure B–1 | Variance–time log–log plot. |
|---|
| Figure B–2 | Aggregation (multiplexing) of multiple substreams produces self–similar traffic. |
|---|