Optical Access Network

This section describes optical access network architectures that are based on Passive Optical Network (PON) [18-22]. The PON provides high bandwidths in access networks. Here we discuss the Ethernet PON (EPON) [20,23], ATM-based PON (APON) [24], Broadband PON (BPON) and Generalize Framing Procedure (GFP) PON (GFP-PON [25]). We also mention wavelength-division multiplexing (WDM)- based PON. The access network is also represented as the “first-mile” network, connecting central offices (COs) to the subscribers.

Issues in Optical Access Architecture

The issues in optical access network are to develop high-capacity backbone networks in which backbone network operators provide the accessing of high-capacity OC-192 (10 Gbps) links (recently OC-768(40 Gbps). The recently developed access network technologies are Digital Subscriber Loop (DSL), which has the limitation to provide broadband services such as video-on-demand, interactive games and video conferencing to end users. The predominant broadband access solutions are the DSL and Community Antenna Television (cable TV)-based networks. However, both of these technologies have limitations because these were originally built for carrying voice and analog TV signals, but their carrying data are not optimal. DSL has another restriction that the distance of any DSL subscriber to a CO must be less than 18,000ft due to having signal distortions.

Cable television networks can also support Internet services by dedicating some Radio Frequency (RF) channels in a co-axial cable for data. Cable networks are mainly built for delivering broadcast services. At high load, the network’s performance is not used for satisfying end users.

The access networks bring fiber to the home with FTTx models - Fiber-to-the- Home(FTTH) [22],Fiber-to-the-Curb(FTTC) [26],Fiber-to-the-Building(FTTB) [26]. etc. These models offer the access bandwidths to end users. These technologies provide fiber directly to the home or very near to the home. The FTTx models are mainly based on the PON. The major developments on PON in recent years are EPON, APON, GFP- PON and WDM-PON which are discussed in this section.

Simple Fiber-Optic Access Network Architectures

Optical fiber transmits up to 50 km or beyond in the subscriber access network. Figure 9.11 shows an FTTH access network in which a logical method accessing optical fiber in the local access network uses a point-to-point (PtP) topology. Three configurations are - dedicated fiber set from the CO to each end-user subscriber for

bandwidth-intensive, integrated, voice, data and video services: (a) simple architecture having connector termination space curb switch connected to a Local Exchange (called as CO) with bidirectional fiber in which N subscribers are connected with a curb switch, (b) architecture having passive optical power splitter connected to the Local Exchange (called as CO) with bidirectional fiber in which N subscribers connect the curb switch at an average distance L kilometers from CO. A PtP structure has 2N transceivers and N x L total fiber length (assuming that a single fiber is used for bidirectional transmission).

Since the curb switch is expensive, the active curb-side switch can be replaced with an inexpensive passive optical split as shown in Figure 9.12b. PON problem [21] can reduce the number of optical transceivers, CO terminations and fiber deployment. A POM is a point-to-multipoint optical network with no active elements in the signal path from the source to the destination. The interior elements used in a POM are passive optical components, such as optical fiber, splices and splitters. An access network based on a single-fiber PON only requires N + 1 transceivers and L kilometers of fiber.

Components of PON Technologies

The key components of PON are optical splitter/couplers, burst mode switches, MAC, etc. These components have been discussed in detail in Chapters 3 and 6. In this section, we try to give an overview of these components and how these components have been related to PONs.

Optical Splitters/Couplers

A PON needs a passive device to divide an optical signal (power) coming from one fiber into several fibers and at the receiver to combine optical signals from multiple

FIGURE 9.12 FTTH deployment scenarios: (a) via curb switch and (b) passive optical switch [26].

fibers into one. This device is an optical coupler/splitter. Section 3.1 already discusses the different optical couplers in detail. An optical coupler is made by using two fibers fused together in which signal can coupler from one fiber to the other with the principle of evanescent wave principle. The N x N couplers are made by staggering multiple 2x2 couplers or by using planar waveguide technology. Couplers are characterized by the following parameters:

• 1. Splitting loss: Power level at a half power 2x2 fiber coupler, and this value is 3dB.
• 2. Insertion loss: Insertion loss arises due to imperfections of the coupler’s fabrication process and coupling loss from 0.1 to 1 dB.
• 3. Directivity: Some amount of input power coming out as a leakage from one input port to another input port. The directivity ~40-50dB.

The couplers made to have only one input to more outputs are referred to as a splitter. A coupler with more inputs and only one output is a combiner. Sometimes, 2x2 couplers are made highly asymmetric (with splitting ratios 5/95 or 10/90). This kind of coupler is used to branch off a small portion of the signal power for monitoring purposes. Such devices are called tap couplers.

PON Topologies

Since a CO has multiple subscribers, multipoint topologies are suitable for the access network, including tree, tree-and-branch, ring or bus (Figure 9.13a). The use of 1:2 optical tap couplers and 1:7V optical splitters, PONs are flexible in ring, bus and tree topologies. All transmissions in a PON are carried out between an Optical Line Terminal (OLT) [26] and Optical Network Units (ONUs) [26], as shown Figure 9.13.

The OLT stays in the CO connecting the optical access network to the MAN or WAN [26], which is a long-haul network. The ONU is located at the end-user location (FTTH and FTTB) or at the curb requiring FTTC architecture.

The advantages of using PONs in access networks are numerous:

• • PONs permit transmission for long distances (up to 20 km or beyond).
• • PONs make minimization of fiber deployment.
• • PONs support higher bandwidth due to the use of fiber.
• • PONs permit for video broadcasting either as Internet Protocol (IP) video or video.
• • PONs remove the necessity of installing active multiplexers.

Burst-Mode Transceivers

Having unequal distances between CO and ONUs, optical signal attenuation in the PON is not uniform for all ONUs. The ONU’s signal power is small at the OLT because of its transmission for longer distance. If the receiver is tuned to properly receive high-power signal from a close ONU, it receives a weak signal from a distant ONU. If the receiver is trained on a weak signal, it differentiates zeros or ones when receiving a strong signal.

Time slot is a period of time during which certain data are transmitted by specific regulations in the network. Time slot is not a periodic time allocation concept as in TDM here. Time slot is a “time slice” that is assigned to a node for transmission of its backlogged packet slot, i.e., it should operate in burst mode. A burst-mode receiver is required only in the OLT. The ONUs study a continuous bit stream (data or idle bits) sent by the OLT and do not need to readjust quickly.

Another approach has ONUs to adjust their transmitter powers where power levels were received by OLT from all ONUs. This method is critical particularly for transceiver design as it makes the ONU hardware more complex, requiring a special signaling protocol for feedback from the OLT to each ONU.

EPON Access Architecture

EPON transmits data traffic encapsulated in Ethernet frames (IEEE 802.3 standard using 8 bit/10 bit line coding), operating at standard Ethernet data rates. Ethernet is used due to low-cost line cards and widely used in LANs [20,23]. Since access networks are focused towards end users and LANs, high-speed Gigabit Ethernet deployment is used. The 10-Gigabit Ethernet products are easily available. Ethernet provides a much more efficient MAC protocol in comparison to ATM imposing a considerable overhead on variable-length IP packets.

Operation of EPON

In the downstream (OLT to ONUs), Ethernet frames are transmitted by the OLT pass via a 1 :N passive splitter and reach each ONU. Typical values of N are between 8 and 32. Packets are broadcast by the OLT and extracted by their destination ONU based on a Logical Link Identifier (LLID), which the ONU is assigned when it registers with the network. Figure 9.13b shows the downstream traffic in EPON [23].

FIGURE 9.13 (CONTINUED) (B) Downstream operation in EPON [26].

(Continued)

In the upstream, data frames originate from an ONU transmitting to the OLT and never send to any other ONU due to the directional properties of a passive optical combiner. The function of EPON is similar to that of a PtP architecture. However, in EPON, data frames from different ONUs transmitted simultaneously may make a collision. Thus, in the upstream direction a contention-based media-access mechanism is required (similar to CSMA/CD), and it is not easy to implement because ONUs are not able to detect a collision in the fiber from the combiner to the OLT due to directional properties of the combiner. An OLT can find a collision and inform ONUs by sending a jam signal; however, the relatively large propagation delay in a PON (where the typical distance from the OLT to ONU is 20km) greatly reduces the efficiency of such a scheme. Figure 9.13c illustrates the operation of upstream, time- shared, data flow' in an EPON [26].

FIGURE 9.13 (CONTINUED) (C) Upstream operation in EPON [26].

Each ONU is assigned a time slot in which all ONUs are synchronized to transmit each time slot capable of carrying several Ethernet frames. When its time slot come as per term wise, the ONU would burst all stored frames at full channel speed. If there are no frames in the buffer to fill the entire time slot, an idle pattern is transmitted. The time-slot assignment is a very crucial step. The possible time-slot allocation schemes could range from a static allocation (fixed time-division multiple access (TDMA)) to a dynamically adapting scheme based on instantaneous queue size in every ONU (statistical multiplexing scheme). In the dynamic scheme, the OLT can play the role of collecting the queue sizes from the ONUs and then issuing time slots. Although this approach leads to a significant signaling overhead between the OLT and ONUs, the centralized intelligence may lead to more efficient use of bandwidth. More advanced bandwidth-allocation schemes are also possible, including schemes utilizing notions of traffic priority [27].