Fiber Wireless Access Networks

FIBER WIRELESS ACCESS NETWORKS

1. INTRODUCTION

The ultimate goal of the Internet and communicationnetworks in general is to provide access toinformation when we need it, where we need it,and in whatever format we need it in. To achievethis goal, wireless and optical technologies play akey role. Wireless and optical access networkscan be thought of as complementary. Opticalfiber does not go everywhere, but where it doesgo, it provides a huge amount of available bandwidth.Wireless access networks, on the otherhand, potentially go almost everywhere, but providea highly bandwidth-constrained transmissionchannel susceptible to a variety of impairment.Clearly, as providers need to satisfy users withcontinuously increasing bandwidth demands,future broadband access networks must leverageon both technologies and converge them seamlessly,giving rise to fiber-wireless (FiWi) accessnetworks.

RoF networks are attractive since theyprovide transparency against modulation techniques and are able to support various digitalformats and wireless standards in a cost-effectivemanner, for example, wideband code-divisionmultiple access (WCDMA), IEEE 802.11 wirelesslocal area network (WLAN), personalhandy phone system (PHS), and Global Systemfor Mobile Communications (GSM). To realizefuture multiservice access networks, theseamless integration of RoF systems with existingand emerging optical access networks isimportant, such as FTTX and wavelength-divisionmultiplexing (WDM) PON networks. RoFnetworks are also well suited to avoid frequenthandovers of fast-moving users in cellular networks.An interesting approach to avoid handoversfor train passengers is the use of an opticalfiber WDM ring-based RoF network installedalong the rail tracks in combination with themoving cell concept, as recently proposed .The concept of moving cells enables a cell patternand a train to move along on the sameradio frequency during the whole connection ina synchronous fashion without requiring handovers.

2.WIRELESS ACCESS NETWORKS

2.1 WIRELESS MESH NETWORKS

Recent advances in wireless communicationstechnology have led to significant innovations that have enabled cost-effective and flexiblewireless Internet access, and provided incentivesfor building efficient multihop wireless networks.A wireless ad hoc network precludes the use of awired infrastructure and allows hosts to communicate either directly or indirectly over radiochannels without requiring any prior deploymentof network infrastructure.Wireless mesh networks (WMNs), on theother hand, are networks employing multihopcommunications to forward traffic en route toand from wired Internet entry point. In contrastto conventional WLANs and mobile adhoc networks (MANETs), WMNs promisegreater flexibility, increased reliability, andimproved performance. WMNs can be categorizedinto infrastructure, client, and hybridWMNs (Fig. 1). A router in an infrastructureWMN has no mobility and performs more functionsthan a normal wireless router. Among others,a router performs mesh functions (routingand configuration) and acts as a gateway. In aclient WMN, clients perform mesh and gatewayfunctions themselves. Efficient routing protocolsprovide paths through the wireless mesh and react to dynamic changes in the topology, somesh nodes can communicate with each othereven if they are not in direct wireless range.Intermediate nodes on the path forward packetsto the final destination. Due to the similaritiesbetween WMNs and MANETs, WMNs canapply ad hoc routing protocols (e.g., ad hoc on demanddistance vector [AODV] and dynamicsource routing [DSR], among others).

Figure2.1.1wireless mesh networks a)infrastructure b)client c)hybrid

2.2 ENABLING TECHNOLOGIES

New technologies and protocols in the physical(PHY) layer, medium access control (MAC) protocols,and routing protocols are required tooptimize the performance of WMNs. In the PHYlayer, smart antenna, multi-input multi-output(MIMO), ultra wideband (UWB), and multichannelinterface systems are being explored toenhance network capacity and further enablewireless gigabit transmission. Recently, gigabittransmission resulting from a combination ofMIMO and orthogonal frequency-division multiplexing(OFDM) has been demonstrated. MACprotocols based on distributed time-division multipleaccess (TDMA) and CDMA are expectedto improve the bandwidth efficiency of carriersense multiple access with collision avoidance (CSMA/CA) protocols.

Currently, IEEE 802.11 a/b/g (WiFi) technologiesare widely exploited in commercial, products and academic research of WMNs dueto their low cost, technological maturity, andhigh product penetration . However, sincethese protocols were originally designed forWLANs, they clearly are not optimized forWMNs. Proprietary wireless technologies andWiMAX have been proposed. Unlike WiFi,IEEE 802.16 allows for point-to-multipointwireless connections with a transmission rateof 75 Mb/s and can be used for longer distances.

Additionally, orthogonal frequency-divisionmultiple access (OFDMA) and smart antennatechnologies extend the scalability of WiMAX.These technologies are exploited to enhance thecapacity, reliability, and mobility of WMNs.

Ultra-high-bandwidth standards such as IEEE802.16m, which aims to provide 1 Gb/s and 100Mb/s shared bandwidth, can be employed to furtherenhance the bandwidth and mobility ofWMNs. Since packets are routed among meshrouters in the presence of interference, shadowing,and fading, a cross-layer design is requiredto optimize the routing in WMNs. For instance,DSR uses link quality source routing (LQSR) toselect a routing path according to link qualitymetrics. LQSR includes three performance metrics:per-hop packet pair, per-hop round-triptime (RTT), and expected transmission count(ETX). ETX shows the best performance in networks ,with fixed nodes, while minimum hopcount shows good performance in networks withmobile nodes.

2.3 FUTURE DEVELOPMENTS

Given the increased demand for mesh networks, a task group was formed in 2004 to define the Extended Service Set (ESS) mesh networking standard; its goal is the development of a flexible and extensible standard for WMNs based on IEEE 802.11. The IEEE 802.11s amendment can be split up into four major parts: multihop routing, MAC enhancements, security, and general topics. It also defines a new mesh data frame format that can be used for transmitting data within the WMN. Traffic in mesh networks is predominantly forwarded to and from wire line gateway nodes forming a logical tree structure. The 802.11s defines a default mandatory routing protocol (Hybrid Wireless Mesh Protocol [HWMP]) that uses hierarchical routing to exploit this tree-like logical structure and on demand routing protocols to address mobility; the on demand routing protocol is based on AODV, which uses a simple hop count routing metric. Alternatively, the standard allows vendors to operate using alternate protocols, one of which is described in the draft (Radio Aware Optimized Link State Routing [RA-OLSR]). RA-OLSR uses multipoint relays, a subset of nodes that flood a radio-aware link metric, thereby reducing control overhead on the routing protocol. Other interesting developments are concerned with the integration of different access technologies; for instance, the authors of presented an approach for integrating WiMAX and WiFi technologies, and discussed several issuespertaining to protocol adaptation and QoS support.

3.OPTICAL ACCESS NETWORKS

Optical fiber provides unprecedented bandwidth potential far in excess of the wireless and any other known transmission medium. A single strand of fiber offers a total bandwidth of 25,000 GHz. More important, optical networks lendthemselves well to offloading electronic equipmentby means of optical bypassing as well asreducing their complexity, footprint, and powerconsumption significantly while providing opticaltransparency against modulation format, bit rate, and protocol.

3.1 FTTX NETWORKS

FTTX networks are poised to become the nextmajor success story for optical fiber communications.Not only must future FTTX access networksunleash the economic potential andsocietal benefit by opening up the first/last milebandwidth bottleneck between bandwidth-hungryend users and high-speed backbone networks,but also enable the support of a widerange of new and emerging services and applications,such as triple play, video on demand,point-to-point (P2P) audio/video file sharing andstreaming, multichannel HDTV, multimedia/multiparty online gaming, and telecommuting.Due to their longevity, low attenuation, andhuge bandwidth, PONs are widely deployed torealize cost-effective FTTX access networks

3.2 PONS

Typically, PONs are time-division multiplexing(TDM) single-channel systems, where the fiberinfrastructure carries a single upstream wavelengthchannel (from subscribers to a centraloffice) and a single downstream wavelengthchannel (from a central office to subscribers).IEEE 802.3ah Ethernet PON (EPON) with asymmetric line rate of 1.25 Gb/s, and InternationalTelecommunication Union — TelecommunicationStandardization Sector (ITU-T) G.984 Gigabit PON (GPON) with an upstreamline rate of 1.244 Gb/s and a downstream linerate of 2.488 Gb/s represent current state-of-threatcommercially available and widely deployedTDM PON access networks, but standardizationefforts have already been initiated in the IEEE802.3av Task Force to specify 10 Gb/s EPON.GPON offers strong operation, administration,maintenance, and provisioning (OAMP) capabilities,and provides security at the protocol levelfor downstream traffic by means of encryptionusing Advanced Encryption Standards. Furthermore,GPON efficiently supports traffic mixesconsisting not only of asynchronous transfermode (ATM) cells but also TDM (voice) andvariable-size packets by using the GPON encapsulationmethod (GEM). EPON aims at convergingthe low-cost equipment and simplicity ofEthernet and the low-cost infrastructure ofPONs. Security and OAMP are not specified inthe EPON standard IEEE 802.3ah, but may beimplemented using the data over cable serviceinterface specification (DOCSIS) OAMP servicelayer on top of the MAC and PHY layers ofEPON. Given the fact that 95 percent of LANsuse Ethernet, and most applications and services(e.g., video) are moving toward Ethernet, in conjunctionwith Ethernet’s low cost and simplicity,EPON is expected to increasingly become thenorm.

Both GPON and EPON are commonly perceivedto carry a single wavelength channel ineach direction. The majority of real-world PONdeployments, however, use an additional downstreamwavelength channel for video distributionaccording to the wavelength allocation in ITU-TRecommendation G.983.3, which specifies a so calledenhancement band from 1539 to 1565 nmplus an L-band reserved for future use. Theenhancement band and L-band can be used toenable additional services such as overlay ofmultiple PONs on a single fiber infrastructure oroptical time domain reflectometry (OTDR) fortesting and troubleshooting.

3.3 FUTURE DEVELOPMENTS

Adding the wavelength dimension to conventionalTDM PONs leads to WDM PONs, which haveseveral advantages. Among others, the wavelengthdimension may be exploited to:

• Increase network capacity

• Improve network scalability by accommodatingmore end users

• Separate services

• Separate service providers

An interesting approach to increasing splitratio (i.e., number of subscribers) and range isthe so-called long-reach PON (LR-PON),which is currently receiving considerableattention from network operators in anattempt to optically bypass central offices andconsolidate optical metro and access networks,resulting in major cost savings and simplifiednetwork operation. LR-PONs can alsobe interesting for new operators wishing onlyto connect the major geographically distributedbusiness clients.

Most of the reported studies on advancedPON architectures have considered standalonePON access networks, with a particular focus onthe design of dynamic bandwidth allocation(DBA) algorithms for quality of service (QoS)support and QoS protection by means of admission control.

4.FIWI NETWORKS

4.1 ENABLING TECHNOLOGIES

Currently, there are two technologies used toimplement fiber-wireless (FiWi) networks:

• Free space optical (FSO), also known as optical wireless (OW)

• Radio over fiber (RoF)

4.1.1 Free space optical (FSO)

FSO is a type of direct line-of-sight (LOS)optical communications that provides point-to pointconnections by modulating visible orinfrared (IR) beams . It offers high bandwidthand reliable communications over shortdistances. The transmission carrier is generatedby deploying either a high-power light emittingdiode (LED) or a laser diode, while thereceiver may deploy a simple photo detector.Current FSO systems operate in full-duplexmode at a transmission rate ranging from 100Mb/s to 2.5 Gb/s, depending largely on weatherconditions. Given a clear LOS betweensource and destination and enough transmitterpower, FSO communications can work overdistances of several kilometres. At both sourceand destination, optical fiber may be used tobuild high-speed LANs, such as Gigabit Ethernet(GbE).

4.1.2 Radio over fiber (RoF)

RoF, on the other hand, allows an analogoptical link to transmit a modulated radio frequency(RF) signal. There are different techniquesavailable to realize RoF networks.Typically, an RoF transmitter deploys a Mach-Zehnder intensity (MZI) modulator in conjunctionwith an oscillator that generates therequired optical carrier frequency, followed byan Erbium doped fiber amplifier (EDFA) inorder to increase the transmission range. RoFnetworks provide both P2P and point-to-multipointconnections. Recently, a full-duplex RoFsystem providing 2.5 Gb/s data transmission over40 km with less than 2 dB power attenuation wassuccessfully demonstrated using the millimeterwave band . There are many cost-efficientoptical approaches to mixing and up convertingmillimetre wave signals.

Table 1 summarizes and compares the salientfeatures of both enabling technologies of FiWinetworks.

Table4.1.1Comparison between wireless segments of FSO and RoF

4.2 ARCHITECTURES

This section deals with the available architecturesfor enabling FiWi integration. For instance, theintegration of EPON and WiMAX access networkscan be done in several ways; according to, the following four architectures can beused.

4.2.1 Independent Architecture

In this approachWiMAX base stations serving mobile clientnodes are attached to an optical network unit(ONU) just like any other wired subscribernode, whereby an ONU denotes the EPON customer premises equipment. WiMAX and EPONnetworks are connected via a common standardizedinterface (e.g., Ethernet) and operate independentof each other.

4.2.2 Hybrid Architecture

This approach introducesan ONU-base station (ONU-BS) that integratesthe EPON ONU and WiMAX BS in bothhardware and software. The integrated ONU-BScontrols the dynamic bandwidth allocation ofboth the ONU and BS.

4.2.3 Unified Connection-Oriented Architecture

Similar to the hybrid architecture, thisapproach deploys an integrated ONU-BS. Butinstead of carrying Ethernet frames, WiMAXMAC protocol data units (PDUs) containingmultiple encapsulated Ethernet frames areused. By carrying WiMAX MAC PDUs, the unifiedarchitecture can be run like a WiMAX networkwith the ability to grant bandwidth finelyusing WiMAX’s connection-oriented rather thanEPON’s queue-oriented bandwidth allocation.

4.2.4 Microwave-over-Fiber Architecture

In thisapproach the WiMAX signal is modulated on awireless carrier frequency, and is then multiplexedand modulated together with the basebandEPON signal onto a common opticalfrequency (wavelength) at the ONU-BS. Thecentral node consists of a conventional EPONoptical line terminal (OLT) and a centralWiMAX BS, called a macro-BS. The OLT processesthe baseband EPON signal, while themacro-BS processes data packets originatingfrom multiple WiMAX BS units.

4.2.5 Other architectures

Besides the aforementioned generic integrationapproaches of EPON and WiMAX networks,several other FiWi architectures based onWiFi technology have been studied, as describedin the following

Figure4. 2.3.1Optical unidirectional fiber ring interconnecting WiFi-based wireless access points.

The network shown in Fig. interconnectsthe central office (CO) with multiple WiFi-basedwireless access points (WAPs) by means of anoptical unidirectional fiber ring . The CO isresponsible for managing the transmission ofinformation between mobile client nodes(MCNs) and their associated WAPs as well as acting as a gateway to other networks. EachWAP provides wireless access to MCNs withinits range. All MCNs take part in the topologydiscovery, whereby each MCN periodically sendsthe information about the beacon power receivedfrom its neighbours to its associated WAP. Indoing so, WAPs are able to estimate the distancesbetween MCNs and compute routes. Multihoprelaying is used to extend the range. Toenhance the reliability of the wireless link, theCO sends information to two different WAPs(path diversity). The proposed implementationcan support advanced path diversity techniquesthat use a combination of transmission via severalWAPs and multihop relaying (e.g., cooperativediversity or multihop diversity).Consequently, the CO must be able to assignchannels quickly and efficiently by using one ormore wavelength channels on the fiber ring toaccommodate multiple services such as WLANand cellular radio network.

Figure4.2.3.2. Optical interconnected bidirectional fiber rings integrated with WiFi-based wireless access points.

Figure shows a two-level bidirectional path protectedring (BPR) architecture for denseWDM (DWDM)/subcarrier multiplexing (SCM)broadband FiWi networks . In this architecturethe CO interconnects remote nodes (RNs)via a dual-fiber ring. Each RN cascades WAPsthrough concentration nodes (CNs), where eachWAP offers services to MCNs. For protection,the CO is equipped with two sets of devices(normal and standby). Each RN consists of aprotection unit and a bidirectional wavelengthadd-drop multiplexer based on a multilayerdielectric interference filter. Each CN contains aprotection unit. The WAP comprises an opticaltransceiver, a protection unit, up/down RF converters, and a sleeve antenna. Each WAP provideschannel bandwidth of at least 5 MHz andcovers up to 16 MCNs by means of frequency divisionmultiplexing (FDM). Under normaloperating conditions, the CO transmits downstreamsignals in the counter-clockwise directionvia RNs and CNs to the WAPs. If a fiber cutoccurs between two RNs or between two CNs,their associated controllers detect the failure bymonitoring the received optical signal and thenswitch to the clockwise protection ring. If a failurehappens at a WAP, the retransmitted signalsare protection switched through other opticalpaths by throwing an optical switch inside theaffected WAP. This architecture provides highreliability, flexibility, capacity, and self-healingproperties.

Figure4.2.3.3Optical hybrid star-ring network integrated with WiFi-based wireless access points.