UNIT V

HIGH CAPACITY NETWORKS

Upgrading Transmission CapacityThere are three approaches towards increasing the

transmission capacity on a link.1. The space division multiplexing ( SDM) approach :

Keep the bit rate the same but use more fibers.2. The TDM approach: increases the transmission

bit rate on the fiber.3. The WDM approach: keep the bit rate the same

but add more wavelengths, each operating at the original bit rate over the same fiber.

The SDM approachIt has two main drawbacks• It requires more fibers, which may not be

available along the desired route. The cost of laying new fiber varies widely.

• If there is space in existing conduits, fiber can be pulled through relatively inexpensively. If new conduits must be laid, the cost can be very expensive, even over short distances.

• If there are no fibers along the desired route ,the link has to be built from the scratch.

• The second drawback of SDM is that even if the fibers are available, a separate set of optical amplifiers or repeaters must be installed for each fiber, which becomes a significant expense over long distances.

• SDM approach can be used when fibers are available and distances are relatively short so that no repeaters or amplifiers are needed.

The TDM Approach• TDM approach requires the bit rate on the fiber

to be increased to 10Gb/s or beyond.• Two major system impairments that these

systems face are chromatic dispersion and polarization mode dispersion (PMD).

• Chromatic dispersion is not a problem if the link uses dispersion shifted fiber.

• With standard single mode fiber, the chromatic dispersion limit is about 60km at 10Gb/s and about 1000 km at 2.5Gb/s.

• With practical transmitters, the distances are even smaller.

• The 10Gb/s limit can be further increased in the presence of self phase modulation.

• Beyond these distances, the signal must be electronically regenerated, or some form of chromatic dispersion compensation must be employed.

• The distance limit due to PMD at 10Gb/s is 16 times less than that at 2.5Gb/s.

• The distance limit is about 25km at 10Gb/s.

• Electronic regeneration is required for longer distances.

• The PMD induced distance limit may be even lower because of additional PMD caused by splices, connectors, and other components along the transmission path.

• TDM is limited by the speeds that can be achieved by the electronics.

• Higher bit rates can be achieved using optical TDM.

• TDM is a competitive approach for short to medium distance links or for links operating over dispersion shifted fiber.

The WDM ApproachThe WDM approach is to keep the bit rate same,

say, at 2.5Gb/s, and add more wavelengths, each carrying data at this bit rate. The advantages of WDM over TDM are the following.

1. Because of the lower bit rates, the distance limit due to chromatic dispersion is much larger for WDM systems than for equivalent TDM systems.

• The transmission capacity can be increased in a modular manner by adding additional wavelengths as and when capacity increases are required, as opposed to a large up front installation in the case of TDM.

• WDM systems can be designed to be transparent systems. This allows different wavelengths to carry data at different bit rates and protocol formats.

• WDM may be preferred to TDM in designing more complicated networks. For example, if there is a network node at which most of the traffic is to be passed through and a small fraction is to be dropped and added, it may be more cost effective to use a WDM optical add/drop element than a full blown set of TDM terminals.

Disadvantages of WDM relative to TDM1. WDM systems are not suitable for deployment

over dispersion shifted fiber because of the limitation imposed by four wave mixing.

2. WDM systems require specially designed optical amplifiers that provide flat gain profiles.

3. WDM systems requires separate terminating equipment for each channel, including an expensive laser and a receiver. A TDM system on the other hand, requires one piece of terminating equipment

4. Transparent WDM systems offer less monitoring and network management capability than TDM systems because they are unaware of the actual format and data rate on the individual channels. Thus they cannot monitor parameters like bit rate or frame errors in the data.

Unidirectional versus Bidirectional Systems• A unidirectional WDM system uses two fibers,

one for each direction of traffic .• A bidirectional system, on the other hand

requires only one fiber, and typically uses half the wavelength for transmitting data in one direction and the other half for transmitting data in the opposite direction on the same fiber.

Comparison1. A unidirectional system is capable of handling

W full duplex channels over two fibers. A bidirectional system handles w/2 full duplex channels over one fiber. The bidirectional system, has half the total capacity, but allows a user to build capacity more gradually than a unidirectional system.

2. If only one fiber is available, then there is no alternative but to deploy bidirectional systems.

3. Unlike unidirectional system, bidirectional system do not require automatic switching protocol between the two ends of the link, since both ends detect a fiber cut simultaneously.

4. BLSR/4 four fiber SONET rings are better supported using bidirectional systems than unidirectional systems.

5. Bidirectional systems provide better support for deploying point to point SONET systems in 1+1 or 1:1 configurations.

6. Bidirectional system can be configured to handle asymmetric traffic. For ex, a large number of wavelengths could be used in one direction and a smaller number in the other direction.

7. It is more difficult to design the transmission system in bidirectional system since more impairments must be taken into account, in particular, reflections.

8. Although amplifiers for bidirectional system may employ more complicated structures than unidirectional systems, they need to handle only half as many channels as unidirectional systems, which means that they can produce higher output power per channel and provide more gain flatness.

9. Bidirectional systems usually require a guard band between the two sets of wavelengths traveling in opposite directions to avoid cross penalties, which may not be needed for unidirectional systems.

10. Bidirectional system can also be operated as a unidirectional system by simply reversing the direction of propagation of half the wavelengths and the corresponding amplifiers.

Application AreasInterexchange networks• Interexchange carriers in U.S have links

spanning several hundred to a few thousand km.

• Among the major carriers, AT&T and sprint have primarily installed standard single mode fiber. Thus WDM is an attractive option for them, and they are actively deploying WDM systems on many of their routes.

• Another major carrier MCI has a number of links using dispersion shifted fiber and is looking to deploy high speed TDM systems on these routes and WDM systems on other routes that use standard single mode fiber.

• In Europe and Japan the links tend to be shorter, several European carriers are considering the deployment of WDM.

Undersea Networks• Undersea links tend to be longer and may

span distances of several thousand km.• Shorter distance repeater less undersea links

are also being deployed in many places.• In long links, the case for WDM over standard

single mode fiber becomes quiet strong.• Reliability is a very important consideration for

undersea links, which are usually designed to operate over a 25 year period.

• Passive WDM multiplexing and demultiplexing technology, combined with 2.5 Gb/s systems, give WDM an edge over higher speed TDM systems.

• WDM also allows segregation of traffic between channels.

Local Exchange Networks• These networks tend to operate over shorter

distances, typically a few tens of km between major cities.

• Local exchange carriers will need to provide high capacity connections to their customers such as large enterprises.

• These customers may want to transmit different types of protocols at different bit rates.

• The bit rates tend to be somewhat lower compared to interexchange networks, typically a few hundred to several hundred Mb/s lower.

• Thus it may be useful to have several lower speed channels than a few high speed channels in the network.

• Several carriers in the U.S are considering the deployment of WDM on selected routes where they need higher capacities.

Enterprise Links• Large enterprises face an increasing need for a

number of high bandwidth connection between their sites.

• Many enterprises maintain two data centers that are separated by a few tens of km. All transactions are mirrored at both sites.

• This allows the enterprise to recover quickly from a disaster where one of the centers fails.

• There may be other reasons for this as well, such as lower real estate costs at one location than at the other.

• This allows peripheral equipment to be placed at the cheaper site.

• The large mainframes at these data centers need to be interconnected by several hundred channels, each at up to a few hundred Mb/s.

• These data centers tend to be located in dense metropolitan areas where most of the installed fiber is already in use.

• These networks use a variety of protocols and bit rates.

• These two factors make WDM an attractive option for these types of networks.

• WDM systems are being installed in large enterprise networks in the united states and all over Europe as well.

• Since the links tend to be short, there is usually no need to use optical amplifiers or regenerators.

• This makes the WDM option highly competitive provided multiple fibers are available. Or when the fiber is being installed from scratch.

OTDM

• Optical signals representing data streams from multiple sources are interleaved in time to produce a single data stream. The interleaving can be done on a bit by bit basis.

• It can also be done on a packet by packet basis.

• In both bit interleaved and packet interleaved case, framing pulse can be used.

• In the packet interleaved case, framing pulse mark the boundary between packets.

• In the bit interleaved case, if n input data streams are to be multiplexed, a framing pulse is used every n bits.

• OTDM networks can be based on a broadcast topology or incorporate optical switching.

• In case of broadcast OTDM networks the topology used is either star or a bus.

• Those photonic packet switching networks that incorporate photonic switching and routing will be referred to as switch based networks.

• In broadcast networks, there is no routing and switching within the network.

• Switching occurs only at the periphery of the network by means of tunable transmitter and receivers.

• Tunability in case of TDM networks refers to the ability to select one of several time multiplexed streams.

• The switch based networks perform routing and switching functions optically within the network in order to provide packet switched service at very high bit rates.

• The broadcast networks suffers from large splitting losses, are not scalable and are generally suitable only for LAN applications.

• The switch based networks on the other hand are scalable and suited for WDM applications.

Multiplexing and Demultiplexing• At the inputs to the network, lower stream data

streams are multiplexed optically into a higher speed stream and at the outputs of the network the lower speed streams must be extracted from the higher speed stream optically by means of a demultiplexing function.

• Functionally optical TDM is identical to electronic TDM. The only difference is that multiplexing and demultiplexing operations are performed entirely optically at high speeds.

• Very short pulses much shorter than the bit interval of each of the multiplexed streams must be used in OTDM systems.

• A periodic train of very short pulses with widths of the order of a few ps can be generated using a mode locked laser.

• Since the pulses are very short, their frequency spectrum will be large.

• Therefore unless some special care is take, there will be significant pulse broadening due to the effects of chromatic dispersion. For this purpose soliton pulses are used in OTDM systems.

• Solitons can propagate for long distances without dispersion induced broadening.

• Almost any pulse shape gets transformed into the soliton pulse shape after propagating some distance through optical fiber.

• The only requirement is that the pulse should have the right peak power.

• The required peak power is inversely proportional to the pulse width and thus directly proportional to the bit rate of operation.

• shape maintenance properties of soliton pulse are true only if the soliton pulses are sufficiently isolated from each other.

Bit Interleaving• The periodic pulse train generated by a mode

locked laser is split, and one copy is created for each data streams to be multiplexed.

• The pulse train for the ith data stream i = 1,2..n is delayed by iτ. This delay can be achieved by passing the pulse train through the appropriate length of optical fiber.

• Since the velocity of light in silica fiber is about 2 * 108 m/s, one meter of fiber provides a delay of about 5 ns. Thus the delayed pulse streams are nonoverlapping in time.

• The undelayed pulse stream is used for the framing pulses.

• Each data stream is used to externally modulate the appropriately delayed periodic pulse stream.

• The outputs of the external modulator and the framing pulse streams are combined to obtain the bit interleaved optical TDM stream.

• The power level of the framing pulses is chosen to be distinctly higher than that of the data pulses.

Demultiplexing• The multiplexed input is split into two streams

using say a 3 db coupler. If the jth stream from the multiplexed stream is to be extracted, one of these streams is delayed by jτ.

• A thresholding operation is performed on the delayed stream to extract the framing pulses.

• The reason the framing pulses were multiplexed with higher power than the other pulses was to facilitate this thresholding operation.

• A logical AND operation between the framing pulse stream and multiplexed pulse stream is used to extract jth stream.

• The output of the logical AND gate is a pulse if during a pulse interval, both inputs have pulses, the output has no pulse otherwise.

Packet Interleaving• A periodic stream of narrow pulses is externally

modulated by the data stream. If the bit interval is T, the seperation between the successive pulses is also T.

• We must somehow devise a scheme to reduce the interval between successive pulses to τ, corresponding to the higher rate multiplexed signal.

• This can be done by passing the output of the external modulator through a series of compression stages.

• If the size of each packet is l bits, the output goes through k = [log2l] compression stages.

• In the first compression stage bits 1,3,5,7 are delayed by T – τ.

• In the second compression stage, the pairs of bits (1,2), (5,6), (9,10).. Are delayed by 2(T – τ).

• In the third compression stage, the bits (1,2,3,4), (9,10,11,12) .. Are delayed by 4(T- τ).

• Each compression stage consists of a pair of 3 dB couplers, two semiconductor optical amplifiers(SOAs) used as on – off switches and a delay line.

• The demultiplexing operation is equivalent to “decompressing” the packet.

• This can be achieved by passing the compressed packet through a set of decompression stages that are similar to the compression stage.

• The on – off switches required in this approach must have switching times of the order of the pulse width τ.

• A more practical approach is to use a bank of AND gates.

• It converts the single high speed data stream into multiple lower speed data streams that can be processed electronically.

• A bank of AND gates is used to break up the incoming high speed stream into five parallel streams each with five times the pulse spacing of the multiplexed stream.

• One input to each AND gate is the incoming data stream, and the other input is a control pulse stream where the pulses are spaced five times apart.

• The control pulse streams to each AND gate are appropriately offset from each other so that they select different pulses.

Optical AND gates• The logical AND operation between two

signals can be performed by an on – off switch if one of the signals is input to the switch and the other is used to control it.

Nonlinear Optical Loop mirror• The nonlinear optical loop mirror (NOLM)

consists of a 3 dB directional coupler, a fiber loop connecting both outputs of the coupler and a nonlinear element (NLE) located asymmetrically in the fiber loop.

• First ignore the non linear element and assume that a signal is present at one of the inputs shown as arm A of the directional coupler.

• Then the two output signals are equal and undergo exactly the same phase shift on traversing the fiber loop.

• In this case both the clockwise and the counterclockwise signals from the loop are completely reflected onto input A: specifically, no output pulse emerges from arm B.

• However if one of the signals were to undergo a different phase shift compared to the other, then an output pulse emerges from arm B.

• In nonlinear amplifying loop mirror (NALM) the pulse traversing the fiber loop clockwise is amplified by an EDFA shortly after it leaves the directinal coupler.

• The amplified pulse has higher intensity and undergoes a larger phase shift on traversing the loop compared to an unamplified pulse.

• These configurations are not convenient for using the NOLM as a high speed demultiplexer.

• First the intensity dependent phase change in silica fiber is a weak nonlinearity and a few hundred meters of fiber are required in the loop to exploit this effect for pulse switching.

• It would be desirable to use a nonlinear effect that works with shorter lengths of fiber.

• Second to realize an AND gate, we require NLE whose nonlinear properties can be conveniently controlled by the use of control pulses.

• NALM has both these properties and is called terahertz optical asymmetric demultiplexer.

• TOAD has another directional coupler spliced into the fiber loop for the purpose of injecting the control pulses.

• The control pulses carry sufficiently high power and energy so that the optical properties of NLE are significantly altered by the control pulse for a short time interval after the control pulse passes through it.

• In particular the phase shift undergone by another pulse passing through the NLE during this interval is altered.

• For the proper operation of the TOAD as a demultiplexer, the timing between the control and signal pulses is critical.

• Assuming the NLE is located such that the clockwise signal pulse reaches it first, the control pulse must pass through the NLE after the clockwise signal pulse but before the counterclockwise signal pulse.

• If this happens the clockwise signal pulse experiences the unsaturated gain of the amplifier, whereas the counterclockwise pulse sees the saturated gain.

• The latter also experiences an additional phase shift that arises due to gain saturation.

• Owing to this asymmetry, the two halves of the signal pulse do not completely destructively interfere with each other, and a part of the signal pulse emerges arm arm B of the input coupler.

• Along with the signal pulse, the control pulse will also be present at the output. This can be eliminated by using different wavelengths for the signal and control pulses and placing an optical filter at the output to select only the signal pulse.

Soliton Trapping AND Gate• The soliton trapping AND gate uses some

properties of soliton pulses propagating in a birefringent fiber.

• In a birefringent fiber, these two modes propagate with different group velocities.

• As a result, if two pulses at the same wavelength but with orthogonal polarizations are launched in a birefringent fiber, they would walk off, or spread apart in time, owing to this difference in group velocities.

• However, soliton pulses are an exception to this walk off phenomenon.

• A pair of orthogonally polarized soliton pulses propagate in birefringent fiber without walk-off.

• What occurs is that the two pulses undergo wavelength shifts in opposite directions so that the group velocity difference due to the wavelength shift exactly compensates the group velocity difference due to birefringence.

• Since the two soliton pulses travel together, this phenomenon is called soliton trapping.

• The logical AND operation between two pulse streams can be achieved using this phenomenon if the two pulse streams correspond to orthogonally polarized soliton pulses.

• Most high speed TDM systems use soliton pulses to minimize the effects of group velocity dispersion so that the soliton pulse requirement is not a problem.

• The orthogonal polarization of the two pulse streams can be achieved by appropriately using polarizers.

• The logical AND operation is achieved by using an optical filter at the output of the birefringent fiber.

• Fig shows the block diagram of such a soliton trapping AND gate. It consists of a piece of birefringent fiber followed by an optical fiber.

• When pulses of both polarization are present at the wavelength λ, one of them gets shifted in wavelength to λ + δλ, and the other to λ – δλ.

• The filter is chosen so that it passes the signal at λ + δλ and rejects the signal at λ.

• Thus the passband of the filter is such that one of the wavelength shifted pulses lies within it. But the same pulse, if it does not undergo a wavelength shift, will not be selected by the filter.

• Thus the filter output has a pulse only if both pulses are present at the input, and no pulse (logical Zero) otherwise.

Synchronization• Synchronization is the process of aligning two pulse

streams in time. In photonic packet switching networks, it can refer either to the alignment of an incoming pulse stream and a locally available clock pulse stream or to the relative alignment of two incoming pulse streams.

• The two periodic pulse streams with period T are not synchronized because the top stream is ahead in time by ΔT. To achieve synchronization, the top stream must be delayed by ΔT with respect to the bottom stream.

• A fixed delay can be achieved by using a fiber of the appropriate length.

• In case of a synchronizer, and in some other applications in photonic packet switching networks, a tunable delay element is required since the amount of delay that has to be introduced is not known a priori.

Tunable Delays • A tunable optical delay line capable of realizing

any delay from 0 to T – 2-k in steps of 2-k .

• The parameter k controls the resolution of the delay achieveble.

• The delay line consists of k-1 fixed delays with the values T/2, T/4 …. T/ 2k-1 interconnected by k 2*2 optical switches.

• By appropriately setting the switches in the cross or bar state, an input pulse stream can be made to encounter or avoid each of these fixed delays.

• If all the fixed delays are encountered, the total delay suffered by the input pulse stream is T/2 + T/4 +…. T/ 2k-1 = T/ 2k.

• This structure can be viewed as consisting of k- 1 stages followed by an output switch.

• The output switch is used to ensure that the output pulse stream always exits the same output of this switch.

• Thus with tunable delay, the two pulse streams can be synchronized.

Optical Phase Lock Loop• Consider a NOLM that does not a separate

nonlinear element bur rather uses the intensity dependent refractive index of silica fiber itself as the nonlinearity.

• Thus if a low power pulse stream, say stream 1, is injected into the loop from arm A of the directional coupler, the fiber non linearity is not excited, and both the clockwise and the counterclockwise propagating pulses undergo the same phase shift in traversing the loop.

• As a consequence, no power emerges from the output in this case . If a high power pulse stream, say stream 2, is injected in phase with, say, the clockwise propagating pulse stream.

• owing to the intensity dependence of the refractive index of the silica fiber, the refractive index seen by the clockwise pulse,and hence the phase shift undergone by it, is different from that of the counter clockwise pulse.

• This mismatch in the phase shift causes an output to emerge from arm B.

• If the high power pulse stream is not in phase with the clockwise propagating pulse stream, the clockwise and counterclockwise pulses undergo the same phase shift and no output emerges from arm B of the directional coupler.

• To achieve snchronization between pulse streams 1 and 2, a tunable delay element can be used to adjust their relative delays till there is no output of stream 1 from the NOLM.

• Since pulses from stream 2 will always be present at the output, in order to detect the absence of pulses from stream 1, the two streams must use different wvelengths or polarization.

• When different wavelengths are used, owing to the chromatic dispersion of the fiber, the two pulses will tend to walk away from each other, and the effect of non linearity will be reduced.

• To overcome this effect, the wavelengths can be chosen to lie symmetrically on either side of the zero dispersion wavelength of the fibre so that the group velocities of the two pulse streams are equal.

Broadcast OTDM NetworksHeader Recognition:• For a header of fixed size, the time taken for

demultiplexing and processing the header is fixed, and the remainder of the packet is buffered optically using a delay line of appropriate length.

• The processing of header bits may be done electronically or optically, depending on the kind of control input required by the switch.

• The packet destination information from the header is used to determine the outgoing link from the switch for this packet, using a look up table.

• If the switch has two inputs and outputs, as shown in fig, the lookup table determines, for each input packet, whether the switch should be set in the cross state or bar state.

• Of course, this leads to a conflict if both inputs have a packet destined for the same output.

• This is one of the reasons for having buffers in the routing node.

• Several techniques have been proposed to simplify the task of header recognition.

• One common technique is to transmit the header at a much lower bit rate than the packet itself, allowing the header to be received and processed relatively easily within the routing node.

• The packet header could also be transmitted on a wavelength that is different from the packet data.

• It could also be transmitted on a separate subcarrier channel on the same wavelength.

Buffering:• A routing node contains buffer to store the

packets from the incoming links before they can be transmitted or forwarded on the outgoing links. Hence the name store and forward for these networks.

• In these networks, the buffer may be present at the inputs only, at the outputs only or at both the inputs and the outputs.

• There are 3 reasons for having to store or buffer a packet before it is forwarded on its outgoing link.

• First, the incoming packet must be buffered while the packet header id processed to determine how the packet must be routed.

• Second, the required switch input and/or output port may not be free, causing the packet to be queued at its input buffer.

• The switch input may not be free other packets that arrived on the same link have to be served earlier.

• The switch output port may not be free because packets from other input ports are being switched to it.

• Third, after the packet has been switched to the required output port, the outgoing link from this port may be busy transmitting other packets, thus making this packet wait for its turn.

• In photonic packet switching networks buffers are scarce resources.

• The best way to construct an optical buffer is to use a piece of optical fiber and delay the signal within it.

• Thus usually very small buffers are used in photonic packet switching networks.

• Unlike an electronic buffer, a packet cannot be accessed at an arbitrary point of time, it can exit the buffer only after a fixed time interval after entering it. This is the time taken by a packet to traverse the fiber length.

• In photonic packet switching networks, the switch operates at the speed of the input and output links but usually not faster.

• This means that the packet arrive to the output links from the switch no faster than the rate at which they can be transmitted on the output links.

• Thus the packets are transmitted immediately on arrival and no buffering at the output link is necessary.

• In a 2* 2 switch, if both the inputs have a packet for the same output, only one of them can be transmitted by the switch immediately.

• There are 3 options• First, option is for the packet to wait, which

necessitates an input buffer beyond what is needed merely for packet header recognition. This is expensive.

• The second option is for the packet to be dropped. This is not attractive as the source has to retransmit the packet.

• The third option is for the packet to be misrouted by the switch, that is, transferred by the switch to the wrong output.

• This option is termed as deflection routing.

Deflection Routing• Deflection routing is also sometimes called hot

potato routing.• Misrouting packets rather than storing them

will cause packets to take longer paths on average to get to their destinations, and thus will lead to increased delays and lesser throughput in the network.

Delay• Due to deflection routing the average delay

experienced by the packets in the network is larger than in store and forward networks.

• Here not only the topology of the network is fixed but the statistics of the packet arrivals between each source – destination pair is also fixed.

• The delay experienced by a packet consists of two components

• The first is the queuing delay, which is the time spent waiting in the buffers at each routing node for transmission. There is no queuing delay in the case of deflection routing.

• The second component of the delay experienced by a packet is the propagation delay, which is the time taken for the packet to traverse all the links from the source node to the destination node.

• The propagation delay is often larger for deflection routing than for routing with buffers owing to the misdirection of packets away from their destinations.

• This increased propagation delay almost always compensates for the lack of queuing delay in deflection routed networks so that the total average delay is higher for the same arrival rate.

Throughput• In deflection routing the throughput of the

network is decreased compared to routing with buffers.

• It depends on the interconnection topology of the network and the data rates on the links.

• In addition it depends on the traffic pattern, which must remain fixed in defining the throughput.

• The traffic pattern specifies the fraction of new packets for each source destination pairs.

• In all theoretical analysis the throughput is evaluated for a uniform traffic pattern, which means that the arrival rates of new packets for all source destination pairs in the network are equal.

• If all the links run at the same speed, the throughput can be expressed as a fraction of the link speed.

• The second property of the network that we must consider is its deflection index.

• The deflection index is the largest number of hops that a single deflection adds to the shortest path between some two nodes in the network.

• A single deflection adds at most four hops to the path length so that its deflection index is four.

• Network topologies with small diameter and small deflection indices are best suited for photonic packet switching networks.

• Performance of deflection routing networks can be further improved by using appropriate deflection rules.

• A deflection rule specifies the manner in which the packets to be deflected are chosen among the packets contending for the same switch output port.

• In the event of a conflict between two packets, both packets are equally likely to be deflected. This deflection rule is termed random.

• Another possible deflection rule called closest to finish, states that when two packets are contending for the same output port, the packet that is farther away from its destination is deflected.

• This reduces the average number of deflections suffered by a packet and thus increasing the throughput.

Small Buffers• Deflection routing, which uses no buffers

except for packet header processing, is also called hot potato routing.

• If the limited buffer is used and if these buffers are full, the packet is again deflected.

• Limited buffer deflection routing achieve higher throughputs than hot potato routing.

Livelocks• When a network employs deflection routing, there is

the possibility that a packet will be deflected forever and never reach its destination. This phenomenon has been called both deadlock and livelock.

• It can be eliminated by suitably designed deflection rules.

• It is difficult to find a deflection rule that is livelock free.

• One way to eliminate livelocks is to simply drop packets that have excedded a certain threshold on the hop count.

Feed Forward and Feedback Delay Lines• The throughput and delay performance of a

photonic packet switched network can be improved by addition of small number of buffers at each node.

• Two ways in which buffering of a few packets can be incorporate into an optical routing nodes are Feed Forward and Feedback Delay Lines.

• In the feed forward architecture a two input, two output routing node is constructed using three 2*2 switches interconnected by two delay lines.

• If each delay line can store one packet, that is the propagation time through the delay line is equal to one slot, the routing node has a buffering capacity of two packets.

• If two packets destined for the same output arrive simultaneously at both inputs, one packet will be routed to its correct output and the other packet will be stored in delay line 1.

• This is accomplished by setting switch 1 in the appropriate state. This packet then has opportunity to be routed to its desired output in a subsequent slot.

• If no packets arrive in the next slot, this stored packet can be routed to its desired output in the next slot by setting switches 2 and 3 appropriately.

• In the feedback architecture the delay lines connect the outputs of the switch to its inputs.

• With two delay lines, the switch is internally a 4*4 switch with two inputs from outside and two from the delay lines.

• If two packets contend for a single output, one of them can be stored in a delay line.

• If the delay line has length equal to one slot, the stored packet has an opportunity to be routed to its desired output in the next slot.

• If there is contention again, it, or the contending packet, can be stored for another slot in a delay line.

• The feed forward architecture is preferred to the feedback architecture since it attenuates the signal almost equally, regardless of the path taken through the routing node.

• This is because almost all the loss is in passing through the switches, and in this architecture, every packet passes through the same number of switches independent of the delay it experiences.