Optical packet switch architectures seminar or presentation report
Active In SP
Joined: Nov 2009
05-01-2010, 03:17 PM
Optical packet switch architectures seminar report.doc (Size: 216.5 KB / Downloads: 47) ABSTRACT
The current fast growing internet traffic is demanding more and more network capacity everyday. As telecommunication and computer communications continue to converge, data traffic is gradually exceeding telephony traffic. This means that many of the existing connection-oriented or circuit-switched networks will need to be upgraded to support packet-switched data traffic. The concept of wavelength division multiplexing has provided us an opportunity to multiply network capacity. Current photonic switching technologies allow us to rapidly deliver the enormous bandwidth of wavelength division multiplexing networks. Of all switching schemes, photonic packet switching appears to be a strong candidate because of high speed data rate/format transparency, and configurability it offers. The goal of this paper is to discuss some of the critical issues involved in designing and implementing photonic packet switched networks. Finally, we conclude by describing some of the emerging technologies with the potential to revolutionize optical packet switching.
The growth -of existing and new services will create a large increase of traffic flow in telecommunication network in the coming years .The demand for bandwidth in a communication network increase continuously .The world wide web alone Ëœfor example require a yearly 8 fold increase in bandwidth per user . All optical networks may be the only solution to cope with such increasing bandwidth .In these network transmitter signals remain in optical format on the way from source to destination
The wave length division multiplexing WDM technology offers a practical way to exploit the bandwidth of fiber optics by partitioning the optical bandwidth into separate channels .Photonic packet switching offers high speed ,data rate transparency ,and configurability which are some of the important characteristics needed in future networks supporting different forms of data
2. TYPES OF SWITCHING
There are three main types of switching:
1. Circuit switching
2. Packet switching
3. Cell switching
Circuit Switching :
This type of switching is the one used in the field of conventional telephony . In an end-to-end dedicated link is set up before communication starts. This link is solely used for this communication and after that it is broken .The steps involved in the process of circuit switching are
1. Circuit setup
2. The actual communication
3. Clearing of the circuit
Here data is sent as packets .Due to this fact, there is an efficient utilization of the communication line. The earlier problem of waiting does not occur here as more than one source can use the line at the same time.
All data to be transmitted is first assembled into one or more messages units called packets, by the source DTE.These packets include both source&destination DTE to its local packets switching exchange (EXE).On receipt of each packets, the exchange first stores the packet and then inspects the destination address it contains . Each PSE containing routing directory specifying the outgoing links to be used for each networkaddress.The PSE forwards the packet on the appropriate link at the maximum available bit rate.
This is a subset of packet switching. It deals with packets of fixed size, for eg. The ATM cells. It buffers in all the switches follow the FIFO discipline, the packets are delivered in order in the case of cell switching, where as in packet switching the packets may not arrive in order. Cells are 53 bytes long, of which 5 bytes are header and 48 bytes are pay load.
3. PHOTONIC SWITCHING
In photonic switching the signals are kept in optical form, while they are being routed from input to the output. It does not convert optical signals into electrical form for switching purpose. Though the lack of interaction among photons would appear to make photonic switching unrealizable, but fortunately optical non-linearity provides bi-stability, which enables optical switching and promises optical computation also similar to digital computers. In fact optical bi-stability enables switching at energies comparable to electronics. Recently, optically controlled photonic switching devices have been developed, which have the potential for switching rates in tetra hertz range.
Photonic switching has three main advantages:
1. No need for optical to electronic conversions.
2. The ability to route high data rate optical signals.
3. Possibility of three dimensional inter-connections.
Though electronic devices need less power, are more controllable and faster than optical logic devices, they rapidly become slowly in operation by electrical connections between them as a result of cross talk, dispersion, and unwanted capacitance. The photonics would be desirable for long wide band inter-connections.
The main disadvantages of photonic switching are:
1. Large physical size of devices.
2. Absence of all optical clock regenerators to remove the loss and cross talk introduced in cross points
3. The polarization sensitivity of devices
3.1 Functions in a switch :
1. Routing: providing networks connectivity information through routing tables.
2. Forwarding: defining the output for each incoming packet based on routing table.
3. Switching: directing each packet to proper output defined by forwarding process.
4. Buffering: resolving contention by storing packets.
3.2 Photonic packet switches:
A WDM optical packet switch consists of four parts; input interface, switching fabric, output interface and control unit .The input interface is mainly used for packet delineation and alignment, packet header information extraction and packet header removal. The switch fabric is the core of the switch and is used to switch packets optically. The output interface is used to regenerate the optical signals and insert the packet header. The control unit controls the switch using the information in the packet headers. Because of synchronization requirements, optical packet switches are typically designed for fixed size packets.
When a packet arrives at a WDM optical packet switch, it is first processed by the input interface. The header and payload of the packet are separated and the header is converted into the electrical domain and processed by the control unit electronically. The payload remains an optical signal throughout the switch. After the payload passes through the switching fabric, it is recombined with the header, which is converted back into the optical domain at the output interface.
4. WAVELENGTH DIVISION MULTIPLEXING (WDM)
This utilizes the full information capacity of the optical fiber. In a normal point to point optical fiber communication link one fiber is needed with one optical source and one detector. The light source works on very narrow spectrum, say 0.85Ã‚Âµm or 1.3Ã‚Âµm etc. The optical fiber has a very low attenuation in the wavelength range of 0.85Ã‚Âµm to 1.6Ã‚Âµm and so several light sources working on a properly spaced peak emission wavelength could be simultaneously used with single fiber provided the arrangement is such that no crosstalk takes place with individual sources. Such a process of sending information is called WDM.
WDM process could be achieved by either of the two processes shown in the figure. The WDM system shown first has an N optical sources being used to separate out individual wavelengths, which are then applied to corresponding detectors 1 to N respectively. This kind of WDM system is unidirectional in nature. The second figure shows another WDM system, which is bidirectional in nature. Here, one information at one wavelength 11 sent in one direction. At the same time, information at wavelength 12 is sent in opposite direction.
The basic performance criteria for a WDM technique are insertion loss, channel width and crosstalk. The insertion loss arises due to the presence of multiplexing and demultiplexing devices in the fiber optic link. This may be at the connection points or inside the multiplexer itself. Insertion loss of only a few dB at each end of the multiplier is normally acceptable. Channel width is defined as the wavelength range allocated to an optical source working with WDM.For a laser diode source, channel width several tens of nanometers are required so that there may not be interchannel interference, which might result due to source instability. For LED sources because of their wider spectral output width 10 to 20 times larger are needed cross talk is the amount of coupling that might take place from one optical source to another Normally a 30 dB is adequate .
There are two varieties of WDMs, namely,
1. The angularly dispersive devices:
2. The filter type multiplexing devices:
1. The angularly dispersive devices
The principle of working of the an angularly dispersive WDM unit is shown in figure. This uses two lenses and an angularly dispersive element .Though this may not be both as multiplexer or de multiplexer of certain advantages, the multiplexer action is preferable. Light signal coming out of the fiber is incident on lens L1 then passed through an angularly dispersive element, which separate the wavelength channel into different spatially oriented beams. Lens L2 focuses this beam in to the appropriate photo detectors. A large number of channels may, thus be combined or separated with such angularly dispersive multiplexing elements. Insertion losses are in the 1 to 3 dBrange and crosstalk level is between 20 and 30 dB.
2. The filter type multiplexing devices:
This type may be used both as multiplexers and demultiplexers. The device has a flat glass substrate upon which multiple layers of different films are deposited for wavelength selectivity. This film work like a filter in the sense that each film is transparent to only one wavelength and reflects all others. While using it as a multiplexer, each wavelength is brought the device by a fiber whose output is collimated by a rod assembly of variable focal length. The collimated beams corresponding to the input wavelength pass through its own filter and are refocused onto the output fiber. All other wavelength under go one or more reflection and are eventually refocused onto the output fiber. In the demultiplxer operation the output fiber becomes the input fiber while all the other fibers provide the output wavelengths of interest.
5. OPTICAL PACKET SWITCHING TECHNIQUES:
5.1 Packet coding techniques:
Several optical coding techniques have been studied. There are three basic categories.
1. Bit serial
2. Bit parallel
3. Out of band signaling
1. Bit serial:
Bit serial coding can be implemented using optical code division multiplexing (OCDM), optical pulse interval, or mixed rate techniques. In OCDM each bit carries its routing information, while in later two techniques; multiple bits are organized into packet pay load with a packet header that includes packet information. The difference between later two techniques is that in optical packet interval, the packet header and pay load are transmitted at the same rate, whereas in the mixed rate techniques, the packet header is transmitted at the low rate than the pay load so that the packet header can be easily processed electronically.
2. Bit parallel:
In bit parallel coding, multiple bits transmitted at the same but on separate wavelengths.
3. Out of band signaling:
This includes sub carrier multiplexing (SCM) and dual wavelength coding. In SCM, the packet header is placed in an electrical sub carrier above the base frequencies occupied by the packet pay load, and both are transmitted in the time slot. In dual wavelength coding, the packet header and pay load are transmitted in separate wavelengths but in the same time slot.
5.2 CONTENTION RESOLUTION
This is necessary in order to handle the case where more than one packet are destined to go out of same output port at the same time. T his is a problem that commonly arises in packet switches and is known as blocking. It is typically resolved by buffering all the contending packets, except one, which is permitted to go out. I n optical packet switch techniques, designed to address the external blocking problem include optical buffering, exploiting the wavelength domain, and using deflection routing.
5 2.1 Optical buffering:
Currently optical buffering can only be implemented using optical delay lines (ODL). An ODL can delay a packet for a specified amount of time, which is related to the length of the delay line. Delay lines may be acceptable in prototype switches, but are not commercially viable. The alternative, of course, is to convert the optical packet to the electrical domain and store it electronically. This is not an acceptable solution, since electronic memories cannot keep up with the speeds of optical networks.
There are many ways an ODL can used to emulate an electronic buffer. TF or instance, a buffer for N packets with a FIFO discipline can be implemented using N delay lines of different lengths. D delay line 1 delays a packet for one time slot. A counter kee the track of the number packets in the buffer. Suppose the value of the counter is j when a packet arrives at the buffer, then the packet will be routed to the jâ„¢th delay line. Limited by the length of the delay lines, this type of the buffer is very small, and does not scale up.
5.2.2. Exploiting the wavelength domain:
In WDM several wavelengths run on a fiber link that connects two optical switches. This can be exploited to minimize external blocking as follows. Let us assume that two packets agree designed to go out of the same output port at the same time. Then they can be still transmitted out, but on two different wavelengths. This method may have some potential in minimizing external blocking; partly since the no of wavelengths that can be coupled together on to a single fiber continues to increase. For instance, it is expected that in a year there will be as many as 200-wavelengt/fibres.
5.2.3. deflection routing:-
Deflection routing is ideally suited to switches that have little buffer spaces. When there is a conflict between two packets, one will be routed to the correct output port. In this way. Little or no buffer is needed. However the deflected packet may end up following a longer path to destination since they are likely to arrive out of sequence.
6. OPTICAL PACKET SWITCH ARCHITECTURES :-
6.1 An architecture with a space switch fabric
Space switch fabric architecture is shown in figure. The switch consists of N incoming and N out going fiber links, with n wavelengths running on each fiber link. The switch is slotted, and the length of the slot is such that an optical packet can be transmitted and propagated from an input port to an out put optical buffer.
The switch fabric consists of three parts; optical packet encoder, space switch, and optical packet buffer. The optical packet buffer works as follows. For each incoming fiber link, three is an optical demultiplexer, which divides the incoming optical signal to different wavelengths. Each wavelength is fed to a different tunable wavelength converter ( TWC), which converts the wave length of the optical packet to a wavelength that is free at the destination optical output fiber. Then, through the space switch fabric, the optical packet can be switched to any of the N out put optical buffers. Specifically the out put of a TWC is fed to a splitter, which distributes the same signal to N different out put fibbers, one per out put fibber. The signal on each of these out put fibbers goes through another splitter, which distributes this in to d+1 different out put fibbers, and each out put is connected through an optical gate to one of the ODLs of the destination out put buffer. The optical packet is forwarded to an ODL by appropriately keeping one optical gate open and closing the rest. The information regarding to which wavelength of an incoming packet and the decision as to which ODL of the destination out put buffer the packet will be switched to is provided by the control unit, which ahs knowledge of the state of the entire switch.
Each out put buffer is an optical buffer implemented as follows. It consists of d+1 ODLs, numbered from 0 to d. ODL1 delays optical packet for a fixed delay equal to 1 slots. ODL0 provides zero delay, and a packet arriving at this ODL is simply transmitted out of the out put port.
Each output buffer is an optical buffer implemented as follows. It consists of d+l ODLs, numbered from 0 to d. ODL1 delays optical packet for a fixed delay equal to 1 slot. ODLO provides zero delay, and a packet arriving at this ODL is simply transmitted out of the output port. Each ODL can delay optical packets on each of the n wavelengths. For instance, at the beginning of a slot, ODL1 can accept up to n optical packets, 1/wavelength, and delay them for 1 slot. ODL2 can accept up to n optical packets at the beginning of each slot, and delay them for 2 slots. That is, at slot t, it can accept up to n packets (1/wavelength) and delay them for 2 slots; in which case, these packets will exit at the beginning of slot t+2. However, at the beginning of slot t+1, t can also accept another batch of n optical packets. Thus a maximum of 2n packets may be in transit within ODL2; similarly for ODL3 through d. let ci denote the number of optical packets on wavelength 1, where 1=1,2,3...n. we note that these 11optical packets may be on a number of different OLDs. To insert an optical packet in to the buffer, we first check all the ci counters to find the smallest one, say cs; then we set the TWC associated with this optical packet to convert the packetâ„¢s wavelengths to 1s, increase cs by 1, and switch the optical packet to ODL cs. If the smallest counter cs is larger than d, packet will be dropped.
6.2. Architectures With A Broadcast And Select Switch Fabric: -
In this section we describe two different architectures with a broadcast and select switch fabric. In these architectures, packets from all input ports, each on a different wavelength, are combined with in the switch and are broadcast to all output ports. Wavelength selectors are then used at each output port to select a wavelength, and consequently a packet, to be sent out the switch. This type of switch fabric lends itself to multicasting.
6.2.1. The KEOPS Switch with a broadcast and select fabric:
This switch is developed as a part of the European ACTS keys to optical switching (KEOPS) project and implimentation. Each input and output fiber carries only one wavelength as shown in figure. Note that the wavelength of an output port is not fixed, and varies with packets. There for, the output interface is responsible for making it meet the requirement of output signal. The switching fabric consists of three blocks: encoder, buffer and selector. The wavelength encoder block consists of N fixed wavelength converters (FWCs), one per input, and a multiplexer. The buffer block consists of a splitter, K OLDs, and a space switching stage implemented by means of splitters, optical gates, and combiners. Finally the wavelength selector block consists of N wavelength channel selectors implemented by means of demultiplexers. These three blocks makes up the broadcast and select switch fabric.
The switch is slotted. At the beginning of a time slot, each wavelength converter in the wavelength encoder block converts the wavelength of the incoming packet to a fixed wavelength. The output of the N converters is combined and then distributed through a splitter into K different ODLs. Each ODL has a different delay, which is an integer number of slots. That is, ODL 1 has a delay of 1 slot. The N optical packets are stored simultaneously to the K different ODLs. At the beginning of the next slot, a maximum of K*N optical packets exit from the K ODLs and up to N of them are directed to their destination output ports with out any collisions. This is achieved through a combination of splitters, optical gates, demultipiexers and multiplexers. Specifically the output signal from each ODL goes through the splitter, which distributes it over N outputs. We recall that this output signal consists of N multiplexed optical packets, one for each wavelength. The signal from output j of each splitter is directed to output port j. since there are K such splitters, there are K such output signals, of which only one is selected and directed to output port j. this selected output signal is fed in to a demultiplexer, which breaks it up in to the N wavelengths, of which only one is transmitted out. A control unit manages the operation of this broadcast and select switch fabric.
In this switch, an optical packet consists of a header, payload and guard time. The header may include information about the destination, payload type, priority and so on. The payload is the user data. A guard time is used to allow for the setup time of the optical devices in the switch. it may be inserted between the header and the payload, or between two successive packets on the same wavelength. Mixed rate coding is used. That is the header is encoded at a low fixed bit rate (e.g.,622 Mb/s), and the payload rate may vary from a few hundred mega bits per second to 10 Gb/s. moreover, the packet length is fixed in time, not in the number of bits. That is the duration of the packet is fixed, but the size of the packet is variable. This packet format has two advantages. First, the processing speed of the logic in the WDM packet switches depends on the header rate, but not on the payload rate. Second, the buffering space in the WDM packet switches, realized by means of ODLs whose length is proportional to the time length of the packet to be stored, does not depend on the payload rate.
This switch architecture can be extended to the case where it has M input and output fibers, and each input and output fiber carries n wavelengths. This is achieved by multiplexing the signal from each incoming fiber to the n wavelengths, and then treating the switch as if it has n*M input wavelengths instead of N represented above. At the output side, each group of n wavelengths can be combined together through a combiner onto the same output fiber.
6.2.2 A Switch With A Broadcast And Select Fabric And Re-circulation Buffer:
The idea of using a re circulation buffer comes from an ATM switch known as a starlite switch. As in the previous switch architecture, there is a single wavelength for each input and output fiber and the wavelength of an output port vary with packets. The broadcast and select switch fabric is implemented through a coupler, which combines up to M input wavelengths and then distributes the combined signal to N tunable optical fibers (TOFs) and M fixed optical fibers (FOFs). Note that M is larger than N. the input to the coupler comes from N input wavelengths and M wavelengths, which are part of the feedback process, explained below.
The switch is slotted and is controlled by a control unit. At the beginning of each time slot, the control unit knows the destination output ports of the incoming optical packets from the input ports and the 1 time slot delay line. Accordingly, it instructs the TWCs at the input ports, the tunable optical filters at the output ports and the optical gates. Up to M optical packets are fed in to the coupler; according to their destinations, up to n of them are passed through the TOFs and out to the output ports, and the remaining packets are re circulated through an ODL. The re circulated optical packets are fed back to the coupler at the beginning of the following slot.
6.3 WITH A WAVELENGTH ROUTING SWITCH FABRIC:
In this section, we describe three architectures based on the wavelength routing switch fabrics. The switching procedure in these three architectures can be divided in to two phases. In the first phase, packets are sent to the ODLs for connection resolution. In the second phase, packets are routed to the correct output ports through the wavelength routing switch fabric.
6.3.1. An Input Buffered Switch:
This switch is shown in the figure. Each incoming and outgoing link carries a single wavelength. The wavelength of an output port varies with packets. The switch consists of the scheduling section and the switching section.
The scheduling section is used for contention resolution and is composed of N TWCs, one for each incoming wavelength, two K*K arrayed wave- guide gratings (AWGs) and M ODLs. Where K=max (N,M). An AWG is a wavelength routing device that can route optical signals from different input ports to different output ports based on their wavelength. The combination of all these optical devices provides for optical buffering of N individual buffers, each of which has M positions. If there are available buffer spaces, a packet entering input 1 of the first AWG will appear at output lof the second AWG after a specific delay. The length of the delay is determined by the wavelength of the packet when it enters the first AWG. Specifically, each TWC converts the wavelength of an incoming optical packet so that the optical packet, when routed through the AWG, joins the ODL with appropriate delay. The delay of an optical packet is selected using the following rules: first, no two optical packets may appear in the same slot at the same switch output; second, no two optical packets may appear at the same buffer output at the same slot.
The switching section is used for switching optical packets to their destination output ports and is made up of an AWG and TWCs. The TWCs are used to assign the optical packet the right wavelength corresponding to the desired output port.
The switch suffers from head of line blocking, which is inherent in input buffering switches. For eg. Suppose that optical packet 1 in input I must be routed to output 1, while optical packet 2 behind optical packet 1 in input must be routed to output 2. if optical packet one must be delayed for 1 time slot, optical packet 2 has to be delayed for at least one time slot due to the second rule, even though optical packet 2 goes to a different output port. However, if optical packet 1 has to be delayed by more than one time slot optical packet 2 need only be delayed by 0 slot as long as packet 2 has no conflict at switch output 2.
6.3.2. An Input Buffered Switch With A Distribution Network:
This switch was developed as a part of the KEOPS project and implimentation, and is shown in figure. Each incoming and outgoing fiber carries a single wavelength. The wavelength of an output port varies with packets. The switch consists of two stages: contention resolution and switching. In the first stage, through the de-multiplexer, each input port is connected with at least one ODL in each of N ODL sets. The TWC in the first stage decides to which ODL an optical packet will be sent. The second stage is used for switching optical packets to the correct output ports. Through the de multiplexer in the second stage, each ODL is connected to each output port. The TWC decides to which output port an optical will be sent.
Logically the first stage can be divided in to two parts: distribution and input buffer. The distribution part distributes optical packets from the same input to different input buffers. Note that if we remove the distribution part (i.e., each demultiplexer can only connect to one ODL set), the switch becomes identical to the one described above. The distribution part helps overcome the head of line blocking. Let us refer to the example of the two optical packets discussed in the previous subsection. In this case, regardless of how many time slots optical packet 1 must be delayed, it has no effect on the delay of optical packet 2, because the two optical packets could be routed to different ODL sets.
An optical packet arriving at the switch at time slot t is routed to the ODL with length d determined using the following three constraints. First, no other optical packet is scheduled to the same output port in time slot t+d. second, no other optical packet is scheduled to the same TWC in the second stage in time slot t+d. third, no optical packet from the same input port and to the same output port is scheduled in di with dl>=d. the ODL with the shortest delay satisfying these three constraints is selected.
In the WDM version of this switch, there are M incoming and outgoing fiber links, and each carries N wavelengths. N plains are located between M demultiplexers and M combiners connected to the incoming and outgoing fibers, respectively. Each plane is an N*N standard wavelength routing switch as described above. However in the WDM switch an outgoing fiber may carry more than one wavelength. The control part of the switch makes sure that there is no wavelength in the combiner.
6.3.3. The WASPNET Switch:-
This switch is proposed as part of the Wavelength Switch Optical Packet Network (WASPNET) project and implimentation. The configuration of the WASPNET switch with single wavelength input and out put is shown in figure. It consists of a 2N*2N AWG, N sets of ODLs, and 4N TWCs. As in the previous two architectures, the switch can be divided in to two phases.
First, optical packets are routed to the ODLs to resolve contention, and then routed to the desired output port. However in this switch these two phases are implemented together by a 2N*2N AWG and N ODL sets. The 2N TWC on the left of the awg are used to select the AWGs out put. The first N TWCs on the right of the AWG are used to select the correct ODLs for the optical packets that will be re circulated. The other N TWCs are used to convert optical packets to the wavelength required by the switch output interface, because there are more wavelengths inside the switch than the incoming and outgoing wavelengths. One advantage of this switch is that it can support optical packet priorities. That is, after leaving the delay line, an optical packet may be delayed again because of preemption by a higher priority optical packet.
The WDM version of this switch is made up of demultiplexers, combiners, and multiple plans of wavelength routing switch fabrics, each with single wavelengths input and outputs. It has N input and output fibers, each with n wavelengths. There are n planes, each corresponding to one of the N wavelengths. For example, wavelength I in each input fiber is always demultiplexed to the plane I. in view of this, the input of each plane have the same wavelength. However different wavelengths may appear at the output of each plane. In one time slot, the switch allows multiple optical packets to leave not only from the same output of the WDM switch, but also from the same output of a single plane. An N*N AWG is inserted between the N TWCs (N+l . . . 2N) on the right of the 2N*2N AWG and the output of a plane for this function. Now the N TWCs carry out more functions. For instance, they make a final routing decision in addition to assigning optical packets to the wavelengths required by the switch output interface. The control part of the switch makes sure that there is no wavelength conflict in the combiner.
7. LOOKING INTO FUTUER
In the future optical tag switching, optical burst switching, MEMS optical switching will likely play an important part in the architecture and system of photonic packet switched networks.
7.1 OPTICAL TAG SWITCHING:
Tag switching, as an alternative approach, has been p proposed to simplify the packet forwarding process. It assigns a short-fixed length label containing routing information, a so- called tag, to multiprotocol packets for transport across interconnected sub networks.
A tag switched network consists of:
1. Tag edge routers
2. Tag switches
3. Tag distribution protocol
When a tag edge router receives a packet for forwarding across the network, it analyses the network layer header, performs applicable network layer services, selects a route for the packets, and applies a tag to the packets. Then it forwards the packets to the tag switch. The tag switch receives the tagged packet and switches the packet bas ed on the tag , without reanalyzing the network layer header .T he packets reaches the tag edge router at the exit point of the network, where the tag is removed and the packet delivered.
7.2 MEMS OPTICAL SWITCHING:
MEMS technology is beginning to impact many areas of science and industry. It has shown a bright future of achieving high quality optical switching .MEMS devices are built in a similar manner to silicon integrated circuits. Various layers of different materials are deposited and patterned to produce complicated, multilayer, three dimensional structure. They can provide low loss and low cross talk while remaining compact in size and providing good economy due to monolithic batch production.
7.3 OPTICAL BURST SWITCHING:
OBS was proposed as another way of implementing packet switching optically to avoid potential electronic bottlenecks. The basic unit of data to be transmitted is a burst, which consists of multiple packets. T he data burst is sent after a control packet reserves necessary resources on the intermediate nodes with out waiting for acknowledgement from the destination node. OBS could achieve high bandwidth utilization with lower average processing and synchronization overhead then pure packet switching since it does not require packet by packet operation. It is also possible to implement quality of services by manipulating the offset time between the control packet and the data burst.
1. No management algorithms to optimize the circuits continuously
2. Problems when traffic between various IP routers will change suddenly can be avoided.
3. As pay load of packets remains optical throughput, both analog and digital transmissions can be done at different pay load formats
4. Since fiber can support very high bit rate, (band width of fiber-40 Hz by conservative estimate)
1. Multiple packets for single destination can be merged using optical TDM techniques into single pay load.
2. The merged packets can be separated at the destination again using optical TD demultiplexing techniques.
5. Unlimited band width
6. Extremely low error bit rates.
Photonic packet switching technology, although promising, is still in its infancy .It is capable of removing the critical limits placed by large electronic systems. The lack of commercially viable optical buffering technology imposes a constraint on the commercial development of optical packet switches. Also, it is not clear how such switches can be deployed to carry IP traffic. For instance, what would be an ideal packet size, and how much memory is required in an optical switch are questions that still need to be addressed. Optical packet switching is promising because it offers much higher capacity and data transparency. Progress has been made in several areas, but not all. Meanwhile, there is a tremendous increase in the processing speeds and capacity of electronic switches and routers. It is important for network designers to reduce the numbers of protocol layers being used in todayâ„¢s network, while preserving the functionality and making use of current optical technology.
1. ËœFiber Optic Communicationsâ„¢ by D.C. Aggarwal, Second Edition.
2. ËœOptical Fiber Communicationâ„¢ by Gerd Keiser, Second Edition.
3. IEEE Communication Magazine, February 2000.
4. IEEE Communication Magazine, January 2001.
5. IEEE Communication Magazine, September 2001.
6. IEEE Communication Magazine, November 2001.
2. Types of switching
3. Photonic switching
3.1 Functions in a packet switch
3.2 Photonic packet switching
4. Wavelength division multiplexing
5. Optical packet switching techniques
5.1Packet coding techniques
5.2 Contention resolution
5 2.1 Optical buffering
5 2.2 Exploiting the wave length domain
5 2.3 Deflection routing
6. Optical packet switch architectures
6.1An architecture with a space switch fabric
6.2 architectures with a broadcast and select switch fabric
6 2.1 The KEPOS switch with a broadcast and select switch fabric
6 2.2 A switch with a broadcast and select fabric and recirculation buffer
6.3 Architectures with wave length routing switch fabric
6 3.1 An input buffered switch
6 3.2 An input buffered switch with a distribution network
6 3.3 The WASPENT switch
7. Looking into the future
Use Search at http://topicideas.net/search.php wisely To Get Information About Project Topic and Seminar ideas with report/source code along pdf and ppt presenaion