WO2003065664A1

WO2003065664A1 - A networking element adapted to receive and output also preambles of data packets or frames

Info

Publication number: WO2003065664A1
Application number: PCT/US2003/001489
Authority: WO
Inventors: Jens P. Tagore-Brage
Original assignee: Vitesse Semiconductor Corporation; Elhoj, Martin
Priority date: 2002-01-30
Filing date: 2003-01-21
Publication date: 2003-08-07

Abstract

Preambles of data packets, frames, or cells, such as Ethernet packets, are used to transport information. This information is received, normally as part of a data packet, by a networking element and output there from, normally also as part of a data packet, in order for this informatin to not only be transferred between two neighbouring networking elements. The preamble information may be altered prior to outputtig and may be 'piggy backing' on different data packets before and after a networking element. The preamble information may be used as e.g. a separate communication channel, for network management or as a means of increasing the bandwidth of the network.

Description

^■ A NETWORKING ELEMENT ADAPTED TO RECEIVE AND OUTPUT ALSO PREAMBLES OF DATA PACKETS OR FRAMES

The present invention relates to a networking element, such as an Ethernet switch, router, or repeater, which is adapted to receive non-trivial preambles of data packets or frames and to output these - such as as parts of the same or other data packets or frames.

US-A-5,615,213 relates to the transmission of small messages, either out-of-band or . collectively as normal packages on the network. A processor will determine, from the size of the message packet, whether to transmit it over the network as a normal packet or to enqueue it for special transmission. The enqueued messages are then transmitted to the receiver depending on the length of that queue.

US-A-4,956,841 relates to source routing where a packet, which is to visit a number of nodes, is provided with a number of arrows which each identifies how a given switch is to switch the packet. The arrow, which was used by one switch for switching the packet, is removed in the subsequent switch. An arrow indicates the output port (artery) by which it must exit the switch.

US-A-5,625,621 relates to a switch where each individual port thereof may be operated in accordance with the type of device connected thereto. The ports may be operated in a port mode or an adapter mode. Changing the mode will entail electrically adapting to the device to obtain that all cabling can be of the same polarity and that no manual . intervention is required.

US-A-6,301 ,257 relates to source routing where a first switch floods a first packet to a destination to all switches connected to the first switch. All switches which are to receive the packet responds with switch ID and port number which is stored in a memory of the first switch (as a destination tag) and is subsequently prepended to packets for the same destination. In this manner, the subsequent switches need not perform look-up when receiving packets having a prepended destination tag.

US-A-5,742,604 relates to an encapsulation of a standard VLAN frame and thereby creating a proprietary non-Ethernet standard compliant data transport between link Patent 20 US 2

partners. The Inter Switch Link packet has, in addition to the Ethernet packet fields, an ISL destination and an ISL^'source address as well as a CRC.

US-A-4,780,870 relates to a switch, which internally provides packets with ID of the destination module and the particular processor on that module which is to receive the packet. This format, however, is only used internally in the switch.

US-A-5,390,173 relates to a non-Ethernet compliant proprietary packet format used only between link partners understanding this particular format. In this format, a local address (shorter than the normal address of the packet) is added in addition to a CRC. The encapsulation may also comprise congestion information, priority, results of address translations or the like. This encapsulation is used in a network segment where all elements understand it but where it is stripped before the packet is transmitted on to a network segment, which understands only standard compliant packet formats. ^" .

WO01/17314 relates to a method where part of a packet header is replaced by a smaller header with which the packet is then transmitted over the network. When being emitted, from the network, the original header is re-established.

EP-A-1 065 834 relates to a switch which may be cascaded with other elements into a • loop configuration where, on the loop, a special data packet format having a pretag with a ring ID is used. When outputting data packets, this ring ID is removed.

EP-A-0 996 256 and 1 003 306 relate to two modes of operating a number of switches in relation to each other. One mode is a oop mode where an oυt-of-band pretag is provided with the ID of the originating switch so that a packet may be identified if it returns to the originator. The other mode' relates to an in-band pretag comprising the ingress port number of the receiving switch and the egress port number of the backbone switch in order for that switch to merely transmit the packet, including the pretag if desired, from the port indicated. In this mode, no look-up is performed in the backbone switch. The in-band pretag completely fills and fully replaces the preamble and start-of-frame delimiter of the incoming packet. This preamble and start-of-frame delimiter are re-generated before outputting form this stacked switch. ' . » Patent 20 US 2

On top of this, the IEEE has formed an IEEE 802.3 EFM Task Force, which looks into the use of Ethernet preambles in link management between two adjacent link partners.

However, if management information and other types of information are to be transported efficiently, it needs to be able to travel over network hops. The present invention relates to a manner of transporting preamble parts over networking elements.

In a first aspect, the invention relates to a networking element adapted to receive and output data packets or frames, the element comprising:

one or more inputs for receiving a first preamble part, means for deriving, from the first preamble part, a second preamble part, a plurality of outputs for outputting preamble parts, and first means for determining from which of the plurality of outputs to output the second preamble part.

Network units may be any unit used for networking, such as Ethernet switches, routers, bridges, repeaters, hubs, Network Interface Cards, MUXes deMUXes, or computers. Naturally, these units may be used in mixed networks where part of the information carried in SONET containers is converted into preambles in an Ethernet stretch of the communication channel and then reconverted when converting into SONET again.

The primary feature of the invention is the fact that the preamble', which in e.g. Ethernet has originally been used for receiver synchronization, is now both received and output by the networking element. This preamble may comprise valuable information relating to the packet, the transmitting networking element, information for the present or a subsequent networking element - or simply be part of a side band channel being a communication channel "piggy backing" on the data frames or packets.

Thus, it should be clear that the preamble is not comprised in the part of the packet or frame originally used for holding data (payload) or addressing information (such as MAC or IP addresses in Ethernet packets). In Ethernet or IEEE 802.3 packets, the original preamble is the 7 bytes of alternating "1"'s and "0"s preceding the SFD and addressing or routing information. If desired, the Start-of-frame delimiter byte following the preamble Patent 20 US 2 may also be used for carrying data and may, in the present context, be taken functionally as part of the preamble.

In the present context, a data packet or frame may be any type of data packet or frame having a preamble and a data part, such as packets or units conforming to a standard having one or more address fields, at least one data field, and a header. Standards of this type may be the Ethernet standard, FDDI or other standards also using preambles (such as Token Bus). •

The preferred networking element has a plurality of output ports and means for determining from which output port(s) to output at least the preamble part of a received packet or frame. A networking element of this type may be a switch, router, repeater, or hub and which is connected to each port to a computer or network (spanning from a single computer to the WWW).

In the present context, a preamble part is at least a part of a preamble (including the SFD). It is not required that the full preamble is provided with information deviating from the original bit pattern. In fact, it is preferred that the first byte of the preamble is left unused due to the fact that certain communication standards overwrite this byte. Preferably, the preambles transmitted have the size determined by the standard in question - where then part of or all of it comprises non-trivial data.

In one embodiment, the deriving means is adapted to derive the second preamble part to be identical to the first preamble part. There are a number of situations where simple transport of the preamble part or information held thereby is desired. .

In another embodiment, the deriving means is adapted to provide the second preamble part as a part of the first preamble part, such as by removing part of the first preamble part. A situation of this type is one where the preamble part comprises routing information for use in a present networking element and subsequent networking element(s) where the routing information of this networking element is no longer relevant and may be removed. Thus, the first determining means could be adapted to base the determination on the part of the first preamble part not being comprised in the second preamble part.

Then, a remaining part of the information in the preamble part may be sent on. Patent 20 US 2

Also - or alternatively, the deriving means could be adapted to perform a mathematical operation on information comprised in the first preamble part and provide the second preamble part with information relating to a result of the operation. Thus, the relevant information may be directly represented in the preamble part or it may be "coded" therein.

A situation of this type is one where information in the preamble part relates to the packet (such as in a multicast packet providing information as to where copies have been sent or are to be sent) or relates to the networking element (for providing other networking elements with a status or other information).

Also, a look-up in a table or memory - based on the first preamble part or part thereof - may provide the second preamble part or information as to how to generate the second preamble part.

In an interesting embodiment, the deriving means comprise means for combining a plurality of received first preamble parts to form combined data and derive, from the combined data, one or more second preamble parts. Thus, if the preamble parts actually form part of a larger data portion or even a stream of data, this portion or stream may be assembled (such as timed, error checked or the like) prior to outputting. This outputting need not be via a normal packet or frame I/O port of the networking element by via a special purpose. output of the networking element.

One situation where this is interesting is that where the plurality of preambles each carries information from each of a number of rietworking elements -'such as congestion information, statistics information, health of network information or the like and where the networking element combines this information and sends the combined information on.

Similarly to that, the present element may comprise means for receiving additional information, to divide the additional information into a plurality of additional information parts, and to provide altered preambles of received data packets or frames by replacing at least part of each preamble part of the received data packets or frames by at least one additional information part, wherein the outputting means are adapted to output at least the altered preamble parts. Patent 20 US 2 .

This additional data could be data packets or frames, which are simply divided and transmitted as side band information in order to utilize the bandwidth on the communication medium between two networking elements. If too many frames or packets desire transmission, part of these may be divided and put into the preambles of others. In that manner, more frames or packets are transmitted on the same communication link.

The additional information may be managing information between networking elements by which a manager pushes information to or requests information from a networking element or via which the networking element provides information to the manager.

In this embodiment, the combining means could be adapted to combine preamble parts from the same port, or transmitter (such as MAC, IP) or for the same receiver (such as MAC, IP) or port lD(s).

In general, it should be noted that ports might be aggregated or virtual so that port ID may be an ID of the actual physical port of the element or an ID covering a number of ports, which are combined in a suitable manner.

Also, the first determining means could be adapted to perform the determination on the basis of the combined data - such as wherein the first determining means is adapted to determine the same output port(s) for all second preamble parts derived from a piece of combined data.

In this embodiment, the deriving means could further comprise means for dividing combined data into a number of second preamble parts and forwarding each second preamble part to an output. In this manner, the information may exit the networking element again as preamble parts

In addition, the deriving means could further comprise means for altering a piece of combined data prior to outputting it.

Preferably, the element further comprises means, for at least one of the one or more outputs, for holding or storing second preamble parts prior to output thereof. This has the advantage that preambles for a given output may be buffered independently of data parts for the same port - and may be added thereto in a given order and when data parts are Patent 20 US 2

available. If no data parts are available, pure preambles may be transmitted - if it is not desired to await the arrival of a data part at the port. An alternative to the pure preambles is dummy frames having no useful data.

Depending on the situation, the first determining means can be adapted to perform the determination on/the basis of an identity of an input port .at which the first preamble part was received and/or on the basis of information comprised in the first preamble part. Alternatively, the switching of a preamble part may be predetermined for a given type of information so that fixed routes can be predefined by e.g. an operator.

The identity of the input port may be used when a specific route through a network is desired. This may be the case when managing the network.

The preamble part may itself comprise routing information as to where to go - and the . determining means may have a look-up table quite as is known for normal Ethernet switching.

The receiver/transmitter of the actual data part may also define where to send the preamble part - even though the data and preamble parts may not be going to the same recipient. This receiver/transmitter may e.g. be seen in the MAC addresses or IP addresses - or higher-level addressing/routing information of the packet or frame. Preambles to or from a predetermined MAC/IP address may be predetermined to go to a given recipient - or to follow a given route.

It should be noted that preambles might be transmitted without data parts attached thereto. Especially within Ethernet, a Reconciliation Sublayer (see further below) will receive the preamble and may be able to derive any information there from before it is determined that the packet received does not conform to any Ethernet standard whereby it is dropped. -

However, normally it is desired that packets comprising both a preamble part and a data part be transmitted. Thus, the one or more inputs are preferably adapted to receive a first packet or frame comprising a first data part and the first preamble part, and Patent 20 US 2

the one or more outputs are preferably adapted to output a second packet or frame comprising the second preamble part and a data part of a received packet or frame.

Thus, the data packet or frame received is now taken as comprising two separate parts, which may be routed or switched independently of each other. This is especially an ^' advantage when the information in the preamble is not related to the rest of the^' packet and/or is desired to take a different route than the rest of the packet. This may be the case when the preamble relates to managing information (for managing network , elements) or an independent communication channel.

In this situation, the first determining means could be adapted to perform the determination on the basis of the first data part.

Also, the element could further comprise second means for determining from which output port to output the first data part.

Also, the element may be adapted to alter the data part, such as changing MAC addresses or tags - as is normal in Ethernet communication. This altered information may also be used for switching preamble parts further on along the communication path.

In an interesting embodiment, the first determining means and the second determining means are adapted to operate independently of each other. Independent operation may mean that one may be idle while the other one performs an analysis. Also, receiving a preamble part and a data part -from the same packet may mean (even though both may be routing the parts on the basis of the. same information - such as MAC addresses) that ^• the parts are to be output from different outputs. Also, the two determinations may base themselves on different parts of a data packet or frame when determining where to send the preamble and data parts of the packet/frame.

The second determining means may be adapted to perform the determination: on the basis of an identity of an input port at which the first data packet or frame was received, on the basis of the first preamble part, and/or on the basis of the first data part. Patent 20 US 2

Preferably, the element further comprises means for adding the second preamble part to a second data part, received as part of a second data packet or frame. In this manner, the preamble part may piggy back different data parts before and after the networking^' element. In this situation, the outputting means can be adapted to output a data packet or frame comprising the second preamble part and the second data part from the output(s) determined by the first determining means.

Preferably: the input ports are adapted to receive a first and a second data packet or frame each comprising a preamble part and a data part, and at least one of the output ports is adapted to output a third data packet or frame comprising the second preamble portion relating to a preamble portion of the first data packet or frame and the data portion of the second data packet or frame.

In accordance with a second aspect, the invention relates to a networking element adapted to receive and output data packets or frames, the element comprising: means for receiving a first and a second data packet or frame each comprising a preamble part and a data part, . means for deriving a second preamble part on the basis of the preamble part of the first packet or frame, and means for outputting a data packet or frame comprising the second preamble part and the data part of the second data packet or frame.

Again, the deriving of the second preamble part may be one where the second preamble part is identical to the first preamble part.

In a third aspect, the invention relates to a networking element adapted to receive and output data packets or frames, the element comprising: means for receiving a plurality of preamble parts, - means for combining the plurality of preamble parts into combined data, a plurality of outputs each adapted to output one or more parts of the combined data, and means for, on the basis of the combined data, determining from which of the one or more outputs to output each of the one or more parts of the combined data. Thus, all parts may be transmitted to the same output - or it may be determined to transmit part thereof to one output and others to other outputs.

Also, at least one of the outputs may be adapted to output the combined data- preferably the full combined data without dividing it. This output may be different from the outputs outputting data packets/frames.

In addition, the element may further comprise means for acting in accordance with . , instructions derived from the combined data. Such instructions could be instructions to alter a wavelength or a fibre used for the communication, new software to run, new contents for look-up tables or other manners of switching (modes turned on or off - mirroring, trunking, switching according to different OSI layers). Then, the element may also comprise means for deriving the instructions and maybe for checking the instructions (conformance to a given protocol or the like). Such means may e.g. be a processor or a CPU.

The element may also comprise means for altering or amending the combined data prior to output thereof, the outputs are preferably adapted to output the one or more parts as second preamble parts, and the receiving means are preferably adapted to receive the plurality of preamble parts as parts of data frames or packets each comprising. a preamble part and a data part, and wherein the outputs are adapted to output data packets or frames comprising a data part from a received data packet or frame and one of the one or more parts (preferably as second preamble parts).

A fourth aspect of the invention relates to a method of operating a networking element adapted to receive and output data packets or frames, the element comprising one or more inputs and a plurality' of outputs, the method comprising:

receiving a first preamble part at one of the inputs, - deriving, from the first preamble part, a second preamble part, and determining from which of the plurality of outputs to output the second preamble part.

^• As mentioned above, the deriving step could comprise deriving the second preamble part identical to the first preamble part or providing the second preamble part as a part of the Patent 20 US 2

first preamble part. In the latter situation, the first determining step could comprise basing the determination on a part of the first preamble part not comprised in the second preamble part. •

The deriving step could comprise performing a mathematical operation on information comprised in the first preamble part and providing the second preamble part with information relating to a result of the operation.

In an interesting embodiment, the deriving step comprises combining a plurality of received first preamble parts to form combined data and deriving, from the combined data, one or more second preamble parts. Then, the. combining step could comprise combining preamble parts from the same port or transmitter (MAC, IP) or for the same receiver (MAC, IP) or port. Also, the first determining step could comprise performing the determination on the basis of the combined data - such as determining the same output port(s) for all second preamble parts derived from a piece of combined data.

In the interesting embodiment, the deriving step preferably further comprises dividing combined data into a number of second preamble parts and forwarding each second preamble part to an output.

Naturally, the deriving step could further comprise altering a piece of combined data before deriving or outputting.

Suitably, the method may comprise the step of holding or storing, for at least one of the one or more outputs, second preambfe parts prior to output thereof.

The first determining mean's may be adapted to perform the determination on the basis of: an identity of an input port at which the first preamble part was received and/or information comprised in the first preamble part.

The receiving step may comprise receiving a first packet or frame comprising a first data part and the first preamble part, and further comprising a step of outputting, from one or more of the outputs, a second packet or frame comprising the second preamble part and a data part of a received packet or frame. Patent 20 US 2

In that situation, the first determining step could comprise performing the determination on the basis of the first data part, and/or a second determining step could be used of determining from which output port to output the first data part.

In the latter situation, the first determining step and the second determining step may be performed independently of each other.

Also, in the latter situation, the second determining step could comprise performing the determination on the basis of: - an identity of an input at which the first data packet or frame was received, the first preamble part, and/or the first data part.

In addition, a step could be added of adding the second preamble part to a second data part, received as part of a second data packet or frame and wherein the outputting step comprises outputting the resulting data packet of frame. In that situation, the outputting step could comprise outputting the resulting data packet or frame from the output(s) determined in the first determining step.

Finally, preferably: the receiving -step comprises receiving a first and a second data packet or frame each comprising a preamble part and a data part, and outputting, from at least one of the output ports, a third data packet or frame comprising the second preamble portion relating to a preamble portion of the first data packet or frame and the data portion of the second data packet or frame.

Another aspect of the invention relates to a method of operating a networking element adapted to receive and output data packets or frames, the method comprising: receiving a first and a second data packet or frame each comprising a preamble part and a data part, deriving a second preamble part on the basis of the preamble part of the first packet or frame, and outputting a data packet or frame comprising the second preamble part and the data part of the second data packet or frame. Patent 20 US 2

Yet another aspect of the invention relates to a method of operating a networking element adapted to receive and output data packets or frames, the method comprising: receiving a plurality of preamble parts, combining the plurality of preamble parts to combined data, - determining, on the basis of the combined data, from which of a number of outputs to output each of the one. or more parts of the combined data, and outputting one or more parts of the combined data from the output(s) determined.

This method could comprise: ^' - the step of outputting the combined data, the networking element acting in accordance with instructions derived from the combined data, the step of altering or amending the combined data prior to determination and/or output thereof, and/or - the outputting step comprises outputting the one or more parts as preamble parts, where the receiving step could then comprise receiving the preamble parts as parts of . data frames or packets each comprising a preamble part and a data part, and wherein the outputting step could comprise outputting data packets or frames comprising a data part from a received data packet or frame and one of the one or more parts (preferably as second preamble parts).

Another aspect of the invention relates to a system comprising at least three networking elements connected in series, wherein

- a first of the networking elemerits is adapted to output to a middle networking element a first data packet having a first data portion and a first preamble part comprising first data, the middle networking element is adapted to: receive the first data packet, - generate, on the basis of the first data in the first preamble part, second data, introduce the second data into a second preamble part, and forward a second data packet comprising a second data portion and the second preamble part to a third networking element, the third networking element is adapted to receive the second data packet and derive the second data from the second preamble part. Patent 20 US 2

Thus, the middle networking element may be one according to the previous aspects of the invention.

It is clear that any number (three or higher) of networking elements may exist in the system.

An especially interesting embodiment is one wherein

- the first networking element is adapted to output a first data packet having a first preamble part comprising instruction data, the middle networking element is adapted to generate, on the basis of the instruction data in the first preamble part, second instruction data, and introduce the second instruction data into the second preamble part, and - the third networking element is adapted to act in accordance with the second instruction data.

Then, the third networking element could be adapted to change a mode of operation in accordance with the second instruction data.

Another aspect of the invention relates to a method of operating a system comprising at least three networking elements connected in series, wherein

a first of the networking elements outputs to a middle networking element a first data packet having a first data portion and a first preamble part comprising first data, the middle networking element receives the first data packet, generates, on the basis of the first data in thd first preamble part, second data, introduces the second data into a second preamble part, and forwards a second data packet comprising a second data portion and the second preamble part to the third networking element, and - the third networking element receives the second data packet and derives the second data from the second preamble part.

Again: the first preamble part could comprise instruction data, Patent 20 US 2

the middle networking element could generate, on the basis of the instruction data in the first preamble part, second instruction data, introduce the second instruction data into the second preamble part, and the third networking element could act in accordance with the second instruction data.

Then, the third networking element preferably changes a mode of operation in accordance with the second instruction data.

Finally, it should be noted that, naturally, introducing a number of networking elements according to the invention into a network comprising also existing networking elements not adapted to provide or receive non-trivial information in the preambles will bring about an incompatibility problem - if not addressed.

In order to take that into account, the invention also relates to a networking element adapted to receive and output data packets or frames each comprising a preamble part and a data part, the element comprising: means for receiving or generating first data to be transmitted, disable able means for providing the first data into one or more preamble parts of one or more first data packets or frames, disable able means for generating one or more second data packets or frames each comprising a data part holding at least part of first data, means for enabling at least one of the generating and the providing means, and ^' -• - ■ means for outputting the ftfst and/or second data packets.

Thus, this element may alter between the mode where information is sent on in the preamble(s) and a mode where the information is instead sent on in the data part of packet(s) or frame(s). It should be noted that it may be possible to, in fact, provide the information both in preamble(s) and data part(s) to old equipment which can not derive information from the preambles but which has no problem with at least part of the preamble not having the trivial bit pattern.

However, the "cleanest" embodiment will be one where the enabling means is adapted to enable one of the generating means and the providing means and disable the other. Patent 20 US 2

In'that respect, a final embodiment of the invention relates to a method of operating a networking element adapted to receive and output data packets or frames each comprising a preamble part and a data part, the method comprising: - receiving or generating first data to be transmitted, performing at least one of the steps of: • o . providing the first data into one or more preamble parts of one or more first data packets or frames, o generating one or more second data packets or frames each comprising a data part holding at least part of first data, and

- ^■ outputting the first and/or second data packets.

Again, the performing step preferably comprises performing only one of the steps.

In the following, preferred embodiments of the invention will, be described with reference to the drawing wherein:

Fig. 1 illustrates a standard Ethernet frame,

Fig. 2 illustrates the frame of Fig. 1 with an altered preamble comprising additional data, .

Fig. 3 illustrates the data frame of Fig. 2 now also comprising error correction data,

Fig. 4 illustrates schematically a set-up accepting preamble data also from non- standardized packets, and

Fig. 5 is an overall block diagram of a preferred networking element according to the invention.

Fig. 1 illustrates a data frartfe in accordance with the Ethernet standard used for transmitting data for example through a network such as a local area network (LAN), metropolitan area network (MAN), wide area network (WAN) or an internetwork (Internet). Ethernet is a widely spread communication medium compatible with a wide multiplicity of types of central processing units. Further, the Ethernet is a good and cheap alternative to normal central processing unit busses such as PCI based solutions, when several physically widely distributed chips or system elements need to be controlled. Patent 20 US 2

The data frame, shown in Fig. 1 and designated in its entirety by reference numeral 50, ^' comprises several separate fields. The first field designated the header or the 'Preamble' field generally carries a7 BYTE sequence of alternating "1_." and "0" values originally used for synchronization between receiver and transmitter. The 'Preamble' field originally served to give devices in a network time to detect the presence of a signal of a new data frame on the network.

A start-of-frame delimiter designated by 'S' and following the 'Preamble' field in Fig. 1 provides one BYTE of a particular bit sequence recognised by devices in the network as a beginning of a new data frame.

A destination MAC address designated by 'DA' and following the 'S' field of the data frame shown in Fig. 1 comprises a 6 BYTE address identifying the device or devices that are to receive the data frame.

A source MAC address designated by 'SA' and following the 'DA' field of the data frame shown in Fig. 1 comprises a 6 BYTE address identifying the device originating the data frame.

The IEEE 802.3 standard permits the source and destination address fields to have a length of either 2 BYTES or 6 BYTES, but virtually all Ethernet implementations used today apply the 6 BYTE address. A destination address 'DA' may specify either an address of a single receiving device, a multicast address of a multiplicity of devices, or a broadcast address identifying, all devices in a network.

A length/type field designated by T and following the 'SA' field shown in Fig. 1 comprises 2 BYTES indicating either the length of the subsequent number of BYTES of the following MAC client data field or the nature of the MAC client protocol (protocol type). The value of the length/type field determines when the length/type field should be understood as a length of the MAC client data field and when the length/type field should be understood as a type. of the MAC client protocol.

As described above the MAC client data field designated by 'DATA' in Fig. 1 follows the ^' T field and contains the data to be transferred from the originating device to the receiving device or devices. The 'DATA' field has a maximum length of 1500 BYTES and a Patent 20 US 2

minimum length of 46 BYTES. In case the 'DATA' field is shorter than 46 BYTES padding is incorporated so as to obtain a 'DATA' field having the minimum length.

A frame check sequence designated by 'FCS' and following the 'DATA' field shown in Fig. 1 comprises a 4 BYTE cyclical redundancy check (CRC) value used for error checking. As a source device assembles a data frame, it performs a CRC calculation on all the bits in the data frame from the destination MAC address through possible padding (that is, all fields except the preamble, start-of-frame delimiter, and frame check sequence). The source device stores the calculated value in the 'FCS' field and thus transmits the value as part of the data frame. When the data frame is received at the destination device or devices, then the receiving device performs an identical calculation and checks if the received value is equal to the new calculated value. If the calculated value does not match the value in the 'FCS' field, the receiving device will assume an error has occurred

In the present context, the preamble part of the packet will be at least part of the

"preamble". Naturally, part of the preamble may be retained with the standard bit pattern. The preamble part will comprise any additional information 71 (See below) as well as any corresponding error check information 92. The addressing part(s) S, DA, SA, T (as well as any higher level addressing information such as IP addresses) will not form part of the preamble. and may be considered^' part of the data part if not denoted a routing part by itself.

It is desired that the additional information 71 , as the data portion of the frame, be protected by error protection information. Thus, as may be seen in Fig. 3, error detection data 92 may be added to the frame 70, now identified by 90. These error detection data 92 are calculated on the basis of the additional data 71 and optionally any other part(s) of the^' frames 50, 70, or 90. This calculation is predetermined so that a receiving means may perform the same calculation on the same data in order to determine whether this data has been transferred correctly. As may be seen, also these error detection data 92 are positioned in the header/preamble in order to not alter any other part of the frame 90 - including the overall length of the frame 90 compared to the frames 50 and 70.

In the presently preferred embodiment, a total of 5 bytes of additional data 71 are introduced into the header or preamble and 1 byte of error check data 92, determined as a CRC of the additional data 71 is used. Patent 20 US 2

In general, the RX/TX side of a switch or other Ethernet networking element has a so- called Reconciliation Sub layer (RS) 1 between the physical transmission medium and the . Media Access Controller (MAC) 2. This is illustrated in Fig. 4.

Normally, these two elements are combined into a single chip, but the functionality is more easily understood if they are separated. Thus, we will describe the preferred functionality with respect to these two elements.

The RS 1 is the element receiving the data packet or frame 3 from e.g. a PHY being the ^' element corresponding directly with the physical transmission medium. This packet or frame 3 will for normal communication have a preamble 4 and a data portion 5 - typically as seen in Figs. 1 -3.

The MAC may be a standard component having a well-defined operation and the RS adapts the communication between the PHY and the MAC.

The preamble information 4 of the data packet or frame 3 may be removed or derived in the RS 1 before the data packet or frame reaches the MAC 2. Different implementations are foreseen depending on the actual operation of the MAC 2. The MAC may search for a SFD or a Start-Of-Packet (used primarily in 10G Ethernet), may count bytes (no analysis of the preamble) in order to determine when the data part starts - or simply rely on the RS 1 to strip the preamble and simply provide the data part to the MAC 2.

One functionality of the MAC 2 is thethecking of the integrity of the data packet or frame, and if the packet or frame is not intact (does not correspond to a standard or if the CRC is not correct), the packet or frame may be dropped.

One example of a packet of this type will be if only a preamble (7-8 bytes) is transmitted on the communication medium, received by the PHY and derived in the RS - and forwarded to the MAC. This is not a standard Ethernet packet, and the data will be dropped in the MAC and will therefore not reach the remainder of the networking element.

However, the preamble data was received and derived and may be treated as if it was . part of any standard Ethernet packet. Patent 20 US 2

One manner of transmitting this type of information on the physical medium and separating preamble information and data packet information may be seen from the US patent application filed on 6 November 2001 by the present Assignee and with the Title: A 5 method and system for transporting information via the XGMII or Fiber Channel standard.

Fig. 5 illustrates an overall description of a preferred switch embodying the invention. The switch 10 comprises four I/O ports 12, a switching core 14 adapted to receive the data parts (including any routing or addressing information), to determine from which port to 10 output the data part and to route the data part to that port for outputting.

A preamble routing element 16 is provided for receiving the preambles of the data packets or frames received and for determining from which port(s), if any, to output the preamble. 15

Means 18 are provided for separating the preamble and data parts of the received data packets or frames and for reassembling these (or relevant parts thereof) prior to outputting.

20. As is described below, one or more of the ports 12 may be configured to send and/or accept standard preambles or preambles holding non-trivial information. Naturally, these ports and the controlling thereof (flow of information) need to take this into account.

The means 14 and 16 may perform their determinations in one of a number of manners: 25 the means 14 may determine the output port on the basis of: o the identity of the input port, o information, in the data part (primarily of the routing information in the data part), and/or 30 o information derived from the preamble part of the packet - transferred from the means 16 via arrow 22 the means 16 may determine the output port on the basis of: o the identity of the input port, o . information in the preamble part, and/or Patent 20 US 2 - o information derived from the data part of the packet - transferred from the means 14 via arrow 22.

Also, the part 16 may perform an altering of the preamble part prior to outputting. This altering may be so as to inform subsequent networking elements of a status of the present networking element or so as to inform the subsequent networking elements of a status of the data part - such as for use during a multicast of the packet for subsequent networking elements to know to which other elements to send the packet - or which to not send it to.

A sideband channel 20 is also provided for e.g. receiving a stream of data or additional data of any kind, where the means 16 divides this, if required, into smaller parts, determined from which port to output the data and transmits corresponding preambles (or data for introduction into preambles) to the corresponding means 18 which will then introduce the preambles or data in data parts received from the means 14 to that output.

The channel 20 may also operate in the opposite direction where preambles or data received by the means 16 is found to be output on the channel 20 whereby it is transmitted thereto by the means 16- optionally first assembled into larger data packets or frames - or into a stream of data.

Assembling preamble information to combined data

In one embodiment, the data of a plurality of preambles may be combined or assembled into a single piece of combined data The routing or other treatment of one or more of the preambles may then be determined on the basis of the combined data. This is an advantage in that e.g. detailed routing information may be difficult to fit into a single preamble whereas the combined data may be as large as required.

Thus, combined data may comprise data from any number of preambles.

The combined data may comprise error checking information, such as a standard CRC in order to identify and possibly correct incorrect data.

Naturally, the combined information may be altered before or after determination of the output port ID(s) and prior to any output thereof. Patent 20 US 2

Subsequent to assembly and e.g. analysis, the combined data may be re-subdivided into a number of parts, which may then be re-transmitted as preambles. There may be more or fewer parts (compared to the number of preambles originally forming the combined data).

One example where there may be fewer preambles after compared to before is an example where the combined data comprises detailed routing information part of which is used in the present networking element and which is not required further on along the path. In this situation, this information may be removed prior to re-transmittal of the preambles. This detailed routing information may be so detailed that each hop on the route is provided with information as to which next port number to transmit the data to.

Another example of the combination of a plurality of preamble parts into a single or fewer preamble parts is one where information from a number of networking elements (statistics, congestion, link/network/element health, ...) is combined and transmitted further on - such as toward a managing entity. .

The introduction of routing information is mostly feasible when all the preambles are going from the same transmitter to. the same receiver - at least part of the way.

• The grouping of preambles may also be used simply between two neighbouring networking elements for management (se below) or simply for transmitting pieces of data larger than a single preamble.--

The current networking element normally desires to know which of the received preambles to assemble into combined data. This may be all preambles from a predetermined transmitter (receiving port ID, source MAC address, source IP address or the like) or specially marked preambles (such as preambles having a predetermined byte . at a predetermined position within the preamble).

Also, the networking element will normally desire information as to what to do with the combined data. This information may be comprised within the data, such as which parts thereof to send on, where to send them (same receiver or different receivers), whether all Patent 20 US 2

or part of the information is management information (see below). Many other uses of such information may be contemplated.

In fact, the combined data may not relate to neither the data packets wherein the preambles are transmitted (if not transmitted as single preambles) nor the networking elements at all. A separate communications channel may be provided by the preambles and the combined data. Data parts, packets, cells, frames or any other assembly of data may be transmitted in the preambles and divided/combined either only at the ends of the , communication route or in one, more, or all networking elements in the communication route in order to e.g. determine the routing thereof or any other evaluation. Naturally, also a continuous stream of data may be provided into the preambles and transmitted as a separate communication channel.

In any case, the combined data may be output by an output of a networking element, either to an external recipient thereof or to e.g. a CPU for management of the networking element. Also, an input may be provided for the data - either in the form of parts adapted to be introduced into preambles, as a constant stream of data, or in larger "lumps" such as standardized packets, Ethernet packets, or any other type of packet.

The combined data may conform to a standard in order to make the analysis, generation, error check and the like easier and standardized.

How to determine where to send a received preamble

Depending on the type of information of a preamble, a variety of manners exist to determine where to send a' preamble - or a plurality of preambles. .

One manner is connected with the assembly of preambles. In this situation, the information may be derived from the assembled data. This information may be a direct port number for each hop of the route or a standard look-up may be performed in order to determine where to send the preambles. In the first situation, only part of the assembled data may be desired transmitted in that e.g. the hop information for use in the actual networking element is not useful in the next hop. Thus, the assembled data may be transmitted in fewer preambles. In other situations, a larger number of preambles are to be output. Patent 20 US 2

Naturally, it may alternatively be determined from the data part of the packet (such as from the address information of the data packet - MAC information, IP information, VLAN information, or higher OSI layers) where to send the preamble. In that manner, the preamble may be transmitted along with the remainder of the packet in which it was received. However, .this is not a requirement.

It is contemplated that preambles of packets with e.g. a given IP address (receiving or transmitting) may be preambles with a certain type of information - and that that information is to be routed to a predetermined recipient or on a predetermined route. Also, when outputting the preamble, it may be desired to await (at the relevant port) a data part having a predetermined (such as the same) address and transmit the preamble with that data part.

Also, the switching/routing information may be derived directly. from the individual preamble. The amount of available data in the preamble is limited, but e.g. a number describing the type of data may be introduced, where the networking element then switches/routes the preamble in accordance with the number. This means that the routing information is indirectly represented in the preamble. Alternatively, the routing information may be directly derivable from the preamble.

Piggy backing

A very interesting aspect of the invention is the fact that the preambles may view the data parts as mere transport mechanisms. The preamble can piggy back any data part as long_, as it is "heading in the right direction" - from hop to hop.

Depending on the type of information in the preamble, the preamble may be totally independent on the data part. The only thing these two parts could have in common is that they are transported on the same link and that they together form a data unit conforming to a standard acknowledged by the link partners of the link. In fact, the preamble and data part may, even though they were received as part of the same packet cell/frame, be output by a networking element on the same output link - but now taking part in two different data units. . Patent 20 US 2

When independent of each other, the preambles and data parts can be stored or buffered independently for output at the relevant port.

Types of data or information which could be transmitted in preambles

In general, the information in- a preamble may or may not relate to the contents of the data part of the same data unit (data packet, frame, cell).

If the preamble information relates to the data part, preferably these are also output as a single data unit. In this situation, the routing or switching of the data packet in the networking element may take place on the basis of the information of the preamble. In- addition, or alternatively, the information of the preamble may relate to other features of the data part, such as to a priority of the data part.

If the preamble information does not relate to the data part, it may relate to one or more of the networking elements of the network - or the network itself. In this situation, the preamble information may be, congestion information from a given networking element, or management software used to manage the network.

Management information may be information from a manager to a networking element as described below.

Another alternative is one where the preamble information is independent of the data parts with which it is transmitted as well as the network and networking elements. This data may follow an independent data flow between two end points, such as a measuring element and a data collector or a separate data communication between two computers.

In fact, this data flow may enter a networking element (initiating the data transport) at an input different from the ones receiving data packets or data parts, and the networking element finally deriving the information may output this information from a special-purpose output (that is, an output different from the outputs outputting the data parts or data packets).

Finally, the preambles may in themselves not represent anything logical but only when assembled into combined data - or the preambles mav constitute a constant flow of data. Patent 20 US 2

When the preambles are to be assembled into combined data, preferably this data acknowledges a standard or protocol in order to facilitate standardization, error check etc.

Network management using the present invention

As mentioned earlier, the preamble information may be used for managing the network.

Management of a network normally comprises a manager instructing networking elements to forward information (congestion information, statistics, memory contents) or the manager sending new information to the networking element (addressing/routing information). Also, the manager may update the software of the networking element using the preamble information.

Naturally, combined data may be used in this connection.

It should be noted that management information might relate to a plurality of networking elements in the network. Thus, a networking element receiving such information may itself act thereon and maybe alter the information before transmitting it to the next networking element. This altering of the information may be in accordance with a predetermined mathematicai equation.

Another situation where the preamble information is altered prior to outputting may be one where a change in information protocol is desired. Thus, different networking elements may interchange information using different protocols. One suitable protocol is the DSCP protocol.

Also, a networking element having received management information may act in accordance with this information - such as change a mode of operation in accordance therewith. A change in mode of operation may be to change a wavelength or fibre used for communication, switch/route packages differently (using different OSI levels or on the basis of an altered look-up table), or the like.

Naturally, an assembly of networking elements may be managed or controlled using this information. Thus, the networking element finally receiving the management information Patent 20 US 2

may output this from a special-purpose output (different from those outputting data parts or data packets) such as to a central managing CPU or system controlling the networking element(s).

The deriving of preamble information from other preamble information

As has been mentioned above, one situation where alternation/changing of preamble information is desired is one where the information to be output has to have a different . protocol than that received. Protocol conversions are known manipulations. ^'

Another situation is one where part of the information received is not to be output. This may be encountered when the information received comprises switching/routing information for use in more than one networking element. Another situation is one where part of the management information is for use with this networking element and the remaining part for subsequent networking elements.

As mentioned above, also the collecting (from the plurality of preamble parts) of information (such as statistics, congestion, network/link/element health) and the providing thereof into preamble parts is a candidate for the providing of one or more preambles from a plurality of preambles.

Finally, if the data part is to be transmitted to a plurality of receivers (multicast or broadcast), the preamble data..may relate to where to send the data part - or where it has been sent to. Thus, when forwarding he data part to the next hop, an updating of the preamble (routing information) is desired.

Dual operation of input/output ports

In order to be backwards compatible to also networking elements not able to derive information from preambles, the individual ports of the present networking element are preferably adapted to run in two modes of operation:

a first mode where preamble information is transmitted and received in accordance with the invention and Patent 20 US 2

a second mode where the information which would otherwise be transmitted in preambles are transmitted using other manners - such as in data packets, control packages or any other manner used hitherto. Naturally, standard preambles may be used - or preambles comprising non-trivial information - but where it is known that that information is not derived at the receiver.

The controlling of how to transmit the data or information may be defined (such as per port) by hardware or software - and software may override a hardware-defined set-up.

Normally, network managing information is transmitted in normal data packets where e.g. a type length field may indicate the type of information. Removing the information from one or more preambles (such as combining the information from a plurality of preambles and providing the combined information in the data part) and generating the preamble part is easy for the skilled person. Also, the receiving of such a data packet and generating one or more preambles is easy once the use of preambles and the switching thereof has been realized.

If the information in the preambles stems from subdivided packets or frames, the recombination thereof and transmission over the network is simple.

If the information in the preambles relates to information independent of the data packets, the network or network elements (such as if it is an additional communication channel), again, providing data packets having the information in the data part (from one preamble or a plurality of preambles) is again simple in that that would be the manner in which that information would have been transmitted hitherto.

The determination of which mode of operation to operate in may be determined by an operator or by a communication between the link partners (for each port) or may be made on the basis of a discovery protocol where the link partners exchange information as to their abilities.

Claims

Patent 20 US 2CLAIMS

1. A networking element adapted to receive and output data packets or frames, the element comprising:

2. An element according to claim 1 , wherein the deriving means are adapted to derive the. second preamble part to be identical to the first preamble part.

3. An element according to claim 1 , wherein the deriving means are adapted to provide the second preamble part as a part of the first preamble part

4. An element according to claim 3, wherein the first determining means are adapted to base the determination on a part of the first preamble part not being comprised in the second preamble part.

5. An element according to claim 1 , wherein the deriving means are adapted to perform a mathematical operation on information comprised in the first preamble part and provide the second preamble part with- information relating to a result of the operation.

6. An element according to claim 1 , wherein the deriving means comprise means for combining a plurality of redeived first preamble parts to form combined data and derive, from the combined data, one or more second preamble parts.

7. An element according to claim 6, wherein the combining means is adapted to combine preamble parts from the same port or transmitter or for the same receiver or port.

8. An element according to claim 6, wherein the first determining means is adapted to perform the determination on the basis of the combined data. Patent 20 US 2

9. An element according to claim 8, wherein the first determining means is adapted to determine the same output port(s) for all second preamble parts derived from a piece of combined data.

10. An element according to claim 6, wherein the deriving means further comprise means for altering a piece of combined data.

1 1. An element according to claim 6, wherein the deriving means further comprise means for dividing combined data into a number of second preamble parts and forwarding each second preamble part to an output.

12. An element according to claim 1 , further comprising means, for at least one of the one or more outputs, for holding or storing second preamble parts prior to output thereof.

13. An element according to claim 1 , wherein the first determining means is adapted to perform the determination on the basis of an identity of an input port at which the first preamble part was received.

14. An element according to claim 1 , wherein the first determining means is adapted to perform the determination on. the basis of information comprised in the first preamble part.

15. An element according to claim 1 , wherein

the one or more inputs-are adapted to receive a first packet or frame comprisinα a first data part and the first preamble part, and the one or more outputs are adapted to output a second packet or frame comprising the second preamble part and a data part of a received packet or frame.

16. An elerήent according to claim 15, wherein the first determining means is adapted to perform the determination on the basis of the first data part.

17. An element according to claim 15, further comprising second means for determining from which output port to output the first data part. Patent 20 US 2

18; An element according to claim 17, wherein the first determining means and the second determining means are adapted to operate independently of each other.

19. An element according to claim 17, wherein the second determining means is adapted to perform the determination on the basis of an identity of an input port at which the first data packet or frame was received.

20. An element according to claim 17, wherein the second determining means is adapted to perform the determination on the basis of the first preamble part.

21. An element according to claim 17, wherein the second determining means is adapted to perform the determination on the basis of the first data part.

22. An element according to claim 15, further comprising means for adding the second preamble part to a second data part, received as part of a second data packet or frame.

23. An element according to claim 22, wherein the outputting means are adapted to output a data packet or frame comprising the second preamble part and the second data part from the output(s) determined by the first determining, means.

24. An element according to claim 1 , wherein: the input ports are adapted to receive a first and a second data packet or frame each comprising a preamble part and a data part, and- at least one of the output ports is adapted to output a third data packet or frame comprising the second preamble porfron relating to a preamble portion of the first data packet or frame and the data portion of the second data packet or frame.

25. A networking element adapted to receive and output data packets or frames, the element comprising: - . means for receiving a first and a second data packet or frame each comprising a preamble part and a data part, means for deriving a second preamble part on the basis of the preamble part of the first packet or frame, and means for outputting a data packet or frame comprising the second preamble part and the data part of the second data packet or frame. Patent 20 US 2

26. A networking element adapted to receive and output data packets or frames, the element comprising: means for receiving a plurality of preamble parts, 5 - means for combining the plurality of preamble parts into combined data, a plurality of outputs each adapted to output one, or more parts of the combined^' ^•data, and^' means for, on the basis of the combined data, determining which of the one or more outputs to output each of the one or more parts of the combined data. 10

27. An element according to claim 26, wherein at least one of the outputs is adapted to , output the combined data.

28. An element according to claim 26, further comprising means for acting in accordance 15 with instructions derived from the combined data.

29. An element according to claim 26, further comprising means for altering or amending the combined data prior to output thereof.

20 30. An element according to claim 26, wherein the outputs are adapted to output the one or more parts as second preamble parts.

31. An element according to claim 30, wherein the receiving means are adapted to receive the plurality of preamble parts as parts of data frames or packets each comprising

"v.

25 a preamble part and a data part, and "wherein the outputs are adapted to output data packets or frames comprising a data part from a received data packet or frame and one of the one or more parts.

32. A method of operating a networking element adapted to receive and output data

30 packets or frames, the element comprising one or more inputs and a plurality of outputs, the method comprising:

receiving a first preamble part at one of the. inputs, . deriving, from the first preamble part, a second preamble part, and Patent 20 US 2

determining from which of the plurality of outputs to output the second preamble part.

33. A method according to claim 32, wherein the deriving step comprises deriving the 5 second preamble part identical to the first preamble part.

34. A method according to claim 32, wherein the deriving step comprises providing the second preamble part as a part of the first preamble part

10 35. A method according to claim 34, wherein the first determining step comprises basing the determination on a part of the first preamble part not comprised in the second preamble part. ^' '

36. A method according to claim 32, wherein the deriving step comprises performing a

15 mathematical operation on information comprised in the first preamble part and providing the second preamble part with information relating to a result of the operation.

37. A method according to claim 32, wherein the deriving step comprises combining a plurality of received first preamble parts to form combined data and deriving, from the

20 combined data, one or more second preamble parts.

38. A method according to claim 37, wherein the combining step comprises combining preamble parts from the same port, transmitter (MAC, IP) or for the same receiver (MAC, IP) or port.

25

39. A method according to claim 37, wherein the first determining step comprises performing the determination on the basis of the combined data.

40. A method according to claim 39, wherein the first determining comprises determining 30 the same output port(s) for all second preamble parts derived from a piece of combined data.

41. A method according to claim 37, wherein the deriving step further comprises altering a piece of combined data. Patent 20 US 2

42. A method according to claim 37, wherein the deriving step further comprises dividing combined data into a number of second preamble parts and forwarding each second preamble part to an output.

5 43. A method according to claim 32, further comprising the step of holding or storing, for at least one of the one or more outputs, second preamble parts prior to output thereof.

44. A method according to claim 32, wherein the first determining means is adapted to perform the determination on the basis of an identity of an input port at which the first

10 preamble part was received.

45. A method according to claim 32, wherein the first determining step comprises performing the determination on the basis of information comprised in the first preamble part.

15

46. A method according to claim 32, wherein the receiving step comprises receiving a first packet or frame comprising a first data part and the first preamble part, and further comprising a step of outputting, from one or more of the outputs, a second packet or frame comprising the second preamble part and a data part of a received packet or frame.

20

47. A method according to claim 46, wherein the first determining step comprises performing the determination on the basis of the first data part.

48. A method according to claim 46, further comprising a second determining step of 25 determining from which output port to^outpύt the first data part.

49. A method according to-^'claim 48, wherein the first determining step and the second determining step are performed independently of each other.

30 50. A method according to claim 48, wherein the second determining step comprises performing the determination on the basis of an identity of an input port at which the first data packet or frame was received.

51. A method according to claim 48, wherein the second determining step comprises 35 performing the determination on the basis of the first preamble part. Patent 20 US 2

52. A method according to claim 48, wherein the second determining step comprises performing the determination on the basis of the first data part.

53. A method according to claim 46, further comprising a step of adding the second preamble part to a second data part, received as part of a second data packet or frame and wherein the outputting step comprises outputting the resulting data packet of frame.

54. A method according to claim 53, wherein the outputting step comprises outputting the resulting data packet or frame from the output(s) determined in the first determining step.

55. A method according to claim 32, wherein: the receiving step comprises receiving a first and a second data packet or frame each comprising a preamble part and a data part, and - outputting, from at least one of the output ports, a third data packet or frame comprising the second preamble portion relating to a preamble portion of the first data ■ packet or frame and the data portion of the second data packet or frame.

56. A method of operating a networking element adapted to receive and output data packets or frames, the method comprising: receiving a first and a second data packet or frame each comprising a preamble part and a data part, deriving a second preamble part on the basis of the preamble part of the first packet or frame, and - outputting a data packet or f name comprising the second preamble part and the data part of the second data packet or frame.

57. A method of operating, a networking element adapted to receive and output data packets or frames, the method comprising: - receiving a plurality of preamble parts, combining the plurality of preamble parts to combined data,

- ^' determining, on the basis of the combined data, from which of a number of outputs to output each of the one or more parts of the combined data, and.

- outputting one or more parts of the combined data from the output(s) determined. . Patent 20 US 2 .

58. A method according to claim 57, further comprising the step of outputting the combined data.

59. A method according to claim 57, further comprising the networking element acting in 5 accordance with instructions derived from the combined data.

60. A method according to claim 57, further comprising the step of altering or amending ^■ the combined data prior to output thereof.

10 61. A method according to claim 57, wherein the outputting step comprises outputting the one or more parts as preamble parts.

62. A method according to claim 61 , wherein the receiving step comprises receiving the preamble parts as parts of data frames or packets each comprising a preamble part and a

15 data part, and wherein the outputting step comprises outputting data packets or frames comprising a data part from a received data packet or frame and one of the one or more parts.

63. A system comprising at least three networking elements connected in series, wherein 20

. a first of the networking elements is adapted to output to a middle networking element a first data packet having a first data portion and a first preamble part comprising first data, the middle networking element is adapted to: 5 - receive the first data packet, generate, on the basis of the first data in the first preamble part, second data, - ^• introduce the second data into a second preamble part, and forward a second data packet comprising a second data portion and the second preamble part to a third networking element, and 30 - ^• the third networking element is adapted to receive the second data packet and derive the second data from the second preamble part.

64. A system according to claim 63, wherein the first networking element is adapted to output a first data packet having a first 5 preamble part comprising instruction data, . Patent 20 US 2 .

the middje networking element is adapted to generate, on the basis of the instruction data in the first preamble part, second instruction data, and introduce the second instruction data into the second preamble part, and the third networking element is adapted to act in accordance with the second 5 instruction data.

.

65. A system according to claim 64, wherein the third networking element is adapted to change a mode of operation in accordance with the second instruction data.

10 66. A method of operating a system comprising at least three networking elements connected in series, wherein

a first of the networking elements outputs to a middle networking element a first data packet having a first data portion and a first preamble part comprising first data, 15 - the middle networking element receives the first data packet, generates, on the basis of the first data in the first preamble part, second data, introduces the second data into a second preamble part, and forwards a second data packet comprising a second data portion and the second preamble part to the third networking element, and the third networking element receives the second data packet and derives the 20 second data from the second preamble part.

67. A method according to claim 66, wherein

the first preamble part comprises instruction data, 25 - the middle networking element generates, on the basis of the instruction data in the first preamble part, second instruction data, introduces the second instruction data into the second preamble part, and the third networking element acts in accordance with the second instruction data.

30 68. A method according to claim 67, wherein the third networking element changes a mode of operation in accordance with the second instruction data.

69. A networking element adapted to receive and output data packets or frames each comprising a preamble part and a data part, the element comprising: 35 - means for receiving or generating first data to be transmitted, Patent 20 US 2

-. . disable able means for providing the first data into one or more preamble parts of one or more first data packets or frames,

. disable able means for generating one or more second data packets or frames each comprising a data part holding at least part of first data, - means for enabling at least one of the generating and the providing means, and means for outputting the first and/or second data packets.

70. A networking element according to claim 69, wherein the enabling means is adapted to enable one of the generating means and the providing means and disable the other.

71. A method of operating a networking element adapted to receive and output data packets or frames each comprising a preamble part and a data part, the method comprising: - receiving or generating first data to be transmitted, performing at least one of the steps of: o providing the first data into one or more preamble parts of one or more first data packets or frames, o generating one or more second data packets or frames each comprising a ^' _. data part holding at least part of first data, and

- ^' outputting the first and/or second data packets.

72. A method according to claim 71 , wherein the performing step comprises performing only one of the steps.