Packet data flow identification for multiplexing

[0001] The present invention is directed to the process of multiplexing packet data flows together into a single aggregate flow. In particular, the present application provides for the identification of individual flows suitable for aggregation into a single, combined flow.

[0002] The concept of multiplexing packet data flows, including realtime transport protocol (RTP) flows together into a single aggregate flow is a concept that is addressed in a number of Internet engineering task force (IETF) Internet drafts. Multiplexing RTP flows would be desirable, because it can reduce the bandwidth usage of wide area network (WAN) links and decrease the RTP packet rate incident on edge routers. In addition, the multiplexing of RTP flows reduces the burden of administering networks, for example in connection with the configuration of quality of service levels. In addition, multiplexing RTP flows can increase the scalability of existing quality of service schemes.

[0003] Although IETF Internet drafts specifying packet formats for aggregate flows, header compression algorithms, and packet loss compensation have been proposed, there are no methods discussed for signaling to a multiplexer which RTP flows to aggregate into a single flow. Although a multiplexer could examine individual RTP flows to determine which packets are candidates for aggregation, doing so would be a significant burden on the multiplexer. In particular, an RTP multiplexer in such a system would be required to examine the destination addresses of individual packets to determine whether they could be aggregated with other packets for transport to another multiplexing device for deaggregation back into individual flows or packets. In addition, requiring a multiplexer to consider the destination addresses of individual packets to determine whether aggregation would be beneficial introduces additional latency into data streams. The addition of latency is particularly undesirable in connection with voice over Internet protocol (VoIP) sessions and may require the addition of costly hardware to the multiplexing device to mitigate or avoid. Furthermore, where explicit information regarding the flows that should be aggregated is not provided to a multiplexer, indeterminancy regarding which flows are aggregated is introduced into the system. This indeterminancy can contribute to delays while the multiplexing devices wait for expected packets that never arrive.

[0004] The present invention is directed to solving these and other problems and disadvantages of the prior art. Generally, according to the present invention, an identifier is assigned to a remote subnetwork having an associated multiplexer. The assigned identifier is communicated to communication devices and to a multiplexer associated with a first subnetwork. Packets originating at communication devices on the first subnetwork that are addressed to endpoints associated with a remote subnetwork for which an identifier has been assigned are associated with that identifier, and provided to the multiplexer on the first subnetwork. The multiplexer then bundles together or aggregates packets that are part of individual communication streams and that are associated with the same identifier, to form a super packet or aggregate stream. The aggregate stream is then passed to a second multiplexer associated with the remote subnetwork. At the second multiplexer, the aggregate stream is unbundled, and packets associated with individual streams are delivered to the appropriate endpoint or destination communication device.

[0005] In accordance with an embodiment of the present invention, the identifier is associated with an individual data packet or data stream using a pre-existing field. For example, a virtual local area network (VLAN) field that is typically used for the purpose of distinguishing between virtual local area networks (LANs) on the same physical LAN and resides in the Media Access Control (MAC) header is used to identify a multiplexer on a remote subnetwork. In accordance with another embodiment of the present invention, a differentiated services field (DS) that is normally used to define a type of service requirement is used to communicate the identifier. This field is also known as the type of service field (TOS) and resides in the Internet Protocol (IP) header. In accordance with yet another embodiment of the present invention, a resource reservation protocol (RSVP) field that is normally used to reserve resources on routers to ensure adequate quality of service for the bearer channel of a VoIP call over a WAN link is used to communicate the identifier. In accordance with still another embodiment of the present invention, the RSVP DCLASS object is used to communicate a differentiated services code point (DSCP) that is placed in the differentiated services field and used to identify individual flows that may be aggregated into a combined flow. In accordance with still other embodiments of the present invention, individual flows that may be aggregated are identified using information associated with the session initiation protocol (SIP) or H.323 signals to identify the parameters of the channel being negotiated to determine whether flows can be aggregated.

[0006] In accordance with an embodiment of the present invention, a system that provides a multiplexer as part of or in cooperation with a subnetwork router is provided. In addition, a control unit may be provided for assigning an identifier to a remote subnetwork. A system in accordance with the present invention may also include a second multiplexer associated with or implemented as a part of a second edge router associated with a remote or second subnetwork. The first and second edge routers may be interconnected to one another by a wide area network (WAN), which may include the Internet.

[0007] These and other advantages and features of the invention will become more apparent from the following discussion, particularly when taken together with the accompanying drawings.

[0008] FIG. 1 is a block diagram depicting a communication system in accordance with an embodiment of the present invention;

[0009] FIG. 2 is a diagram depicting packet flow through a subnetwork in accordance with an embodiment of the present invention;

[0010] FIG. 3 is a flow chart illustrating the aggregation of data packets in accordance with an embodiment of the present invention;

[0011] FIG. 4 is a flow chart depicting the identification of packets suitable for aggregation in accordance with an embodiment of the present invention;

[0012] FIG. 5 is a flow chart depicting the identification of packets suitable for aggregation in accordance with another embodiment of the present invention;

[0013] FIG. 6 is a flow chart depicting the identification of packets suitable for aggregation in accordance with another embodiment of the present invention;

[0014] FIG. 7 is a flow chart depicting the identification of packets suitable for aggregation in accordance with another embodiment of the present invention; and

[0015] FIG. 8 is a flow chart depicting the identification of packets suitable for aggregation in accordance with another embodiment of the present invention.

[0016] The present invention is directed to the identification of data packets suitable for aggregation into a combined flow.

[0017] In FIG. 1, a communication system 100 in accordance with an embodiment of the present invention is illustrated. In general, the communication system 100 includes a first subnetwork 104 and a second or remote subnetwork 108. The first subnetwork 104 is interconnected to the second subnetwork 108 by a communication network 112. The communication network 112 generally comprises a packet data network or combination of packet data and switched circuit networks. As an example, the communication network 112 may comprise the Internet, an intranet, or a combination of one or more intranets and the Internet.

[0018] The first subnetwork 104 generally includes a plurality of communication devices 116. The communication devices 116 are each interconnected to a multiplexer 120. A control unit 124 may be provided for coordinating activity between or among the communication devices 116 and the multiplexer 120. The first subnetwork 104 additionally includes an edge router 128 that interconnects the first subnetwork 104 to the communication network 112.

[0019] Like the first subnetwork 104, the second subnetwork 108 may include a plurality of communication devices 132, a multiplexer 136, a control 140, and an edge router 144.

[0020] In accordance with an embodiment of the present invention, the subnetworks 104, 108 may comprise local area networks (LANs). Accordingly, one or both subnetworks 104, 108 may be implemented as Ethernet networks.

[0021] The communication devices 116, 132 associated with the system 100 may include any device suitable for sending packet data across a communication network or subnetwork. Accordingly, the communication devices 116, 132 may comprise Internet protocol telephones, general purpose computers, personal digital assistants (PDAs) or other computing devices configured to function in connection with media communications, including communications that occur in realtime. The multiplexers 120, 136 generally function to aggregate media or data streams into bundled streams comprising super packets containing data packets that may originate from a number of communication devices (e.g., communication devices 116) associated with the first subnetwork 104 and that are addressed to one or more communication devices (e.g., communication devices 132) associated with a remote subnetwork 108. In connection with the receipt of aggregated data streams, a multiplexer (e.g., the second multiplexer 136) unbundles the aggregated streams to allow delivery of individual streams to recipient communication devices 132. As can be appreciated, the above described functions are generally reversed in connection with data packets originating at or through communication devices 132 associated with the second subnetwork 108 that are transmitted to communication devices 116 associated with the first subnetwork 104. In accordance with embodiments of the present invention, the multiplexers 120, 136 are implemented as general purpose or server computers running software implementing the bundling and unbundling functions. In alternative embodiments, hard-wired circuitry may be used in place of or in combination with software instructions to implement a multiplexer 120, 136. In accordance with another embodiment of the present invention, the multiplexers 120, 136 may be implemented as a part of the respective edge routers 128 or 144. In accordance with yet another embodiment the multiplexers 120, 136 may be implemented as a specialized device employing appropriate specialized hardware and configured as the network gateway for other VoIP devices to provide transparent access to those devices' data streams (e.g., RTP streams) if necessary.

[0022] The controls 124, 140 generally function to coordinate activity between the communication devices 116, 132 and the associated multiplexer 120 or 136. For example, the first control 124 may assign an identifier for use in connection with packets originating at or through communication devices 116 associated with the first subnetwork 104 that are addressed to communication devices 132 associated with the second subnetwork 108. The controls 124, 140 may be implemented as general purpose computers. In accordance with a further embodiment of the present invention, each control 124, 140 may be implemented as part of the corresponding multiplexer 120, 136 and/or edge router 128, 144.

[0023] The edge routers 128, 144 generally function to interconnect the respective subnetwork 104 or 108 to a communication network 112. In addition, an edge router 128, 144 in accordance with an embodiment of the present invention may incorporate the multiplexer 120, 136 and/or control 124, 140 of the respective subnetwork 104, 108.

[0024] With reference now to FIG. 2, the composition of data packets transmitted in accordance with an embodiment of the present invention are illustrated. In general, each of the communication devices 116, 132 is capable of transmitting (or receiving) data packets that include various fields. In particular, each of the data packets or individual data streams 204 includes various header information fields 208, such as the medium access control, Internet protocol, universal datagram protocol (UDP), realtime transport protocol (RTP), and medium access control cyclic redundancy check (MAC CRC) fields. In addition, each of the data packets 204 generally contains a payload 212, such as a G.729 payload. Furthermore, in accordance with the present invention, each of the data packets 204 includes a destination subnetwork Id 216. As will be described in greater detail below, the destination subnetwork Id 216 may be included within another of the fields, including any of the header information fields 208. The individual data packets 204 are provided to (or received from) the multiplexer 120, 136 associated with the subnetwork 104, 108.

[0025] The multiplexer 120, 136 aggregates the individual data packets 204 together into one or more super packets or aggregate data streams 220. As can be seen from FIG. 2, the super packet 220 includes its own header information. In addition, the super packet 220 includes the header information 208 and payloads 212 from the individual data packets 204. The super packet 220 is provided to (or received from) the edge router 128, 136.

[0026] As can be appreciated by one of skill in the art, the aggregation of individual data packets 204 into a super packet 220 generally reduces the burden on the edge router 128, 136. In particular, by aggregating multiple individual packets 204 into a lesser number of super packets 220, the number of individual packets that must be routed by the edge router 128, 136 is reduced. Furthermore, because it is the ability of a router 128, 136 to classify and route data packets, and not the bandwidth of the router's interfaces 128, 136 that typically limits a router's effective bandwidth, the present invention can increase the data transmission capabilities of a network or subnetwork 104, 108 carrying data streams having a large number of small packets.

[0027] With reference now to FIG. 3, a flow chart depicting the operation of a system 100 in accordance with an embodiment of the present invention is shown. Initially, at step 300, an identifier 216 is assigned to a remote subnetwork 108. At step 304a, voice data addressed to a first communication device 132a on the remote subnetwork 108 is received at a first communication device 116a on the first subnetwork 104. Similarly, at step 304b, voice data addressed to a second communication device 132b on the remote subnetwork 108 is received at a second communication device 116b on the first subnetwork 104. In addition, voice data may be received at an n^th communications device 116n that is addressed to an n^th communication device 132n on the remote subnetwork 108 (step 304n). At steps 308a-308n, the received data is packetized by the respective communication device 116a-116n. At steps 312a-312n, the assigned identifier 216 for the remote subnetwork 108 is associated with the packetized data by the respective communication device 116a-116n, and the communication devices 116a-116n send their data packets 204 to the multiplexer 120.

[0028] At step 316, the multiplexer 120 on the first subnetwork 104 bundles the packets 204 associated with the same identifier 216 together with one or more super packets or flows 220, and sends them to the remote subnetwork 108 across the communication network 112. As can be appreciated by one of skill in the art, by sending the various data packets 204 received from the communication devices 116a-116n as a single super packet 220, the processing load placed on the edge routers 128, 124 and the switches and routers associated with the communication network 112 is lessened. For instance, in the example illustrated in FIG. 3, instead of the three individual packets 204, only the header information associated with the single super packet must be considered by the routers and switches.

[0029] At step 320, the multiplexer 136 on the remote subnetwork 108 receives the super packet 220, unbundles the aggregate flow, and sends the individual flows (i.e., the individual packets 204) to the destination communication devices 132a-132n. At steps 324a-324n, the packetized voice data is received at the appropriate communication device 132a-132n on the remote subnetwork 108.

[0030] With reference now to FIG. 4, the operation of a system 100 in accordance with an embodiment of the present invention is illustrated. In general, the embodiment illustrated in connection with FIG. 4 utilizes the virtual local area network (VLAN) identification field provided in connection with data packets 204 transmitted across the subnetwork 104 or 108 to associate or identify individual packets 204 with a multiplexer 136 at a remote subnetwork 108. More particularly, the VLAN Id field in the MAC header is used to identify a data packet 204 as being addressed to a communication device 132 associated with a remote subnetwork 108 having a multiplexer 136 in accordance with the present invention. As can further be appreciated, the assigned VLAN Id is selected so that it does not correspond to a valid virtual local area network established in connection with the first 104 or second 108 subnetworks.

[0031] Initially, at step 400, a VLAN Id is assigned in connection with a remote subnetwork 108. In particular, the first control 124 may function to select a VLAN Id to identify a remote subnetwork 108 having a multiplexer 136 in accordance with the present invention. Furthermore, the control 124 communicates the assigned identifier to each of the communication devices 116 on the first subnetwork 104, so that the communication devices 116 may utilize the identifier when they have data packets 204 addressed to communication devices 132 on the remote subnetwork 108.

[0032] At step 404, the first multiplexer 120 receives data packets 204 addressed to communication devices 132 on the remote subnetwork 108 that are marked with the assigned VLAN Id associated with the remote subnetwork 108. The mutliplexer 120 aggregates packets 204 having the same VLAN Id together into one or more super packets 220 (step 408).

[0033] The first multiplexer 120 then sends the super packet 220 to the second multiplexer 136 on the remote subnetwork 108. The second multiplexer 136 deaggregates the super packet 220, and sends the individual packets 204 to the appropriate communication devices 132 on the remote subnetwork 108 (step 416) as individual streams.

[0034] As can be appreciated from the description given herein in connection with FIG. 4, the transmitting multiplexer 120 is not required to consider whether the destination addresses of the individual packets 204 make those packets 204 suitable for aggregation into a super packet 220. Instead, the multiplexer 120 is only required to consider the relatively short VLAN Id field within each packet 204 to determine whether it contains an identifier associated with a second multiplexer 136 on a remote subnetwork 108. Furthermore, it can be appreciated that an embodiment in accordance with that described in connection with FIG. 4 does not require the definition of new data packet 204 information or header fields. Instead, a preexisting field is utilized. As can also be appreciated, the super packet 220 is sent to an address associated with the second multiplexer 136. Upon receiving a super packet 220 addressed directly to it, the second multiplexer 136 then unbundles the individual data streams 204 of the super packet 220 and delivers the individual packets 204 as noted above.

[0035] With reference now to FIG. 5, the operation of a system 100 in accordance with another embodiment of the present invention is illustrated. In general, the embodiment illustrated in FIG. 5 utilizes the differentiated services field (DS) or Type of Service (TOS) field in the IP Header in order to identify individual data streams 204 that are addressed to communication devices 132 on a remote subnetwork 108 having a multiplexer 136 in accordance with the present invention.

[0036] At step 500, a Diffserv Code Point (DSCP) is assigned to the remote subnetwork 108. At step 508, the first multiplexer 120 receives data packets 204 addressed to communication devices 132 on the remote subnetwork 108 and marked with the DSCP assigned to that remote subnetwork 108. The first multiplexer 120 then aggregates the individual data packets or flows 204 having a common DSCP into one or more super packets or aggregate flows 220 addressed to the second multiplexer 136 (step 512). The super packet 220 is then sent to the second multiplexer 136 on the remote subnetwork 108 (step 516). The multiplexer 136 on the remote subnetwork 108 deaggregates the super packet 220, and sends the individual packets 204 comprising individual data streams to the appropriate communication devices 132 on the remote subnetwork 108 (step 520).

[0037] From the description given herein with respect to FIG. 5, it can be appreciated that an embodiment utilizing the DS field allows a multiplexer 120 to identify data packets 204 suitable for aggregation into a super packet 220 simply by considering the relatively short DSCP field associated with such data packets. Upon identifying multiple data packets 204 associated with a DSCP that has been assigned to a remote subnetwork 108, the multiplexer 120 bundles the individual packets 204 together, and addresses the resulting super packet 220 to the multiplexer 136 associated with the remote subnetwork 108. The multiplexer can place an appropriate DSCP value in the DS field of the super packet to nominate the required per hop behavior for the WAN portion of the communications network.

[0038] With reference now to FIG. 6, the operation of a system 100 in accordance with yet another embodiment of the present invention is illustrated. In general, the embodiment described with reference to FIG. 6 utilizes the resource reservation protocol (RSVP) protocol to identify the source and destination address and port associated with individual data packets 204 in order to determine what RTP data flows can be expected. Individual data packets 204 that are part of such flows and that are addressed to communication devices 132 on the same subnetwork 108 are then aggregated into one or more super packets 220.

[0039] As an example of the operation of such an embodiment, an RSVP enabled VOIP communication device 116 is configured to use the first multiplexer 120 on the first subnetwork 104 as its default IP gateway (step 600). At step 604, a voice over Internet protocol session or other communication session is initiated. At step 608, an RSVP request (PATH message) to provide a certain quality of service for the RTP stream is sent to the multiplexer 120 by the communication device 116.

[0040] The multiplexer 120 determines the destination of the individual flow or RTP stream that will be received from the communication device 116, and assigns that flow to a new or existing aggregate flow (step 612). At step 616, a return reservation message (RESV message) to the communication device 116 is emulated by the multiplexer 120 and delivered to the communication device 116. The communication device 116 then sends individual packets 204 to the multiplexer 120. The individual packets 204 are received at the multiplexer 120, and are placed in a super packet 220 for delivery to the remote subnetwork 108 under a single RSVP reservation (step 620) that is maintained between the multiplexer and the remote multiplexer. The parameters of the RSVP reservation may be changed to accommodate the different requirements of the aggregate flows. The super packet 220 is received at the multiplexer 136 associated with the remote subnetwork 108, which deaggregates the super packet 220, and sends the individual packets 204 to the appropriate communication device or devices 132 on the remote subnetwork 108 (step 624).

[0041] As can be appreciated from the description set forth herein, an embodiment of the present invention utilizing the RSVP field allows the multiplexer 120 to know what VoIP flows to expect from the communication devices 116. However, the multiplexer 120 must examine the destination address of at least one individual packet 204 within an identified flow to determine the destination address of that flow. Therefore, the destination address serves as the initial identifier of individual data packets 204. RSVP information may then be used to identify packets suitable for aggregation. The information provided as part of the RSVP request also communicates when the individual flows begin and end, so that the multiplexer 120 can determine how many resources must be requested in connection with the created super packet 220 in order to ensure a desired level of service.

[0042] In FIG. 7, a flow chart depicting the operation of a system 100 in accordance with yet another embodiment of the present invention is illustrated. In general, the embodiment illustrated in FIG. 7 utilizes the DS field to identify flows, and also uses information included as part of RSVP requests provided in connection with individual data streams to make a single RSVP request for an aggregate stream. According to such an embodiment, at step 700, a DSCP is assigned to identify a remote subnetwork 108. At step 704, an RSVP enabled VoIP communication device 116 is configured with the multiplexer 120 on the first subnetwork 104 as the device's default IP gateway. The VoIP session is then initiated (step 708).

[0043] At step 712, the communication device 116 sends an RSVP request to the multiplexer 120 on the first subnetwork 104 (step 712). The multiplexer 120 then assigns the flow 204 associated with the RSVP request to a position within a new or existing aggregate flow 220 (step 716).

[0044] At step 720, the multiplexer 120 requests an RSVP reservation for the aggregate flow 220. The multiplexer 120 receives a return RSVP message, emulates the correct reservation message back to the communication device 116, and inserts a DCLASS object containing the assigned DSCP (step 724).

[0045] The communication device 116 then marks individual packets with the assigned DSCP, and sends the packets 204 to the multiplexer 120 (step 728). The multiplexer 120 aggregates the packets 204 having common DSCPs into one or more super packets 220, and delivers the super packets 220 to the remote subnetwork 108 under a single RSVP reservation (step 732). The super packets 220 are received at the second multiplexer 136 on the remote subnetwork 108, which deaggregates the super packets 220, and sends the individual packets 204 to the appropriate communication devices 132 on the remote subnetwork 108 (step 736).

[0046] From the above description, it can be appreciated that the multiplexer 120 is only required to consider the DS fields in order to determine whether an individual data packet 204 is suitable for aggregation into a super packet 220. In addition, the embodiment discussed in connection with FIG. 7 allows a single RSVP reservation to be established in connection with a super packet 220. Accordingly, a desired level of service may be maintained, without requiring that a separate RSVP reservation be established for each of the individual data packets 220.

[0047] With reference now to FIG. 8, a flow chart illustrating the operation of a system 100 in accordance with another embodiment of the present invention is illustrated. Initially, at step 800, packets associated with an RTP flow are received at a first multiplexer 120. The first multiplexer 120 inspects setup packets, defined according to any telephony signaling protocol, such as session initiation protocol (SIP) or H.323, used to initiate the media channel associated with the received packets (step 804). For example, in connection with a flow utilizing the session initiation protocol, the INVITE and ACK packets would be inspected to determine the parameters of the channel being negotiated. The multiplexer 120 then aggregates individual packets 204 that are addressed to a common remote subnetwork into one or more super packets 220 (step 808). The super packet 220 is then delivered to the remote subnetwork 108 (step 812). The multiplexer 136 associated with the remote subnetwork 108 then deaggregates the super packet 220, and sends the individual packets 204 to the appropriate communication devices 132 on the remote subnetwork 108 (step 816). The embodiment described in connection with FIG. 8 allows a router 128, 144 to identify individual flows 204 that are suitable for aggregation into a super packet 220, and also explicitly signals, according to the particular telephony signaling protocol used, the start and end of a realtime transport protocol flow.

[0048] Although the descriptions given above in connection with the various embodiments of the present invention have generally used the transmission of voice data in accordance with voice over Internet protocols or other realtime transfer protocols as examples, it should be appreciated that the present invention is not so limited. In particular, the present invention may be applied in connection with any transmission of data across a packet data network.

[0049] The foregoing discussion of the invention has been presented for purposes of illustration and description. Further, the description is not intended to limit the invention to the form disclosed herein. Consequently, variations and modifications commensurate with the above teachings, within the skill and knowledge of the relevant art, are within the scope of the present invention. The embodiments described hereinabove are further intended to explain the best mode presently known of practicing the invention and to enable others skilled in the art to utilize the invention in such or in other embodiments with various modifications required by their particular application or use of the invention. It is intended that the appended claims be construed to include the alternative embodiments to the extent permitted by the prior art.