WO2001033779A1

WO2001033779A1 - Voice packets in ip network

Info

Publication number: WO2001033779A1
Application number: PCT/SE2000/002089
Authority: WO
Inventors: Lars-Göran Petersen
Original assignee: Telefonaktiebolaget Lm Ericsson (Publ)
Priority date: 1999-11-03
Filing date: 2000-10-26
Publication date: 2001-05-10
Also published as: GB2372403B; KR20020059670A; CN1387716A; CN1157901C; SE9903982D0; SE9903982L; AU1424901A; GB2372403A; SE517961C2; US7397820B1; DE10085160B4; DE10085160T1; GB0209358D0; KR100740743B1

Abstract

The present invention relates to a method for transportation of user information in a core network (CN). The core network is located between a source access network (SAN) and a destination access network (DANB). The method comprises the following steps: collecting in a source centre (MSCa) in the source network (SAN) of voice packets from source units (1a, 2a, 3a, 4a, 5a) in the source network. The voice packets have the destination network (DANB) specified as receiving network; arranging of the collected voice packets into a payload part (IPPL) of an IP-packet (IP) in the core network (CN); transferring of the IP-packet in the core network (CN) from the source centre (MSCa) to a destination centre (MSCb) in the destination network (DANB).

Description

VOICE PACKETS IN IP NETWORK

TECHNICAL FIELD OF THE INVENTION

The present invention relates to a signal format in an IP bearer used in a core network for transmission of user data. The invention also relates to a method and an arrangement to create said signal format.

DES_CRIPTION OF RELATED ART

The use of asynchronous transfer mode networks "ATM" for transfer of information, such as video and voice, has become increasingly popular due to the high data rate and flexibility of such networks. ATM networks use a packet switching technique. An ATM network is a connection-oriented network in which each transfer of data between network devices is preceded by an order to establish a virtual connection between the devices.

In contrast, other types of networks such as internet are referred to as being connectionless because data is transferred over the network from a source device to a destination device without first establishing a connection as is required in ATM networks. In a connectionless network, data is transmitted by a source device with an address of the destination device and the connectionless network will route that data by any number of network paths to the desired destination device.

The differences in these data transfer techniques has been a major obstacle in linking a source device with a destination device via a network, where the source and destination routines generate and receive data in ATM-formatted frames and the connectionless network transmits data in Internet Protocol "IP" packets. In the US patent US 5,623,605 such data transfer is accomplished by using encapsulators and decapsulators to encapsulate ATM- formatted frames in data portions of IP packets for transmitting on the network.

In another prior art example, a Universal Mobile Telecommunications Network UMTS architecture comprises access networks and a core network. For UMTS, ATM has been standardized as bearer for compressed voice. The Internet Protocol is used as core network. Mobiles in one access network have independent and simultaneous voice calls to mobiles in another access network. The two access networks are connected to the IP core network and compressed voice packets are issued e.g. every 10 ms . The IP-based core network provides connectionless end-to-end service by means of routing in each traversed node. The IP format comprises a header and a payload and comes in two versions, namely IPv4, well known and established on the market today, and Ipv6, newly developed and beginning to be introduced on the market. An Ipv4 packet will have a header of 28 octets and an Iρv6 packet will have a combined header of 52 octets. When the IP-based core network issues an IP packet for every voice packet regularly every 10ms, the header-payload ratio between the average voice packet of 6 octets in the payload will become very unbalanced. For IPv4 the figures are 28 to 6. For Ipvβ the figures are 52 to 6. This unbalanced ratio will lead to bandwidth waste.

SUMMARY OF THE INVENTION

The present invention solves the problem with unbalanced ratio between header and payload in an IP packet when information is transferred between access networks of ATM- type, via a core network of IP-type.

The problem is solved by the invention by arranging multiple voice packets, having the same destination address, in the payload part of the IP packet. In a method according to the invention, voice packets from an ATM voice bearer, having the same access destination network addresses, are gathered in an access source network and arranged into the payload part of an IP bearer in a core network. The IP bearer transports the packets to the destination access network address . In the access destination network, packets are distributed to their final destination within the access destination network.

More in detail, the IP signal format includes:

- a header, comprising routing information,

a payload, comprising voice information.

The voice information includes multiple voice packets with the same destination address . The voice packets are arranged into the payload part from an ATM voice packet bearer.

One object of the present invention is to increase the bandwidth when transmitting voice packets in an IP core network between two access networks of ATM-type.

Another object is to enhance the overall voice quality by leaving voice packets untouched from source to destination access networks, even though a core network of different type has to be passed.

An advantage with the invention is the bandwidth saving that can be made in the core network .

Another advantage is that the packet format used in the source and destination access networks can be kept during the transport in the core network . DESCRIPTION OF THE DRAWINGS

Figure 1 is a perspective illustration of a telecommunication system having a source access network of ATM-type, a core network of IP-type and a destination access network of ATM-type.

Figure 2a discloses an ATM signal format including ATM cells with a header part and a payload part.

Figure 2b discloses an IP signal format including a header part and a payload part.

Figure 3 is a block schematic illustration of a source and a destination centre, communicating via a core network CN.

Figure 4 is a flow sheet illustrating a method to create the IP signal format.

Figure 5 is a block schematic illustration of a telecommunication system having a source access network of ATM-type, a core network of IP-type and several destinations access networks of ATM-type.

DESCRIPTION OF PREFERRED EMBODIMENTS

Figure 1 discloses a telecommunication system TS used for mobile communication . The telecommunication system includes a source access network SAN and a destination access network DANB, both of ATM-type (Asynchronous Transfer Mode) . The telecommunication system comprises a core network CN of IP- type (Internet Protocol) . The core network CN is used for transportation of user data such as voice packets VP between the source and the destination access networks SAN and DANB. The source access network SAN in figure 1 includes base transceiver stations bla, b2a and b3a communicating with mobile stations la, 2a, 3a, 4a and 5a. The communication between the base stations and the mobile stations constitute the radio interface. The radio interface in this embodiment is based on WCDMA (Wideband Code Division Multiple Access).

The ATM-networks SAN and DANB are connection-oriented networks in which each transfer of data between network devices is preceded by an order to establish a virtual connection between the devices . The ATM networks in figure 1 are AAL2/ATM based access networks. AAL2 (Adaption Layer 2) specifies ATM transport of connection-oriented circuit and high bit-rate packetized audio and video. The communication protocols in an ATM network are designed in a layered fashion with lower layer protocols providing services to the next higher layer. The physical layer is the lowest layer protocol, which provides access to the transmission medium, specifying details concerning physical interface signalling and timing. Data link layer protocols allow communication with the physical layer and provide link-by-link error detection/correction. Network layer protocols then take over by providing end-to-end addressing, flow control and integrity checking. The transport layer provides multiplexing onto the network. The session layer establishes a connection between systems. Finally, the presentation layer manipulates data into different forms for the highest layer, the application layer.

The base stations bla, b2a and b3a in the source access network SAN are connected to ATM/AAL2 switches Sla and S2a . The first and second base station bla and b2a are connected to a first switch sla. The first switch sla and the third base station b3a are both connected to a second switch s2a. The switches direct voice packets through the access network via the established virtual connection. The second switch s2a is connected to a mobile switching centre MSCa in the source access network SAN. The mobile switching centre MSCa acts as a gateway to the IP-based core network CN. The destination access network DANB has the same configuration as the source access network SAN previously mentioned. The mobile stations lb, 2b, 3b, 4b and 5b communicate with base stations bib, b2b, and b3b via a WCDM_A radio interface. The base stations in the destination access network DANB are connected to ATM/AAL2 switches Sib and S2b in the same manner as in the source access network SAN. The second switch in the destination access network is connected to a mobile switching centre MSCb which acts as a gateway to the IP-based core network CN.

The IP-based core network CN is connectionless. In a connectionless network, data is transmitted by a source device, which in this case is the source centre MSCa in SAN. An IP data header comprises a destination device address i.e. the address to the destination centre MSCb in DANB. The connectionless network will re-route the data by any number of network paths to the desired destination device. The rerouting takes place in IP-routers IPR1-IPR3. The IP-routers verify an address specified in the header in a received IP- packet and re-route the packet to the correct destination. In figure 1 is shown a general outline how voice packets VP are transported from the first base station bla in the source access network SAN to the first base station bib in the destination network DANB via the IP routers IP1-IP3.

The building up of ATM cells in the ATM access networks can be seen in figure 2a. The figure shows the basics of AAL2 packets multiplexed on ATM. ATM is connection oriented to its nature meaning that a connection must be established before any cell can be traversed. In the following example a connection in the source access network SAN is established between the first base station bla and the mobile switching centre MSCa. Compressed voice packets blal-bla5 from mobile stations communicating with the first base station bla is shown in figure 2. In the ATM adaptation layer AAL2, each voice packet blal-bla5 is provided with a header AAHD in which the destination address is specified. Only one header is shown in figure 2a . Voice packet and header together constitute an AAL2 packet. Before transportation in the ATM source access network SAN, an ATM cell is created. In figure 2a, two ATM cells are shown. The ATM cells have a similar design as an AAL2 packet, with a header part ATMHD1 and ATMHD2 and a payload part ATMPL1 and ATMPL2, but unlike the AAL2 packet, the ATM cell has a fixed payload size. The payload size of an ATM cell is 48 octets, while the payload size in an AAL2 packet can vary from 1 to 64 octets. In the example in figure 2a, the first two voice packets blal and bla2 are arranged in its entirety into the payload part ATMPL1 of a first ATM cell ATMC1. A third voice packet bla3 is partly arranged into the payload ATMPL1 of the first ATM cell ATMC1. The remaining part of the third voice packet bla3 is arranged into the payload part ATMPL2 of a second ATM cell ATMC2. The voice packet bla3 thereby overlaps the first ATM cell boundary and continue with the remainder in the next ATM cell ATMC2. After establishment of a connection, the ATM cell is transported through the source access network SAN from the first base station bla to the mobile switching centre MSCa.

Figure 3 discloses a block schematic illustration of the source centre MSCa communicating with the destination centre MSCb via the core network CN. The source centre MSCa comprises a multiplexer MUXa having outputs connected to "First In First Out" registers FIFO-b, FIFO-c and FIFO-d. Data sent to a first register FIFO-b belong to the destination centre MSCb, while data to the other two registers FIFO-c and FIFO-d belong to other destination centres which so far have not been discussed. Voice packets blal, b3al, b2al, bla2 and bla5 received from source units are received to the multiplexer MUXa and distributed to the registers FIFO-a, FOFO-b or FIFO-c in the source centre MSCa. Sending of control information from the source network to the destination network, via an established control channel precedes connection establishment. A call centre CC in the source network SAN analyses requested set-up of calls and distributes call set-up data to an address analyser AA in the source centre MSCa . The address analyser AA is informed by the call centre CC of the specified destination for received voice packets. The address analyser is connected to the multiplexer MUXa and directs after connection establishment voice packets to the correct register FIFO-b, FIFO-c and FIFO-d. The transportation of the voice packets from the source network SAN to the destination network DANB takes place over an IP core network CN in the payload part IPPL in an IP packet IP.

The destination centre MSCb is located in the destination network DANB and comprises a First-In-First-Out register FIFO to which voice packets from the first register FIFO-b is received. The register FIFO in the destination centre MSCb is connected to an input of a multiplexer MUXb in the centre . The multiplexer distribute voice packets to addressed destination units lb, 3b, 4b located in the destination network DANB. Voice packets in the payload part of one or more IP-packets with the same base station as destination, are arranged into an ATM cell in the destination network DANB. After establishment of a virtual connection in DANB specified in the voice packet header, the ATM cell is transported to the correct destination base station and voice packets are distributed to specified destination units .

According to the invention, voice packets with the same destination address, i.e. having the same access network as destination, are gathered in the mobile switching centre MSCa and transferred to the payload part of an IP packet after a specified time period. In figure 2b, the building up of an IP packet is shown. Compressed voice packets gathered in the mobile switching centre MSCa are transferred to the payload part IPPL of the IP packet IP. In figure 2b can be seen how the voice packets have been arranged in the IP packet. The voice packets having the same destination address, i.e. which all belong to the destination access network DANB, have been arranged in the payload part of the IP packet. The voice packets blal, bla2 and bla5 arrive from the first base station, b3al from the third base station and b2al from the second base station.

A method according to the invention will now be described more in detail. The telecommunication system used in the method is the system already shown in figure 1 and figure 3. References that already have been used in previous figures will also be used in the explanation of the method. The most essential steps of the method can be seen in figure 4. The method comprises the following step:

A call is set-up from a user of the source unit la in the source network SAN to a user of destination unit lb in the destination network DANB

The call set-up is analysed by the centre CC . The call centre send information to the address analyser AA in the source centre MSCa that voice packets received in the source centre from source unit la is to be sent to the destination centre MSCb in the destination network DANB.

- A first voice packet blal is sent over the air interface from the source unit la to the base station bla in the source network SAN.

The voice packet blal is transported from the base station bla to the source centre MSCa in a payload part of an ATM cell. The ATM cell is transported through the ATM source access network SAN, via the switches Sla and S2a. A voice packet b3al with destination unit 2b is transported from the base station b3 to the source centre MSCa in an ATM cell . A voice packet b2al with destination unit 3b is transported from the base station b2 to the source centre MSCa in an ATM cell.

Two voice packets bla2 and bla5 with destination units 4b and 5b are transported from base station bl to the source centre MSCa in one ATM cell . All voice packets are handled in the same manner as the previously mentioned first voice packet blal.

The voice packets blal, b3al, b2al, bla2 and bla5 are received from the base stations bl, b2, b3 by the source centre MSCa .

The address parts in the received voice packets are analysed by the address analyser AA. By checking the address of the voice packet blal and use the information already received from the call centre CC, the address analyser AA influence the multiplexer in the source unit to switch over and transmit the received voice packet blal to the first register FIFO- b.

All received voice packets blal, b3al, b2al, blal and bla5 having the destination units in the destination network DANB specified as receiving units will be switched over by the multiplexer MUXa to the register

FIFO-b intended for the destination centre MSCb.

A predetermined time 10ms lapses in the timer TIMER.

The voice packets blal, b3al, b2al, blal and bla5 gathered in the first register FIFO-b are arranged into the payload part IPPL of the IP-packet IP. The IP-packet is re-routed from the source centre MSCa to the destination centre MSCb through the core network CN, via the IP-routers IPR1, IPR2 and IPR3.

The IP-packet is received by the First- In-First-Out register FIFO in the destination centre MSCb.

Each voice packet is analysed and the multiplexer MUXb in the destination centre MSCb is arranged to distribute the received voice packet to the specified destination units lb, 2b, 3b, 4b, 5b.

In a second embodiment, the telecommunication system TS includes several destination networks (DANBx, DANCx and DANDx) . The destination networks are disclosed in figure 5 and represented by the destination centres MSCbx, MSCcx and MSCdx. According to the invention, the embodiment which now will be explained is used when the number of AAL2 packets to be sent to one access network are too few to constitute a cost efficient transfer in the payload part of an IP-packet. Figure 5 discloses a block schematic illustration of a source centre MCSax communicating with destination centres MSCbx, MSCcx and MSCdx via a core network CNx. The receiving input of a multiplexer receives voice packets from units in the source access network. The inputs of First-In-First-Out registers FIFO:bx-FIFO: kx are connected to an output of the multiplexer MUXax. The received voice packets in this embodiment have different destination centres MSCbx, MSCcx and MSCdx specified as receivers . In the same manner as discussed in figure 3, voice packets belonging to the destination centre MSCb are multiplexed into a First-In- First-Out-register FIFO:bx. In the same way, voice packets belonging to the destination centre MSCc is multiplexed into a First-In-First-Out-register FIFO: ex and voice packets belonging to MSCd is multiplexed into FIFO:dx. In this second embodiment, a control unit CU selects which registers FIFO:bχ-kx contain only a small number of voice packets after the lapse of a timer TIMERx. The number of voice packets, representing a small number could be an interval, for example 1-10. The voice packets in the selected FIFO- registers, in this embodiment the registers FIFO:bx, FIFO : ex and FIFO:dx, are all multiplexed into a register FIFO-bx located within MSCa but belonging to MSCb. The reason that FIFO:bχ-FIFO:dx have been selected is that the registers together have a number of voice packets suitable to transfer in the payload part of an IP-packet IP. The reason that MSCb has been selected to be the receiving centre for the IP- packet, which also contains voice packets to MSCc and MSCd, is that the voice packets to MSCb outnumber the number of voice packets to MSCc and MSCd. The voice packets gathered in the FIFO-register FIFO-bx, are re-arranged to the payload part of the IP-packet. The IP-packet is routed in the core network CNX to the centre MSCb in the destination access network DANBx. The destination centre MSCbx comprises a First-In-First-Out register FIFOx to which voice packets from the register FIFO-bx is received. The register FIFOx in the destination centre MSCb is connected to an input of a multiplexer MUXbx in the centre MSCB . The multiplexer distributes voice packets to addressed units lbx, 3bx and 4bx within the destination network DANBx. The voice packets that do not belong to units located within the network, i.e. voice packets that belong to the other two destination networks DANCx and DANDx are distributed via multiplexer to the FIFO-registers that transfer information to MSCcx and MSCdx via the core network CNx . In figure 5, the same core network CNx is disclosed twice. Voice packets sent to MSCcx and MSCdx are distributed to receiving units within respective network DANCx and DANDx. This second embodiment presumes some kind of organiser that keeps track of the sequence in which voice packets are sent from the source network. This can be done by, for example, adding information to the AAL2 header.

As a variant of this second embodiment, instead of first selecting which registers FIFO:bχ-kx contain only a small number of voice packets, voice packets from different chosen FIFO-registers are always sent together. Different destination networks are combined in clusters and voice packets in FIFO-registers belonging to a cluster is always put together and sent in one IP-packet to one of the destination networks in the cluster, i.e. a selected destination network, a so called master network. After arrival to the selected destination network, the voice packets belonging to other networks are further distributed to the other networks in the cluster. Instead of always in one IP-packet send voice packets belonging to clusters, as a further variant, change over is done between that method and a method where transportation from one source network to one destination network takes place in an IP-packet that carries only voice packets intended for that destination network. If the number of IP-connections between the source network and the destination network is less than, for example, five, transition is done to the transportation to cluster method. The number of single IP-connections from a source to a destination network thereby decides whether or not it is cost efficient to continue.

In a third embodiment not shown in any of the figures, an AAL2-router is located in the core network and connected to one of the routers in the core network CNx. In the previously discussed second embodiment, voice packets in the source centre MSCa were moved from FIFO-registers containing a low number of voice packets to the register FIFO-bx. In this third embodiment, voice packets from FIFO-registers having a low number of voice packets are moved to the AAL2- router . In the AAL2-router in the core network, the received voice packets are distributed to the source centres MSCbx, MSCcx and MSCdx to which the voice packets belong. Like in the second embodiment, this embodiment contributes to a more cost efficient transfer of voice packets within the core network. By using the above described methods, the ratio between the header part and the payload part in the IP- packets become more balanced.

Different variations are of course possible within the scope of the invention. Any type of IP-packet i.e. all variants of IP packets, are possible to use in the invention. The discussed core network of IP-type can be situated between two base stations, i.e. the source and destination access networks in such embodiment are two radio interfaces including base stations and mobile units . AAL2 packets have been used as examples of packets to transport in the core network between access networks. It is also possible to use variants of AAL2 packets in the payload of the IP packet. These variants can be created by reducing the AAL2 packet header to just two octets. Removing the header Error Control and reducing the user to user indication can for example do this. In other words, the invention is not restricted to the above described and illustrated exemplifying embodiments, and modifications can be made within the scope of the claims .

Claims

1. A method for transportation of user data in a core network (CN) , between a source access network (SAN) and a destination access network (DANB), which method comprises the following steps :

collecting in a source centre (MSCa) in the source network (SAN) of voice packets from source units (la, 2a, 3a, 4a, 5a) in the source network, which voice packets have the destination network (DANB) specified as receiving network;

arranging of the collected voice packets into a payload part (IPPL) of an IP-packet (IP) in the core network (CN);

- transferring of the IP-packet in the core network (CN) from the source centre (MSCa) to a destination centre (MSCb) in the destination network (DANB) .

2. A method for transportation of user information according to claim 1, which method comprises the following further step :

expiring of a predefined time, set in a timer (TIMER) in the source centre (MSCa) .

3. A method for transportation of user information according to claim 1 or 2, whereby the access networks (SAN, DANB) are of ATM-type.

4. A method for transportation of user information according to claim 1, 2 or 3, whereby the user information being of AAL2-type.

5. A method for transportation of user information according to any of claim 1-4, which method comprises the following further step :

distribution in the destination centre (MSCb) of the voice packets from the IP-packet (IP) to specified destination units (lb, 2b, 3b, 4b, 5b) in the destination network (DANB) .

6. A method for transportation of user information in a core network (CNx) from a source access network (SANx) to at least one destination access network (DANBx- DANDx), which method comprises the following steps:

collecting in a source centre (MSCAx) in the source network (SANx) of voice packets from source units in the source network, which voice packets have the at least one destination network (DANB -DANDx) specified as receiving network;

expiring of a predefined time, set in a timer (TIMERx) in the source centre (MSCAx);

moving of the collected voice packets into a payload part (IPPLx) of an IP-packet (IP ) in the core network

(CNx);

transferring of the IP-packet in the core network (CN) from the source centre (MSCAx) to a destination centre (MSCBX) in a first destination network (DANBx) of the at least one destination networks (DANBx-DANDx ) .

7. A method for transportation of user information according to claim 6, which method comprises the following further steps:

Forwarding of received user information from the first destination network (DANBx) to a second destination network (DANCx) of the at least one destination networks (DANBx-DANDx) .

8. A method for transportation of user information in a core network (CNX) from a source access network (SANx) to at least one destination access network (DANBx- DANDx), which method comprises the following steps:

collecting in a source centre (MSCAx) in the source network (SANx) of voice packets from source units in the source network, which voice packets have the at least one destination network (DANBx-DANDx) specified as receiving network;

moving of the collected voice packets into a payload part (IPPLx) of an IP-packet (IPx) in the core network (CNx);

transferring of the IP-packet in the core network (CN) from the source centre (MSCAx) to an IP-router in the core network (CNx) .

9. A method for transportation of user information according to claim 8, which method comprises the following further steps:

Forwarding of received user information from the IP- router to at least one destination network (DANBx-

DANDx) .

10. A signal format for use in IP bearers to increase transmission bandwidth between access networks in a telecommunication system, which IP signal format includes :

a header (IPHD), comprising routing information;

a payload (IPPL), comprising voice information;

c h a r a c t e r i s e d in that

- the voice information includes multiple voice packets (blal, b3al) having the same destination address, which packets have been arranged into the payload (IPPL) from an ATM voice bearer.

11. A signal format according to claim 10 whereby an adaptation layer AAL2 is used to transport the compressed voice from the ATM voice bearer.

12. Arrangement for transportation of user information in a core network (CN) between a source access network (SAN) and a destination access network (DANB), which arrangement comprises : means for collecting in a source centre (MSCa) in the source network (SAN) of voice packets from source units (la, 2a, 3a, 4a, 5a) in the source network, which voice packets have the destination network (DANB) specified as receiving network;

means for arranging the collected voice packets into a payload part (IPPL) of an IP-packet (IP) in the core network (CN);

means for transferring of the IP-packet in the core network (CN) from the source centre (MSCa) to a destination centre (MSCb) in the destination network (DANB) .

13. Arrangement for transportation of user information according to claim 12, which arrangement further comprises :

a timer (TIMER) in the source centre (MSCa) .