Multiplex communications – Communication over free space – Combining or distributing information via time channels
Reexamination Certificate
1999-05-05
2003-04-15
Hunter, Daniel (Department: 2684)
Multiplex communications
Communication over free space
Combining or distributing information via time channels
C370S310000, C370S431000
Reexamination Certificate
active
06549530
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to integrated signalling mechanisms for ATM-like services in shared medium networks. A shared medium network is a network that contains at least one multiple access (MA) segment. Users of a MA segment share a common medium link to access the network.
Shared medium segments may be classified according to their topologies, for example, ring topology (eg. a LAN), bus topology and star topology. The present invention is mainly applicable to shared medium segments having a star topology, which means that end systems on the network which share the shared medium cannot communicate directly with each other. Instead all end systems in the MA segment communicate with a headend, which is essentially a switch. The headend is responsible for routing traffic from the end systems of the shared medium segment to other end systems in the segment or to other parts of the network (the wide area network) and for routing traffic from the wide area network to the end systems.
This type of segment is essentially asymmetric in the sense that end systems share a common medium uplink to communicate with the headend, while the headend has a dedicated downlink to communicate with all the end systems. Problems are created by the reception of the same ATM stream (containing all traffic directed to all end systems in the segment) by all the end systems in the segment and by all the end systems in the segment having to compete to establish connections or calls on the common medium to communicate with the headend. These problems are exacerbated when there are large numbers of users in a segment.
The ATM stream may carry data traffic, ATM signalling traffic (eg. information for setting up a connection or call on the network) and ATM management traffic (eg. network administration information, including registration of end users).
An example of a shared medium network with star topology is a network, as shown in FIG.
1
. The network in
FIG. 1
has one MA segment, or common medium beam (
2
).
In adapting ATM to shared medium segments, a number of problems have to be solved. These include, maximising the useable amount of the shared medium bandwidth for data traffic, identifying the source of traffic in order to control access to the shared medium and for billing purposes and reducing delays to end systems for access to the shared medium. In addition where the headend comprises a geostationary satellite (
4
) supported by a ground based network controller (
6
), as shown in
FIG. 1
, propagation delay effects have to be minimised, for example, by reducing the number of messages exchanged to set up a call. Also, when the headend is a satellite a large number of end system have to be catered for on the common medium.
Furthermore, standard ATM virtual channels (VCs) which may have to be modified in order to transmit them over the shared medium segment, must appear to be standard ATM VCs to the wide area network and the end systems, if the shared medium segment is to be able to operate as part of a wide area network which uses conventional end system processing. Management VCs originating and terminating within the shared medium segment may be modified within the ATM Adaptation layers, but must be supported in an unmodified manner above this layer. In particular, the SNMP (Simple Network Management Protocol) layer of the ILMI (Interim Local Management Interface) must be unmodified. Also, signalling VCs originating and terminating within the shared medium may be modified within the ATM Adaptation layers but must be supported in an unmodified manner above this layer. In particular the call signalling layer (eg. Q.2931, PNNI, B-ISUP, BICI, AINI) should remain unchanged, ie. the SAAL (Signalling ATM Adaptation Layer) must provide a conventional SAAL service.
In a geostationary satellite system, round trip delay (ie. propagation delay for messages sent from an end system to the satellite and for response from the satellite to the end system) is high. Wireless telephony solutions such as GSM (Global System Mobile) tend to require multiple exchanges between an end system and an intermediate system in order to establish a call and so can generate unacceptable delays.
Conventional wireline ATM systems are known which operate in a point to point manner whereby a dedicated two-way simultaneous link exists between a port on an end system and a port on an intermediate system. The intermediate system performs four basic functions; cell relay, ATM end system registration, ATM peer intermediate system registration and ATM signalling. In a shared medium segment the activities of the intermediate system will have to be dealt with by the headend of the segment in combination with a NCC (Network Control Centre). For a network segment with a satellite headend the NCC will generally be ground based and thus physically separated from the headend.
The cell relay function consists of receiving a cell, examining the cell header, forwarding the cell to the appropriate output port and replacing the cell header with a new cell header.
The ATM end system registration function consists of allocating a unique address to each end system and associating the address with the appropriate intermediate system port. This is done by concatenating the end system's unique address with the intermediate system's assigned address. An automatic process using the SNMP of the ILMI conventionally does this over a standard VC identified by Virtual Path Identifier VPI=0 and Virtual Channel Identifier VCI=16 during the set up of the link between the port of the intermediate system and the end system.
The ATM peer intermediate system registration function consists of exchanging routing information with peer intermediate systems and associating routing information with each link to another intermediate system.
The ATM signalling function consists of exchanges between end systems and peer intermediate systems to establish call routing information which is subsequently used by the cell relay function. The system relies on having individual duplex signalling and ILMI channels to each end system. The ATM signalling protocol is run over a standard VC identified by VPI:VCI (0:5). The ATM signalling function also relies on the Service Specific Connection Orientated Protocol (SSCOP) to provide assured data delivery between peer signalling entities, eg. end systems and intermediate systems or peer intermediate systems.
SSCOP provides reliable transport of ATM signalling messages between two signalling entities using timeouts and retransmissions. The sender periodically sends a POLL Protocol Data Unit (PDU) to enquire about the state of the receiver. The receiver replies with a STAT PDU to tell the sender which packets were correctly received. The sender then uses the STAT messages to adjust its window size and retransmit lost packets. Once an SSCOP connection is established between two signalling entities it is kept alive by exchanging ‘keep alive’ messages periodically.
Conventional wireline ATM systems use the physical port on the intermediate system to identify the end system at the interface between a single end system and its associated intermediate system.
If multiple users are required per port an ATM Forum UNI 4.0 can be used to support the multiple users using a virtual UNI concept. The virtual UNI concept assigns each end user one or more Virtual Path Connections (VPCs) using the VPI addressing field and the UNI uses a VP cross-connect to combine the user VPCs so that the intermediate system effectively sees a single end system. As the VPI addressing space comprises a single octet, there is a limit of 256 end systems per port.
An improvement of the Virtual UNI scheme is possible where the 24 bits of VPI (8 bits) and VCI (16 bits) are re-partitioned to give more to the VPI. For example an even division giving the VPI 12 bits could support 4000 end systems with 4000 VCs. It may happen that the 24 bit VCI/VPI space is already reduced to fewer bits by some systems that wish, for example, to use
Abu-Amara Hosame Hassan
Cable Julian
Minato Alessandro
Rosenberg Catherine
Barnes & Thornburg
Chow C.
Hunter Daniel
Nortel Networks Limited
LandOfFree
Integrated signalling for asynchronous networks does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Integrated signalling for asynchronous networks, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Integrated signalling for asynchronous networks will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3017877