GB2442261A - Termination of packet based communication on detection of traffic congestion - Google Patents

Abstract

A communication system (100) for providing packet data communication including a first host (101), a second host (121) and a communication network (131) operable to provide a communication path (151) between the first host and the second host, wherein the first host is operable to send a packet data communication to the second host via the communication path and the second host is operable to receive a packet data communication from the first host, wherein each of the first and second hosts is operable to detect traffic congestion in the network along the communication path between them and, in response, to terminate a communication between the first and second hosts. Also described is a terminal (101) and a method (300) used in the system 100.

Description

1 2442261 TITLE: COMMUNICATION SYSTEM, TERMINAL AND METHOD The present

invention relates to a communication system and a terminal and a method for use therein. In particular, the invention relates to packet data communication between hosts of the system.

BACKGROUND OF THE INVENTION

Communication systems which provide communications in the form of data packets between a sending terminal and receiving a receiving terminal are now in wide use.

The internet, which operates according to standard procedures defined in the Internet Protocol (IP), operates in this way. Some wireless communication systems are also being designed to employ procedures which operate in accordance with Internet Protocol (IF) compatible standards to manage communications within the system. For example, some wireless systems which are TETRA systems, which operate according to the TETRA standard protocols defined by ETSI (the European Telecommunications Standards Institute), are being designed in this way. TETRA systems are primarily designed for application by professional radio users such as the emergency and security services. Other systems designed for application by such professional radio users and which employ methods which are compatible with the Internet Protocol include APCO 25 systems which operate in accordance with the APCO Project 25 standard defined by the Association of Public-Safety Communications Officials-International, Inc. in IP compatible systems, communicated information, including traffic information, i.e. user communicated information such as speech or text, picture or video informaLion, and system control information, is sent in the form of IP compatible data packets between terminals of the system by use of IP addresses. Unicast (point-to-point) II addressing may be used where only a single target terminal is to receive a communication sent via the system from a single sending terminal. Multicast (point-to-multipoint) IP addressing may be used where more than one target terminal has to receive communicated information from a sending terminal. Target terminals are generally user terminals, e.g. mobile stations, although in wireless communication systems they may also be base stations serving user terminals.

In addition, in wireless communication systems one of the terminals may comprise a dispatcher terminal used for receiving and sending messages relating to operation of the organisation, e.g. police, in which the communication system is used. Alternatively, or in addition, one of the terminals may comprise a recorder, e.g. voice recorder, to record traffic information being communicated. Alternatively, or in addition, the system may include telephony gateways and gateways to other communication systems. Each of these gateways may comprise a terminal in the system. Generally, user terminals and the BSs that serve them, as well as the other terminals which have been referred to, are known as hosts' . Generally, a host is a station or node that can take part in two way communication with other hosts.

Communications between hosts are delivered by routers of the system arranged in a network.

in some known communication systems there is little or no attempt to guarantee quality of communication service between an originating (sending) host and a target host to ensure that a communication reaches the target host. In other words, the communication is sent without knowledge of any congestion which might be present in the system which might lead to loss of some of the information of the communication. In some other known communication systems quality of communication service between a sending host and a target host is achieved by use of resource reservation, particularly where the system is a telephony system such as a wireless communication system. For example, in a mobile telephony system such as a TETRA system, a call set up procedure operated by a system controller is conventionally applied to establish a fixed bandwidth full duplex connection along a path between an originating host and a target host. The appropriate channel resource has to be reserved in each intermediate node, e.g. each router, of the system along that path.

Each of these intermediate nodes has to have state information relating to the state of the communication.

Unfortunately, this envisaged requirement is not easily applied in large networks and has not been successfully applied in large networks. Knowledge of the distribution of traffic and capacity available may be unknown from one zone of a large network to the next. Furthermore, where the communication system comprises a wireless communication system providing mobile connectivity, changes in topology of the system can cause difficulties by requiring new resource reservaLions to be made for existing communications. Often existing calls are dropped when such topology changes occur.

SUMMARY OF THE INVENTION

According to the present invention in a first aspect there is provided a communication system as defined in claim 1 of the accompanying claims.

According to the present invention in a second aspect there is provided a communication terminal as defined in claim 2 of the accompanying claims.

According to the present invention in a third aspect there is provided a method of operation in a communication system, the method being as defined in claim 22 of the accompanying claims.

Further features of the invention are as defined in the accompanying dependent claims and in the embodiments of the invention to be described.

Embodiments of the present invention will now be described by way of example with reference to the accompanying drawings, in which:

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a block schematic diagram of an illustrative wireless communication system which may be adapted in accordance with embodiments of the invention.

FIG. 2 is a block schematic diagram of an illustrative layout of a base station of the system of FIG. 1.

FIG. 3 is a flow chart of a method embodying the invention for use in the system of FIG. 1.

FIG. 4 is a flow chart of another method embodying the invention for use in the system of FIG. 1.

FIG. 5 is a flow chart of another method embodying the invention for use in the system of FIG. 1.

FIG. 6 is a flow chart of another method embodying the invention for use in the system of FIG. 1.

DESCRIPTION OF EMBODIMENTS OF THE INVENTION

In accordance with embodiments of the invention to be described, a communication system for providing packet data communication includes a first host, a second host and a communication network operable to provide a communication path between the first host and the second host. The first host is operable to send a packet data communication to the second host via the communication path and the second host is operable to receive the packet data communication from the first host. Each of the first and second hosts is operable to detect congestion in the network along the communication path between them and, in response, to terminate a communication between the first and second hosts.

The first host may thus be an originating host and the second host may be a target host.

The communication system may comprise a wireless communication system. The first and second hosts in such a system may comprise base stations (base transceiver stations) The communication system may operate by procedures which are compatible with the IP (Internet Protocol) standard.

FIG. 1 is a block schematic diagram of a communication system 100 adapted In accordance with an embodiment of the invention. The system 100 is a communication system in which communication between hosts of the system 100 comprises data packet wireless communication. The hosts comprise base stations, also referred to herein as BSs' (or BS' in the singular), which serve mobile stations, also referred to herein as MSs' (or MS' in the singular) . It will be apparent to those skilled in the art that the system 100 and the components which are to be described as operating therein, particularly the hosts therein, e.g. base stations, may take a number of forms. Thus, the form of the system 100, and of its operational components, to be described should be regarded as illustrative rather than limiting. The system 100 includes a ES (base station) 101 which provides radio communication service to user terminals within range of the ES 101, i.e. within a cell or area having the ES 101 at its centre. The BS 10]. may serve many user terminals. Two such terminals in the form of NSs (mobile stations), namely an MS 103 and an MS 105, are shown having respectively radio links 107 and 109 to the ES 101. The system 100 also includes another ES 111 which provides radio communication service to user terminals within range of the BS 111, i.e. within a cell or area having the ES 111 at its centre. The BS 111 may serve many user terminals. Two such user terminals in the form of MSs, namely an MS 113 and an MS 115, are shown having respectively radio links 117 and 119 to the ES 111. The system 100 also includes a ES 121 which provides radio communication service to user terminals within range of the BS 121, i.e. within a cell or area having the BS 121 at its centre. The ES 121 may serve many user terminals. Two such user terminals in the form of MSs, namely an MS 123 and an MS 125, are shown having respectively radio links 127 and 129 to the BS 121.

The ES 101 is operationally associated via a link 104 with a router 102, e.g. a core router, which routes communications from the BS 101 (on behalf of MSs served by the BTS 101) to other terminals, e.g. MSs served by other BSs, within the system 100 and in other systems (not shown) operably connected to the system 100. The router 102 also routes incoming communications to the BS 101. The ES 111 is operationally associated via a link 114 with a router 112, e.g. a core router, which routes communications from the BS 111 (on behalf of MSs served by the BS 111) to other terminals, e.g. NSs served by other BSs, within the system 100 and in other systems operably connected to the system 100. The router 112 also routes incoming communications to the ES lii. The BS 121 is operationally associated via a link 124 with a router 122, e.g. a core router, which routes communications from the BS 121 (on behalf of MSs served by the BS 121) to other terminals, e.g. MSs served by other BSs, within the system 100 and in other systems operably connected to the system 100. The router 122 also routes incoming communications to the BS 121. The routers 102, 112 and 122 are operably connected via further links 132, 134 and 136 to a network 131 which may for example include a piurality of further routers and/or other nodes (not shown) which may include one or more telephony gateways and/or gateways to other communication systems. The routers 102, U2 and 122 are mutually connected to one another and to other routers (not shown) via the network 131. The links 104, 114, 124, 132, 134 and 136 and individual links (not shown) between nodes and/or routers (not shown) in the network 131, may comprise wired and/or wire'ess links. The routers 102, 112 and 122, together with any routers in the network 131, form a router network in the system 100. Although each of these routers is shown in FIG. 1 serving one associated BS, each router of the router network may serve each of a plurality of associated ESs, as will be apparent to those skilled in the art.

FIG. 2 shows an illustrative layout 200 of operational components in each of the BSs 101, 111 and 121. A similar layout may be used in any other base station (not shown) of the system 100. A controller 201 of the base station having the layout 200 controls functional operations of the base station. A processor 202, e.g. a digital processor, operably connected to the controller 201 processes information sent in RF signals to and from the base station. The controller 201 and the processor 202 are operably connected to a timer 205 which provides operational synchronisation and timing and to a memory 206 which stores data and programs needed in operation by the controller 20] and the processor 202.

The processor 202 is operably connected to a plurality of RF transceivers two of which are shown, namely an RE transceiver 203 and an RF transceiver 207.

Each of the RF transcei\Ters 203 and 207 transmits and receives RE signals including signals carrying information sent to and from user terminals, including MSs, served by the BS 101. The signals are delivered over-the-air to and from an antenna 204 connected to the RE transceiver 203 and to and from an antenna 208 connected to the RE transceiver 207. The base station having the layout 200 may include one or more further RE transceivers and associated antennas (not shown) When the RE transceiver 203 receives via the antenna 204 an RE signal including information representing communicated information, the signal is passed to the processor 202. Similarly, when the RE transceiver 207 receives via the antenna 208 an RE signal including communicated information, the signal is passed to the processor 202. The processor 202 converts each signal including communicated information from the transceiver 203 or the transceiver 204 into a demodulated electronic signal including the communicated information. The communicated information processed in this way includes system control information as well as user communicated traffic information for onward delivery. Where the communicated information comprises system control information the demodulated electronic signal produced by the processor 202 is passed to the controller 201. Where the demodulated electronic signal produced by the processor 202 comprises user communicated information for onward delivery it is delivered to a router 212 which routes the electronic signal towdrd its destination. For the BS 101, the router 212 may provide the function of the router 102 (FIG. 1) or it may be an add]tional router. For the BS 111, the router 22 may provide the function of the router 112 (FIG. 1) or it may he an additional router.

For the BS 121, the router 212 may provide the function of the router 122 (FIG. 1) or it may be an additional router.

Each incoming electronic signal received by the router 212 which includes communicated traffic information, i.e. user information to be sent to one of the user terminals including mobile stations served by the base station having the layout 200, is routed by the router 212 to the processor 202. The processor 202 processes the electronic signal into a form suitable for inclusion in an RF signal for transmission by the transceiver 203 via the antenna 204 or for transmission by the transceiver 207 via the antenna 208.

The processor 202 also prepares and receives system control messages received from the controller 201.

The base station layout 200 includes a power supply 211, e.g. from the main (mains) electricity supply, which provides a source of electrical energy for all active components of the BS layout 200.

The base station layout 200 includes, operably coupled to the controller 201 and to the processor 202, a priority indication processor 213 and a congestion indication processor 215. The processors 213 and 215 process information relating respectively to indications of: (i) a leve] of pciority in communications sent to and from the processor 202; and of (ifl detected congestion in communications sent to and from the processor 202. The functions of the priority indication processor 213 arid the congestion indication processor 215 are described further as follows.

Communication of traffic information is referred to herein as a call' (even though the traffic information to be communicated may be other than speech information) . Communication between an originating BS and one or more target BSs to establish a call, but before any traffic information is sent in the call, is referred to herein as a call start' . A call start may comprise sending of call set up data messages in a known format. Each of the base stations of the system 100 having the layout 200, including the BS 101, the BS ill and the BS 121, is able to set, by its priority indication processor 213, a priority level of calls or call starts that it makes on behalf of the mobile stations that it serves. Each of these base stations is also able to detect by its priority indication processor 213, a priority level of calls and call starts that it receives on behalf of the mobile stations that it serves. The priority level in each case may be indicated in signalling employed in making and delivering calls or call starts, for example in a data packet header for a packet data transmission. Such an application and indication of a priority level may be known per se. For example, the indication may comprise a six bit number in the Differentiated Services Field of an IP Version 4 Header as defined in the document entitled RFC (Request For Comments) 2474' . That document has been issued by the IETF (Internet Engineering Task Force) which is a body that defines Internet operating standards and is supervised by the Internet Society's Tnternet Architecture Board. Use of priority levels in calls and call starts sent and received by the base stations having the layout 200 in the system 100 in embodiments of the invention is described later.

Each of the base stations of the system 100 having the layout 200, including the ES 101, the BS ill and the ES 121, is able to detect, by its congestion indication processor 215, traffic congestion in the system 100.

Furthermore each of the base stations of the system 100 having the layout 200, Including the BS 101, the BS 111 and the BS 121, is able to insert, by the congestion indication processor 215, in communications sent by the base station an indication of traffic congestion in the system 100 known to the base station, e.g. from detections made by its congestion indication processor 215. The congestion detected by or known to the congestion indication processor 215 of each base station of the layout 200 may have been caused in a number of ways including one or more of the following: (a) new calls being established in the system 100; (b) mobile hosts, e.g. MSs, moving from one part of the system 100 and attaching to another part of the system; (c) changes in network topology of the system 100; and (d) changes of link speeds in the system 100 where adaptive wireless links are used.

Detected congestion may be indicated in signalling employed in the system 100 in making and delivering calls and call starts, for example in a packet data header for a packet data transmission. Such an indication of congestion may he known per Se. For example, the indication may comprise a two bit number in the Explicit Congestion Notification Field of an IP version 4 Header as defined in the document entitled RFC (Request For Comments) 2474' referred to above.

Such an indication may be added in the signalling by one of the base stations having the layout 200 or by one of the routers in the system 100, including the routers 102, 112 and 122 and any routers (not shown) in the network 131, upon detection of traffic congestion in the system 100.

Alternatively, or in addition, each of the base stations of the system 100, including the BS 101, the BS 111 and the BS 121, is able by its congestion indication processor 215 to monitor the quality of received data packets in a known manner, e.g. using a known quality monitoring protocol, such as the RTCP (Real Time Transport Control Protocol) defined by the ILETF in its published document entitled RFC 3550', to determine in a known manner that there has been loss of data packets and therefore there is congestion in the system 100.

Use of congestion detection, particularly by base stations, in calls sent and received in the system 100 is described later in embodiments of the invention.

It will be apparent to those of ordinary skill in the art that two or more of the processors of the base station layout 200, including the controller 201, the processor 202, the priority indication processor 213 and the congestion indication processor 215 may he implemented in the form of a common processor, e.g. a digital signal microprocessor, programmed to carry out the functions of the different processors.

A first illustrative method 300 of operation of the system 100 in accordance with an embodiment of the invention will now be described with reference to FIG. 3 which is a flow chart of the method 300. In the method 300, it is assumed that the MS 105 is to be an originating MS which intends to send a communication and that the BS 101 serving the MS 105 is therefore an originating ES. It is also assumed that the intended communication is to be sent to the MS 125 which will be a target receiving MS. The BS 121 serving the MS 125 is therefore a target ES. Information required to be communicated by the originating MS 105 may comprise speech information or data, e.g. a data stream representing aiphanumeric text characters or picture or video data. The information to be communicated is herein referred to as traffic information', and the signal carrying such information is herein referred to as a traffic signal' . Communication between the originating ES 101 and the target BS 121 takes place along a communication path 151 between the originating BS 101 and the target ES 121. The communication path 151 extends via the link 104 to the router 102, via the link 132 to the network 131, via the link 136 to the router 122 and via the link 124 to the BS 121.

The method 300 illustrates how an embodiment of the nventiori is applied in the case of a call start procedure. In a step 301 of the method 300, the originating MS 105 sends a call set up request message to the originating BS 101. In a step 303, the originating 135 101 sets, by its priority indication processor 213, a priority level of the intended call and forwards a call set up request message including an indicaLion of the priority level it has set. The priority level may be a high priority level if the intended call is an emergency call or a lower priority level if the intended call is a non-emergency call. The request message is passed along the communication path 151 between the BS 101 and the BS 121. In a step 305, each of the routers along the path 151 monitors for congestion in a known way. If congestion which will affect the intended call is detected by a router of the path 151, the router adds a congestion notification in the call set up request, e.g. by adding an indication in a data packet header as described earlier. Each router forwards the request message with a speed that depends on the priority level indicated in the request message.

The target BS 125 receives the call set up request in a step 307. In a step 309, the target BS 125 determines by its congestion indication processor 215 whether there is a congestion notification in the received request message. If step 309 produces a NO' result, i.e. no congestion notification is detected, a step 311 follows in which the target BS 121 sends a notification to the target MS 125 so that the target MS can prepare to take the intended call. The target ES 121 also sends a response message to the originating BS 101 in a step 313 to indicate that the call set up request is accepted. The response message includes an indication that no congestion has been detected along the communication path 151. If the response message is returned via the communication path 151 without encountering any congestion, a step 315 follows in which the intended cal] takes place. The originating MS 101 sends traffic information via the link 107 to the ES 101, and the BS 101 sends the traffic information via the communication path 151 to the target ES 121. The target BS 121 forwards the traffic information to the target MS 125.

If step 309 produces a YES' result, i.e. congestion notification is detected, a step 317 follows in which the target BS 121 determines whether to terminate the call start procedure. The target BS 121 may carry out this determination in one of the ways described later. The result of this determination may be to terminate the call start, a YES' result, or not to terminate the call set up, i.e. call start, a NO' result. If step 317 produces a YES' result, the call start is terminated by the target 55 121 in a step 319.

The BS 121 may also send in a step 321 a notification message to the originating ES 101 to indicate that the call start is terminated. The notification message may also indicate that congestion was detected on the path 151 between the BS 101 and the ES 121. Following receipt of the notification message sent in step 321, the originating BS 101 may send re-send the call set up request in a step 323, e.g. after a suitable delay period. The re-sent call set up request is sent via the communication path 151. Steps 303 onward are then repeated.

If step 317 produces a NO' result, e.g. because the received call set up request has an indicated high priority level, the target ES 12 sends in a step 325 a response message, which may include a congestion notification, to indicate that the call set up request is accepted. In a step 327, the originating ES 101 in response to receiving the response message with a congestion notification itself determines whether to continue with the call start or to terminate the call start. The determination may be carried out in one of the ways described later. This determination may result in the call start being terminated or continued with as appropriate.

FIG. 4 is a flow chart of another illustrative method 400 of operation embodying the invention in one of the base stations having the layout 200, e.g. the ES 101. The method 400 begins at a step 401 which represents normal operation in the BS. The ES is receiving for the NSs it serves call start messages and also traffic information in a number of existing incoming calls. It is also sending on behalf of the MSs that it serves at least one call start message as well as traffic information in a number of existing outgoing calls. In a step 403, the BS detects, by its congestion indication processor 215, congestion in the system 100 from the incoming calls and call start messages. For example, congestion notifications may have been added in the incoming calls by one of the routers which have routed the incoming calls. In a step 405, the BS applies, by its priority indication processor 213, priority levels to the existing outgoing calls it is handling arid also to call starts it is sending. Step 405 may be applied in response to the BS detecting the congestion notification or it may be a step applied continuously. The priority levels applied may be indicated in the calls or call starts as described above. These priority levels may be applied to all outgoing calls and call starts. Alternatively, the priority levels may be applied selectively, e.g. only for calls and call starts that are known to be going to or through a part of the system 100 known to be congested, but not for calls and call starts that are known to be going to and through a part of the system known not to be congested. Examples of priority levels which may be applied are given later. In response to step 405, the ES applies in a step 407 selective termination of outgoing calls and call starts depending on the priority level of each call or call start. The termination may be applied by issue of a signal from the controller 201. Examples of ways in which the ES may apply selection of calls and call starts for termination are given later.

In a step 409, which may be applied in response to detecting the congestion in step 403 or may be applied continuously, the ES may detect by the priority indication processor 215 priority levels of the existing incoming calls it is handling and also of any incoming call starts. Examples of priority levels detected are given later. In response to steps 403 and 409, the BS may apply in a step 411 selective termination of incoming calls and call starts, depending on the priority level of each incoming call or call start.

Examples of ways in which the BS may apply selection of calls and call starts for termination are given later. Step 407 may he carried out at the same time as step 411 as illustrated

in FIG. 3. Alternatively, one of these steps may be carried out before the other. For example, outgoing calls and call starts may be selectively terminated before any incoming calls or call starts are selectively terminated.

In a step 413, the BS detects from incoming calls and call starts that there is reduced congestion in the system 100, for example by detecting that fewer incoming calls per unit time have a congestion notification. In response, in a step 415, the ES applies a revised selective termination to calls and call starts (outgoing and/or incoming) . In other words, more calls or call starts are allowed to avoid termination. The priority level at which calls or call starts avoid termination may be lowered so that more calls or call starts are allowed to continue, thereby providing an activity level that better matches the reduced congestion level.

Eventually, in a step 417, the BS detects no congestion in the system 100 and, in response, in a step 419 the BS ceases termination of calls or call starts, e.g. by issue of a control signal by the controller 201.

The methods 300 and 400 described above are useful when a unicast call to a single target BS such as the ES 121 is being requested or delivered by an originating ES, e.g. the BS 101. The methods 300 and 400 are also useful when a multicast call is being requested or delivered from an originating ES such as the BS 101 to a plurality of target ESs such as the ES ill Ond the BS 121. Where the cell intended or being delivered is a multicast call, the steps in the method 300 or the method 400 which are to be applied by a ES as a target BS may be applied individually by each of the target BSs of the multicast call, e.g. including the BSs 121 and 111. Thus, each of the target BSs in a step 311 and step 315 of the method 300 or in similar steps for existing calls may terminate its own participation in the call or call start. Alternatively, or in addition, each target ES may send to the originating BS a message similar to the message sent in step 321 of the method 300 indicating that congestion has been detected. In response, the originating BS may decide whether to selectively terminate the entire multicast call or call start.

FIG. 5 is a flow chart of another method 500 embodying the invention illustrating how selective termination of calls and call starts may be applied by a ES. The ES may be an originating BS or a target BS or both. The ES in the method 500 may for example be the BS 101. In a step 501, the BS detects, by its congestion indication processor 215, congestion in the system 100 in one of the ways described earlier and therefore becomes aware that it must begin selective termination.

In a step 503, in a first period Ti following the congestion detection in step 501, the BS terminates all call starts. A termination control signal may be issued by the controller 201. The cafl starts may include all call starts relating to intended outgoing calls requested by MSs served by the BS and may also include all call starts relating to intended incoming calls which have been requested. In a step 505, in a period T2 following the period Tl, the ES terminates existing calls having an age less than a given threshold age Al.

The given threshold age Al may be a fixed threshold age pre-recorded in the BS, e.g. in the memory 206, or it may be an age which is dynamically set by the ES according to current detected operational conditions in the system 100. The ES also terminates any new call starts which appear in the period T2. The existing calls having an age less than Al which are terminated during the period T2 may be existing outgoing calls or existing incoming calls or both. In some cases, the threshold age Al applied for selective termination of incoming calls may be different from that applied for selective termination of outgoing calls. In a step 507, the ES terminates in a period T3 following the period T2 existing calls having an age less than a given threshold age A2. The given threshold age A2 is greater than the given threshold age Al. The given threshold age A2 may be a fixed threshold age pre-recorded in the BS or it may be an age which is dynamically set by the BS according to current detected operational conditions in the system 100. The BS also terminates any new call starts which appear in the period T3. The existing calls terminated during the period T3 may be existing outgoing calls or existing incoming calls or both. In some cases, the age A2 applied as threshold for selective termination of incoming calls may be different from that applied as threshold for selective termination of outgoing calls.

Eventually, in a step 509, no congestion is detected by the BS. In d step 511 which follows step 509, the ES ceases termination of calls and call starts, e.g. by issue of a control signal by the controller 201.

In the method 500, the periods Ti, T2 and T3 may be of equal length. Alternatively, at least two of these periods may have different lengths. The length of each of the periods Ti to T3 may be fixed or may be selected dynamically by the ES, e.g. by the controller 201, according to current operational conditions in the system 100.

It will be apparent from the above description that in the method 500 calls and call starts are prioritised in terms of their age. Call starts generally are younger than existing calls and may be given a (lowest) priority level 1. Existing calls having an age less than Al are the youngest existing calls and may be given a priority level 2. Existing calls having an age less than A2 but not less than Al may be given a priority level 3.

Existing calls having an age not less than A2 may be given a (highest) priority level 4.

Calls which are emergency calls may be given a higher priority level than calls which are not emergency calls. Thus, for example, non-emergency calls may be prioritised in the priority levels 1 to 4 described above. Emergency calls may be prioritised as follows.

Emergency call starts may be given a priority level 5 higher than the highest priority level 4 for non-emergency calls. Existing emergency calls having an age less than Al may be given a priority level 6. Existing emergency calls having an age Jess than A2 but not less than Al rriay be given a priority level 7. Existing emergency calls having an age not less than A2 are given a (highest) priority level 8. Alternatively, all emergency calls may be given the same priority level which may be the highest priority level for non-emergency calls, or a single higher priority level.

FIG. 6 is a flow chart of another illustrative method 600 embodying the invention. The method 600 is an alternative to the method 500. In the method 600, congestion is detected in a step 601 by the BS, e.g. the 35 101, in one of the ways described earlier. In an optional step 611, the ES may estimate a current level of congestion in the system 100. The step 611 may be carried out by the congestion indication processor 215 of the BS. In response to step 601, the ES operates an algorithm in a step 603 to select calls and call starts randomly for termination. This algorithm may be operated by the congestion indication processor 215 or another processor of the BS. The ES may apply in the algorithm of step 603 a fixed probability to random selection of calls and call starts until congestion ceases. The probability applied may be the same for calls and call starts or it may be different for the two, e.g. the probability applied for call starts may be higher than that applied for calls. Alternatively, the BS may apply a varying probability for each of calls and call starts, e.g. in the following manner. The BS may apply a first probability P1 during a first time period t1 following detection of congestion in step 601, then a second probability P2]ower than P1 during a second time period t2 following tl, then a third probability P3 lower than P2 during a third time period t3 following t2 and so on, until congestion ceases. The time periods ti, t2 and t3 may he of equal length. Alternatively, at least two of these periods may have different lengths. The length of each of the time periods t1 to t3 may be fixed or may be selected dynamically by the BS according to current operational conditions in the system 100. The values of each of the probabilities P1, P2, P3 and so on may be different for calls and call starts. For example, the values of each of P1, P2 and P3 may be higher for call starts than for calls. The values of each of P1, P2 and P3 may be previously selected and fixed values which are recorded in the BS, e.g. in the memory 206, or may be values which are dynamically adjusted by the BS. These values may be selected to give a curve relating the number of terminated calls and/or call starts following detection of congestion in step 601, as a pre-determined function of time, e.g. to be or to approximate to an exponential decay curve.

In an alternative algorithm applied in step 603, the value of each of the probabilities P1, P2 and P3 may be different for each different call priority level. The different priority levels may be those described earlier. Thus, the probability may be low for a high priority level and the probability may be high for a low priority level.

In an alternative algorithm applied in step 603, where step 603 is applied following step 611, the ES may apply a probability to selection of calls and call starts to be terminated which matches the level of congestion estimated in step 611. Thus, there may be a high probability in step 603 that: each call or call start wifl be selected for termination if the level of congestion in the system 100 estimated in step 611 is high. On the other hand, there may be a relatively low probability in step 603 that each call will be selected for termination if the level of congestion in the system estimated in step 611 is relatively low. The probability may be dynamically varied by the BS to match the congestion level estimated in step 611. The probability applied for call starts in a given period may be different from, e.g. higher than, that applied for calls.

In a step 605 the BS terminates the calls and call starts selected in step 503. This operation may be undertaken by the controller 201 of the BS.

Eventually, in a step 607, the BS detects that there is no longer congestion arid, in a step 609, the BS ceases, e.g. by issue of a control signal from its controller 201, termination of calls and call starts.

A change in network topology or other operational change may occur in the system 100. Such a change may cause a sudden increase in congestion within the system 100. The method 500 or the method 600 may need to apply an adjustment to deal with the increase. For example, in operation of the method 500 where step 507 has been reached, one or more of steps 503, 505 and 507 may need to be repeated. Alternatively, where a reducing probability of selection 1s being applied by the algorfthm in step 603, the probability may need to be returned to a higher level and reduced again progressively from that higher level.

In the embodiments of the invention described above, the system 100 may be a trunked wireless system, e.g. a TETRA system as referred to earlier or an APCO 25 system as referred to earlier. The system 100 could alternatively be a system operating according to another protocol for communications, e.g. it may be an ad hoc In the embodiments of the invention described above resource allocation for calls in the system 100 is carried out in a distributed manner by the individual hosts, e.g. base stations, that are to participate in the call. It is not necessary for communication resources required for calls in the system 100 to be allocated by a system controller such as a zone controller as in a conventional TETRA system. However, the system 100, if a wide area system, may optionally include at least one controller which provides inter-zone trunking between geographical zones of the system.

However, such a controller does not need to allocate resources for individual calls within a zone. Thus design of the system 100 can be simplified since it does not depend on the complex operations of a system controller.

Furthermore, by use of the above described embodiments of the invention, changes in attachment, e.g. following a service attachment handover, of mobile stations between different serving base stations of the system as a result of movement of the mobile stations has a reduced impact on calls in progress since it is not necessary for new resources to be allocated by a system controller following the re-attachment.

Furthermore, by use of the above described embodiments of the invention, changes in topology of the system, e.g. following a service failure in one part of the system, has a reduced impact on calls in progress since it is not necessary for new resources to be allocated by a system controller following such a topology change.

The routers included in the system 100 need to forward calls and call starts according to priority levels indicated in the data packets comprising the calls and call starts. The routers in the system 100 may mark packets to indicate congestion, or in some cases may drop packets, using known procedures. However, the embodiments of the invention that have been described above allow good overall quality of communication service between originating and target hosts to be obtained by operations of the originating and target hosts without intermediate routers needing to know the state of each communication to make call resource reservations. Thus, in the embodiments of the invention, the routers may be operationally simpler than those envisaged in the prior art for use in a packet data network giving a guaranteed quality of service.

Claims (26)

1. A communication system for providing packet data communication including a first host, a second host and a communication network operable to provide a communication path between the first host and the second host, wherein the first host is operable to send a packet data communication to the second host via the communication path arid the second host is operable to receive a packet data communication from the first host, wherein each of the first host and the second host is operable to detect traffic congestion in the network along the communication path between them and, in response, to terminate a communication between the first and second hosts.

2. A system according to claim 1 wherein each of the first and second hosts is operable to detect congestion in the network by detecting a notification of congestion in a data packet header field of a received packet data communication.

3. A system according to claim 1 or claim 2 wherein each of the first and second hosts is operable to detect congestion in the network by detecting loss of data packets in a received packet data communication.

4. A system according to any one of the preceding claims wherein each of the first arid second hosts is operable to assign a priority to communications it is sending or is to send and to terminate a communication selectively according to the assigned priority of the communication.

5. A system according to claim 4 wherein each of the first and second hosts is operable to provide in each packet data communication it sends an indication of a priority of the communication in a data packet header

field of a sent packet data communication.

6. A system according to any one of the preceding claims wherein each of the first and second hosts is operable to detect a priority of communications it receives and to terminate a communication selectively according to the detected priority.

7. A system according to claim 6 wherein each of the first and second hosts is operable to detect a priority of a received communication by detecting an indication of priority in a packet data header field of a received packet data communication.

8. A system according to any one of claims 4 to 7 wherein the priority assigned or detected depends on an age of the communication.

9. A system according to claim 8 wherein communications have an assigned priority level selected from a discrete number of priority levels, wherein youngest communications have a lowest priority level and oldest existing communications have a highest priority level.

10. A system according to claim 9 wherein each of the first and second hosts is operable whereby in a first period in which congestion is detected by the host communications having the lowest priority level are terminated by the host, and in a second period following the first period in which congestion is detected by the host communications having a higher priority are terminated by the host.

ai. A system according to claim 10 wherein in each of a set of consecutive periods in which congestion is detected by the host communications are terminated by the host, the priority level of the communication terminated increasing from period to period as the number of periods when congestion is detected increases.

12. A system according to any one of the preceding claims 4 to 11 wherein emergency communications are assigned a higher priority than all non-emergency communications.

13. A system according to any one of claims 1 to 3 wherein each of the first and second hosts is operable to select communications randomly for termination with a fixed probability of selection for communications of a given priority level for each given period that congestion is detected by the host.

14. A system according to any one of claims 1 to 3 wherein each of the first and second hosts is operable to select communications randomly for termination with a probability of selection, for communications of a given priority level for each given period that congestion is detected by the host, as a function of time that is or approximates to an exponential decay curve.

15. A system according to any one of claims 1 to 3 wherein each of the first and second hosts is operable to estimate a level of congestion in the communication system and to select communications for termination with a probability of selection which depends on the estimated level of congestion.

16. A system according to any one of the preceding c'aims wherein the first host is operable to send a communication set up request message to the second host via the communication path between them and the second host is operable to detect a state of congestion of the communication path from receipt of the set up request message and to send to the first host a response message indicating the detected staLe of congestion.

17. A system according to any one of the preceding claims which is operable according to procedures compatibie with the Internet Protocol.

18. A system according to any one of the preceding claims wherein the system comprises a wireless communication system.

19. A system according to claim 18 wherein the system comprises a TETRA system, an APCO 25 system or an ad hoc network.

20. A system according to claim 18 or claim 19 wherein at least one of the first and second hosts comprises a base station.

21. A communication terminal operable as the first or second host in the system according to any one of the preceding claims.

22. A method of operation in a communication system for providing packet data communication between a first host and a second host via a communication network operable to provide a communication path between the first host and the second host, wherein one of the first and second hosts detects traffic congestion in the network along the communication path between them and, in response, terminates a communication between the first and second hosts.

23. A method according to claim 22 wherein communications are terminated according to a priority depending on an age of the communication.

24. A system according to any one of claims 1 to 20 and substantially as herein described with reference to any one or more of the accompanying drawings.

25. A communication terminal according to claim 21 and substantially as herein described with reference to any one or more of the accompanying drawings.

26. A method according to claim 22 or claim 23 and substantially as herein described with reference to any one or more of the accompanying drawings.