Note: Descriptions are shown in the official language in which they were submitted.
CA 02358413 2001-06-26
WO 00/41416 PCT/GB99/04304
1
CONCURRENT IP BASED VOICE AND DATA
SERVICES VIA WIRELESS NETWORKS
Field of Invention
The present invention relates to communication of audio information
in a wireless communications network - that is, a network having wireless
links between some communication devices (such as mobile telephones,
communications-centric PDAs or communications-enabled computers) and
network access nodes (commonly referred to as base stations in cellular
networks). Each network access node receives information from and
transmits information to local wireless devices.
Background
Today's advanced wireless networks, especially the digital cellular
networks such as the Global System for Mobile communication (GSM) and Code
Division Multiple Access (CDMA) networks, provide a variety of bearer
services available to the user, each with different characteristics and
costs of use. While such networks differ in many aspects, the present
invention is applicable to all of them and so can be described with
reference to any one - for example GSM.
GSM provides timeslots for both control information and users' calls
within its 8 time slots per radio channel architecture which is a
variation of Time Division Multiple Access (TDMA). A Channel in this
context is a particular operating frequency and parameters defining an
end-to-end transmission path at that frequency. Each time slot provides a
nominal capacity of 22.8kbps which includes the necessary channel coding,
which results in a nominal l3kbps for voice services and l2kbps for the
fastest data service. This latter l2kbps data rate is further reduced to a
nominal 9600bps by use of the Radio Link Protocol which provides extra
data error correction.
The so called ~~control channel~~ (which in GSM is one of the eight
time slots per channel, rather than a truly separate channel) is used by
mobiles and the network for call control, namely registering the users
active presence on the network when first turning on the telephone or
entering an area of network coverage, making or answering a call. Any
spare control channel capacity is used for low rate data services. Two low
speed or narrow band data services are known as SMS (Short Messaging
Service) or Unstructured Supplementary Service Data (USSD). The SMS and
USSR data services are, to all intents, low speed packet data services,
since the messages can be sent whenever the network has the spare capacity
to deliver them. The voice and high speed data services using a whole
CA 02358413 2001-06-26
WO 00/41416 PCT/GB99/04304
2
timeslot are known as circuit-switched services since a circuit is setup
from the user to the end-point, which may be another telephone in the case
of a voice call or another computer if a data call.
Users have typically used the voice services without particular
knowledge or care about the complexity of service provision. They simply
use the keypad to enter the desired telephone number and press the
appropriate PROCEED key, the telephone and network taking care of the
rest. Users often receive SMS messages from the network's voice response
systems in response to other users leaving messages when they have been
unable to get answers to their calls. More recently SMS, USSR and
equivalent services have been used to deliver information such as traffic
information to users who subscribe to such services. However, if users
have wanted to send messages they either needed a computer to interface to
the SMS or USSR access of the phone or use complex sequences of keystrokes
to generate and send such messages using the basic phone functions. If a
user wants to use the circuit switched data service then an integrated or
attached computer is necessary to handle the data applications.
Many data applications have been developed and used via GSM and
other networks using conventional Internet-based and other communications,
but the Internet Protocols (TCP/IP and UDP/IP) are becoming the defacto
standard. These are far from optimal for this use at the present time.
The Wireless Application Protocol (WAP) Forum is an industry forum
with an aim of delivering advanced telephony and information services to
users of mobile wireless devices such as phones, pagers, smart phones and
personal digital assistants. The wAP Forum has produced a set of
specifications to meet these aims and continues to complete and enhance
this task. The fundamental concept of WAP is to deliver services using
Internet-based technology, the user interacting with the phone and
associated services using a micro-browser, the information being delivered
by communication protocols similar to those of the Internet's Internet
Protocol (IP) and HyperText Transfer Protocol (HTTP). The wAP protocols
are known as WDP (Wireless Datagram Protocol) which is equivalent to the
Internet's UDP/IP, WTP (Wireless Transaction Protocol) which provides
acknowledged delivery and optionally segmentation and reassembly, and WSP
(Wireless Session Protocol) which is similar to HTTP.
There is additionally WTLS (wireless Transport Layer Security) to
deliver authentication and secure data delivery between WAP client and WAP
proxy (a proxy is a server computer acting as an intermediary between the
client and the communications network, and in particular the WAP proxy is
responsible for delivery of data content in a form which is independent of
the network's data communications protocol). The communications protocols
CA 02358413 2001-06-26
WO 00/41416 PCT/GB99/04304
3
have been designed to operate over the bearer services available in
today's networks, such as the narrow band SMS and USSD data services as
well as the higher data rate circuit-switched data services at 9600bps or
lower. Content is in the form of Wireless Markup Language (WML) and
wMLScript, which are based on the Internet's extended Markup Language
(XML), and HyperText Markup Language (HTML), and JavaScript respectively
(though WML is both a subset and superset of HTML and WMLScript is a
subset and superset of JavaScript). Thus applications and services can be
provided in WML or WMLScript without having to worry about which bearer
service is being used - such as SMS, USSD, or Circuit Switched. WML
content is in the form of decks of cards (collections of one or more
'cards', each comprising a fully specified piece of WML or wMLScript
content or function).
The following description of the present invention will refer to WAP
by way of example, but a similar approach could be adopted using
conventional TCP/IP or UDP/IP communications over an HTTP session and an
HTML, XML or XMLScript based application.
Now consider an enhanced voice based service, such as a voicemail
service as provided by IBM Corporation's DirectTalk products. when voice
messages arrive in the user's intelligent voice message service, the
messages are recorded and via some method irrelevant to this discussion
the callers' identities are obtained (calling line ID or voice recognition
are two options for this identification). The message service generates
the appropriate wML to describe the various callers' identities and call
control options (for example 'listen', 'discard', 'save', 'forward'), and
sends it to the user either when solicited (PULL model) or unsolicited
(PUSH model). The delivery could be via any of the bearer services
available to both the WAP client and the wAP proxy, which has the
responsibility of delivering content to users securely, efficiently and
reliably. The use of SMS or USSR has advantages since they can be used
while a circuit switched voice call is in progress and they do not require
a call to be set up. If the user responds to this message with call
options requiring a voice call to be set up then the wAP based phone has
all the capabilities to establish the call using the WML and internal
supporting WAP libraries. At any time the user can use the telephone's
micro browser interface to effect changes in the function of the service
required (STOP, REPEAT, etc) and to convey such command again via the
established SMS or USSR service.
To enable the user to work with an in-progress voice call, it is
necessary for the WAP-enabled phone to implement an interface to the DTMF
function (Dual Tone Multi Frequency signals are the signals generated by
pressing a telephone's touch keys in a ~~TOUChtone~~ phone) or some other
CA 02358413 2001-06-26
WO 00/41416 PCT/GB99/04304
4
signalling level function, the availability of which might be very
dependent on the network type being used or even the facilities provided
even if the network architecture catered for it.
If SMS, USSD or an equivalent service is being used for signalling
while the voice call is in progress then the user will be paying for both
services - the voice call connection and the SMS or equivalent service
used for control signals.
US-A-5799251 discloses a potential problem in radio telephone
systems in that Short Data Messages which are sent on the control channel
among control signals can block the control channel, causing interference
in the control signalling and potentially affecting the speech traffic. A
solution is then proposed which involves reserving a radio channel
specifically for transmission of user's data messages, this reserved
channel operating like a second control channel. A problem with this
solution is network resource usage, and associated cost to the user, since
the user requires an additional channel.
US-A-5790551 discloses dynamic assignment of an available packet
data traffic channel for transmission of packetized data based on which
channels will be free for a specified time period, the dynamically
assigned channel being separate from the data control channel. Thus, no
dedicated channel is required for data transmission, permitting more
efficient and flexible use of available communication channels.
Summarv of Invention
According to a first aspect of the invention, there is provided a
method for use in the communication of audio information between a
wireless communication device and a remote communication device via a
wireless communication network, the method comprising: at one of said
communication devices, providing audio information in digital form in
discrete packets and providing call control information in digital form in
discrete packets; establishing a call connection between the wireless
communication device and the remote communication device; and transmitting
both the discrete audio packets and the discrete call control packets via
said single call connection.
Transmitting audio information (voice or synthesized voice data) and
control information via a single logical end-to-end connection avoids the
need for a separate connection to be maintained throughout the call for
control information or for a separate connection to be established at
numerous times within the call. The invention avoids the need to use two
bearers throughout each call for each direction of communication between
CA 02358413 2001-06-26
WO 00/41416 PCT/GB99/04304
the communication devices and this reduces the cost of the user's call (in
particular, by avoiding the need for two connections between a mobile
device and a network access node) and increases the available network
capacity for other users. That is, by removing the need for concurrent use
5 of SMS or USSD low-bandwidth or narrow band bearer services or a separate
higher rate bearer to convey control information for the call, the
bearer's capacity may be utilised for other SMS and USSD message services
or to make additional capacity available for voice traffic and its
associated controls. Furthermore, reducing the demand for SMS or USSR type
bearer services to manage ongoing control of services reduces the
likelihood of needing to dynamically assign additional low rate data
service capacity.
As well as using less network resource than prior art solutions, the
present invention has additional advantages. Firstly, each call drains the
telephone's battery power more slowly if the invention is used, because
only one connection is maintained after call set-up. Secondly, in
embodiments of the invention in which call control information is sent in
packets over a circuit switched call connection, the latency is both lower
and more predictable than the latency of USSR or SMS messages. For
interactive voice applications, the long latency of SMS is unacceptable:
USSR is faster, but nevertheless has a latency which is longer and less
predictable than a circuit switched connection. Ensuring that voice data
and control data have the same latency is, of itself, beneficial since it
avoids voice data having to be held to await control information. For
example, if a user selects 'FORWARD' to next message, they want to be
moved to the next message without delay.
The invention goes against the general teaching of the prior art
which has dedicated channels or timeslots for voice data and for control
data and which uses for each of these data types a communications protocol
which has been optimised over time for the respective data type. In other
words, the focus of the prior art has been on optimising voice processing
and associated transmission protocols and otherwise enhancing the voice
services within the constraints of the available data capacity of the
respective channel architecture standards, without focusing on how to
reduce the cost of voice services. This can be understood in the context
of the available data communication capacities which have in the past
required considerable development effort merely to keep pace with
conflicting demands for high voice quality, comprehensive error
correction, etc.
According to a preferred embodiment of the invention, the step of
establishing a call connection follows the user of a calling device (for
example, a mobile telephone) requesting access to a messaging centre for
CA 02358413 2001-06-26
WO 00/41416 PCT/GB99/04304
6
use of interactive messaging services, which in turn follows either a PUSH
notification of new messages or a regular user or mobile device initiated
check (content PULL). Establishing the connection preferably comprises
sending an initial notification (for example, via SMS or USSD) from the
calling device to a called device (for example, the messaging centre),
informing the called device of the calling device's intention to establish
an interactive voice oriented session, and then the called device and
caller device establishing a connection (possibly after sending an
acknowledgement to the caller). Upon establishment of this connection, and
verification of identities, the devices notify each other that the new
channel will be used for subsequent transfers of control information. An
alternative and simpler method which involves the same steps except that
it omits the initial notification stage is equally acceptable.
After the connection has been set up, only a single call connection
is required for transmission of both voice data and associated ongoing
call control information. The voice data packets and control packets are
transmitted using a common transport protocol. This may be, for example, a
protocol based on current WAP protocols over IP using WML or WMLScript as
data content formats. At the end of the interactive session, the calling
device may indicate the end of the session or simply cease the connection,
one of these actions being predefined as the condition for resumption of
the use of the shared low bandwidth bearers for subsequent notification
and access.
Voice communication requires regular delivery of content, but humans
are also tolerant of minor disturbances in speech such as clicks and
'holes' where a small amount of information is not delivered. The
invention preferably transmits speech content and control information as
discrete packets via a single circuit-switched connection, allowing
regular delivery to be achieved with low overhead and complexity. A
similar approach over packet networks requires a little more control over
regular delivery and contention, to achieve error and packet loss
resilience, including some buffer control for voice while the control
information is small and can fill the gaps in buffer managed time.
The invention preferably uses 'voice over IP', which entails
digitising the voice information and sending it across the network in
discrete packets using the Internet Protocol (in contrast to the
traditional circuit-switched protocols of the telephone networks, but
nevertheless preferably using a circuit switched connection). However, the
invention is not limited to any particular implementation of 'voice over
IP' and includes any means of conveying suitably encoded voice over an
IP-based connection.
CA 02358413 2001-06-26
WO 00/41416 PCT/GB99/04304
7
In network architectures such as GSM which divide their
communications channels into multiple timeslots, the invention provides
most benefits if combined with an appropriate coding of voice information
which leaves spare capacity within a single time slot, or alternatively
using a non-standard data transmission rate which provides spare capacity,
such that voice and control information can be provided within the
capacity of a single timeslot.
In a second aspect of the invention, there is provided a wireless
communication device (such as a mobile telephone, or a PDA with
communications facilities) including: means for encoding audio information
in a digital form in discrete packets and means for generating discrete
packets of digital call control information; means for establishing a call
connection via an access node of a communications network; means for
transmitting both the discrete audio packets and the discrete control
packets to the local access node via said call connection; means for
receiving discrete packets containing audio information and discrete
packets containing call control information from the network access node
via said call connection; and means for decoding received audio
information packets and control information packets.
A third aspect of the invention provides a messaging centre for
providing voice-based communications services in a communications network
supporting wireless communications, the network including access nodes (or
base stations) for receiving signals from, and for transmitting signals
to, mobile telephones within their local cell, the messaging centre
including: means for encoding audio information in a digital form in
discrete packets and means for generating discrete packets of digital call
control information; means for establishing a call connection between the
messaging centre and a mobile telephone within the network; means for
transmitting the encoded discrete audio packets and discrete control
packets to the mobile telephone via said call connection; means for
receiving discrete packets containing audio information and discrete
packets containing call control information sent across the network from
the mobile telephone; and means for decoding received audio information
packets and control information packets.
The messaging centre preferably includes means for transmitting to a
wireless communications device via a wireless communications network a
first message including, within the message content, one or more menus of
selectable operations, and means responsive to receipt of a message, sent
from said wireless communications device in response to the first message
(either with or without user interaction), for establishing said call
connection for provision of a selected operation. The messaging centre can
thus provide a unified messaging service to deliver a number of different
CA 02358413 2001-06-26
WO 00/41416 PCT/GB99/04304
8
data content types (e-mail, voice, e-commerce transactions, database
lookup operations) with the requests made using selectable menus delivered
as WML, wMLScript or some other equivalent language, and the response to
such actions can be converted to a voice format using text-to-speech
technology for example.
Brief Description of Drawings
A preferred embodiment of the present invention will now be
described in more detail, by way of example, with reference to the
accompanying drawings in which:
Figure 1 is a schematic representation of a data communications
network according to the prior art; and
Figure 2 is a schematic representation of a data communications
network implementing the invention according to a preferred embodiment.
Description of Preferred Embodiment
A wireless communications network (i.e. a network including wireless
links) according to the prior art is shown schematically in Figure 1. A
mobile telephone 10 is shown communicating with a remote Messaging Centre
50, via two different transmission paths 70,80. The telephone communicates
with network 30 via a network access node 20. The Messaging Centre is an
example of a remote communication device to and from which voice data is
sent and received, which provides voice services such as voicemail
recording when a user's telephone is not contactable (e. g. switched off,
on another call, or out of range of a network access unit). Communications
between the mobile telephone and either the Messaging Centre or another
communications device involves a first connection being established for
transmission of control information in the form of SMS or USSD alert
messages via a Short Message Service Centre (SMSC) or Unstructured
Supplementary Services Data Centre (USSDC) 40. A second connection via a
different transmission path is then established in response to the alerts,
and this is used for the communication of voice data between the mobile
telephone and the Messaging Service or another communications device.
Communications with the Messaging service is via a conventional
Interworking Function (IwF) 60 which converts between the data transfer
protocols of fixed telephone lines and of wireless communications. As well
as being used for the alert messages to trigger establishment of the
second connection for transfer of voice data, this first connection will
be used for subsequent transfers of call control information. A wAP
communications environment 90 provides control data access to various
CA 02358413 2001-06-26
WO 00/41416 PCT/GB99/04304
9
bearers including SMS, USSD, etc. Conventional voice access (e.g. PSTN or
ISDN) is used for conventional access to messaging services.
This conventional arrangement has the problem that two separate
connections must be maintained throughout the call, even if there is spare
capacity on each connection. This use of network resources is typically
charged back to the user.
Figure 2 shows a communications network suitable for implementing an
embodiment of the present invention, in which a first control path 170 is
used for transfer of alerts between a mobile telephone 100 and a Messaging
Centre 140 via a Short Messaging Service Centre or Unstructured
Supplementary Services Data Centre 130. After an initial alert has been
received, or after a SMS or USSR response message has been returned, a
separate call connection 180 is established between the mobile telephone
and Messaging Centre via a Remote Access Server or Gateway 190 (described
below). The messaging centre is assumed to have a suitable data
communications environment for the delivery of notifications, menus to
control any options to manage the messages, etc, and this is assumed to be
wAP based.
Having established this separate connection, all subsequent session
control information as well as voice data is sent via this connection and
the connection used for transfer of alerts is no longer required. (Of
course, if the call connection terminates during the call, then a new
connection will have to be established and the communications path may be
different on reconnection - for example the mobile telephone may have
moved to a new cell of a cellular network. However this should not cause
any problems as the new connection would be accompanied by a new identity
(IP address) notification for the mobile telephone).
The transmission of voice and control data via a single connection
involves processing voice signals (digitizing, encoding including
compressing, and packetizing the audio data) such that the voice and
control data can be carried over a common communication protocol. The
replacement of the initial control channel by the single voice and control
connection is notified to both communicating devices as part of the
process of establishing the voice connection and terminating the initial
control connection.
Once the main interaction between mobile telephone and messaging
centre is no longer required, the call is ceased and all further control
passed back to the initial control path 170.
CA 02358413 2001-06-26
WO 00/41416 PCT/GB99/04304
The roles of the various components of a communications network
implementing the present invention according to the preferred embodiment
are as follows:
5 The Remote Access Server: provides the mobile telephone with access to IP
connectivity. The Remote Access Server provides a route between the mobile
network and thereby the mobile telephone, whether the actual
communications use packet based bearers or circuit-switched bearers, and
it can allocate the IP address of the client and the IP addresses of other
10 IP facilities such as Internet Symbolic Names resolution services and it
can additionally provide authentication services, thereby only allowing
authorised access.
The Short Messaging Service Centre: provides the message switching service
for the short message service offered by GSM and many other networks,
whereby short messages of some 150bytes can be sent from mobile telephone
to mobile telephone or a defined fixed location reliably, as SMS has an
optional store and forward mechanism which is used when either the mobile
telephone or fixed device is not available or has no current spare
capacity to receive the SMS.
The USSDC: provides a message switching service for USSR similar to that
provided by the SMSC for SMS. However, the nature of USSR, with less
reliability and no store and forward capability, permits simpler and
faster delivery than SMS.
The messaging centre: contains a service environment containing a voice
message store as a minimum and could encompass an entire unified messaging
environment in more complex embodiments. To access this messaging centre
from mobile telephones and other suitable calling devices requires
communication access which conventionally would be telephone or ISDN lines
etc, but in this case would be using packetized encoded voices as
described earlier using IP as the communications protocol to deliver the
packets of voice. The messaging centre also needs the control functions of
a WAP communications environment 150, or conventional TCP/IP and HTML
environment, which provides the means of notification and management of
the users interaction with the messaging services provided, for example
notifying which messages you have, whether old, or new and who from, etc,
options to listen, move to next, delete, save, forward, and many other
basic controls, and optionally more complex functions such as conversion
from/to voice to/from text, fax etc. Although the wAP gateway for protocol
handling of SMSC is shown as an integral component of the Messaging
Centre, it could equally be external of the Messaging Centre.
CA 02358413 2001-06-26
WO 00/41416 PCT/GB99/04304
11
Figure 2 also illustrates that the network infrastructure is
unchanged when using the existing InterWorking Function to convert between
the network specific protocols and those more commonly expected outside of
a mobile network, for example PSTN or ISDN. The modified IWF function 200
additionally would contain the IP voice packetization protocols necessary
for provision of more general voice services to mobile thereby removing
the conventional voice encoding and replacing it with just IP voice like
services over a number of bearer service options, e.g. conventional
circuit -switched or high speed general radio packet bearers
According to preferred embodiments of the invention, calling devices
such as the Messaging Centre thus provide means for performing an initial
notification (via, for example, SMS or USSD), followed by the caller and
called devices setting up a call in which voice data traffic and
associated ongoing control information can be handled by the one call. The
user thus requires only one call and the network has more SMS and USSD
capacity available for other users and services. From the user's
perspective, the cost of use of the service is only that of the voice call
during the period when the enhanced voice services are being used. For the
operator, the improved performance provided by this invention increases
customer satisfaction while the reduction in SMS or USSR traffic during
such periods, although reducing revenue if the bearer services are being
paid for rather than being amortised into the service costs, generates
better response for those still using the service within the network cell.
This benefits the other users and potentially avoids the costs of adding
additional channels or dedicated SMS, USSD capacity which would increase
costs without necessarily revenue.
Recent developments in voice and Internet technology have brought
about facilities for managing the delivery of voice based information
using Internet Protocol based communications. This involves sending voice
information in digital form in discrete packets rather than using the
traditional circuit-switched protocols of the telephone networks. Any of
the emerging packet based services may be used (such as GPRS being
deployed by GMS, CDMA Packet, or UMTS Packet, etc).
A specific mechanism for delivery of voice data over IP (referred to
hereafter as 'VOIP') was derived from the 'VOIP Forum', an effort by major
equipment providers including Cisco, VocalTec, 3Com, and Netspeak to
promote the use of communications protocol standard ITU-T H.323. This is
becoming the standard for sending audio and video using IP on the public
Internet and within intranets. The Forum also promotes the use of
directory service standards so that users can locate other users and the
use of touch-tone signals for automatic call distribution and voice mail.
CA 02358413 2001-06-26
WO 00/41416 PCT/GB99/04304
12
In addition to IP, VoIP uses the real-time protocol (RTP) to help
ensure that packets get delivered in a timely way. Using public networks,
it is currently difficult to guarantee Quality of Service (QoS). Better
service is possible with private networks managed by an enterprise or by
an Internet telephony service provider (ITSP).
A technique used by at least one equipment manufacturer (Netspeak)
to help ensure faster packet delivery is to ping all possible network
gateway computers that have access to the public network and choose the
fastest path before establishing a TCP socket connection with the other
end .
Using VoIP, an enterprise positions a "VOIP device" (such as Cisco's
AS5300 access server with the VoIP feature) at a gateway. The gateway
receives packetized voice transmissions from users within the company and
then routes them to other parts of its intranet (local area or wide area
network) or, using a T-1 or E-1 interface, sends them over the public
switched telephone network.
For VoIP, the voice information is sampled after suitable filtering,
most commonly 8-bit (or more) samples at 8K samples/second which results
in 64kbps or higher. This level of speech quality is consistent with the
norms of the telecommunications industry which normally uses 8K samples/s
with 8 bits/sample. The subsequent encoding in GSM brings the transmitted
data rate down to below l3kbps for full rate speech, or 6.5kbps for a
recently proposed coding and decoding of voice which achieves a
half-transmission-rate voice service with acceptable voice quality.
Framing this within IP packets of modest size, balancing the need to use
small packets of perhaps 64 or 128 bytes/packet to minimise loss of speech
and recovery, and allowing for overheads which would tend towards the use
of larger packet results in full rate speech of approximately 16-19 kbps
with packets of 64-128 bytes of data, or 8-9.5 kbps for half rate speech.
Standard full rate GSM speech coding and decoding has too high a
data rate to squeeze into current 9600bps data calls. However, with
increases in GSM's basic data speeds, IP encoded full rate speech can be
considered. However the option of using a half rate speech codec with a
9600bps or higher data circuit will allow sufficient spare data capacity
to provide low bandwidth data applications such as WAP enhanced services
to coexist in the same IP communications circuit. This is made easier to
achieve by the existence of natural quiet periods in speech, the presence
of which can be relied upon to some extent.
As noted above, protocol H.323 has been established for the delivery
of IP Voice based services. While more complex than is absolutely
CA 02358413 2001-06-26
WO 00/41416 PCT/GB99/04304
13
necessary for IP Voice in the case of GSM or other mobiles networks using
a point-to-point circuit switched call based form of connection it would
be much more relevant where GSM's GPRS (General Packet Radio Service) is
considered and hence this is one option for implementation of the
packetized voice delivery of the present invention.
Thus, the problem is solved in the following way for the enhanced
voice messaging service used above for illustration.
A mobile phone user has for some reason not been able to receive
some incoming calls. This might have been for one of a number of reasons,
such as: the phone was in use, turned off or not in coverage of the
network. The messaging center generates a PUSH alert to the user in the
form of a WML deck which is sent to the user via the wAP proxy 150 using
the default bearer service available, e.g. USSD. Having received the USSD
content the phone displays it via the micro-browser user interface. The
user can now decide whether to defer this notification or take some action
upon it. Assuming the latter, the user chooses to listen to the message.
By selecting the action LISTEN TO MESSAGE, the phone could optionally send
a response to the messaging service via the WAP proxy and must instruct
the phone to establish a data call to the Remote Access Server gateway
giving IP access to both the wAP proxy and the messaging service. Upon
establishment of the IP connection via the RAS gateway, a message is sent
to the messaging service to establish the new identity if the message was
sent earlier, and, additionally, to send the instructions to LISTEN if the
message was not sent earlier. Having set up all the identities and
instructions the messaging service can now start to provide the IP Voice.
At some point the user can interrupt with new instructions etc via the IP
data connection to the messaging center via the WAP proxy. At the end of
the session, the client can send one final message to the WAP proxy to
delete the current temporary RAS identity and to resume the default
connection, then disconnect the call to the RAS gateway.
To illustrate the ease with which the present invention can be
implemented, a set of changes which can be made to the components involved
in a WAP based service provision over GSM for implementing the invention
will now be described:
The mobile telephone can use existing voice processing functions for
both the transmission and reception processing (voice encoding and
decoding) to achieve the necessary raw data rate supported by this method
over the available connection bit-rate. The mobile telephone would
additionally need:
CA 02358413 2001-06-26
WO 00/41416 PCT/GB99/04304
14
1. Internet Protocol communications capability (e.g. UDP/IP) and PPP
based data protocol connection establishment in addition to any SMS or
USSR support for the wAP environment which would be required for the basic
services function. This is unlikely to be an increase on functional
requirement for the phone in many instances as the wAP environment will
ideally use IP for some services demanding larger amounts of data, e.g.
'over the air programming', or even basic services when SMS or USSR may
not be available.
2. the ability to packetize the encoded voice into IP packets for
transmission via the connection. For received information there is a need
to receive the packetized voice and to deliver this to the voice decoding
circuits of the phone.
3. the ability to multiplex the voice and data control packets. This
should be a normal IP function, but some buffering of the voice packets
would ideally be provided to permit the synchronously (regularly) sampled
voice to be sent plesiochronously (of a nominal rate but not
synchronously) over the network.
The wAP communications functions of the messaging centre, or even a
separate gateway, would not require any changes since the ability to
resume connections on different bearers is already defined.
The messaging centre would require no changes to its control
operation other than the support of WAP, which is typically required
regardless of whether the present invention is implemented. However, the
messaging centre would require the same function as the phone to encode,
decode, buffer and manage the packetization of voice traffic over IP.
The WAP Proxy is provided with the capability to support identity
updates/changes (this is catered for in the wAP specifications). The
routing between the messaging center, RAS gateway and preferably, though
less importantly, the wAP Proxy requires sufficient capacity to avoid
congestion if the voice over IP protocol being used over the mobile link
does not have the full capabilities of H.323 to buffer etc.
Throughout the examples above, it has been assumed that the
established Internet principles of port numbers are used to identify the
application, namely voice over IP, and therefore no explicit service
identification bits are required. However this invention recognises that
signalling bits within the IP packet conveying the encoded and packetized
voice could convey other useful information (such as quiet period
duration, encoding algorithms etc) in order to maximise the options and
performance over available bearer capacity.
CA 02358413 2001-06-26
WO 00/41416 PCT/GB99/04304
There are alternative ways in which the cost problem associated with
multiple dedicated channels could have been addressed, but each
alternative has its problems.
5 Firstly, it would be possible to use DTMF signalling within the
audio band. Simple choices could certainly be managed this way but complex
messages or further WML messages might be very inefficient. One potential
problem is that in some networks signalling during a call; such as DTMF,
is achieved using USSD with reconstruction in the network and thus no gain
10 is made, just an increase in complexity. As indicated previously such
features are network type and deployment dependent.
Secondly, the call could be split into two portions, one portion
being the conventional voice call and the other being a data call purely
15 for the conveyance of control information. At any time either the voice or
data call can be utilised, each with their bespoke communication format
and protocol, but only one at a time. This could reduce users' costs, but
it is more complex than the present invention and places greater demands
on network infrastructure resources even if the number of channels
required at any one time is the same as for the present invention. Whether
the user would then be charged for one or two calls is a network service
provisioning choice. Features aimed at supporting such concurrent voice or
data calls are emerging, but the required network features are complex and
not all networks will have this capability.
The invention has been described above as using an initial
notification via narrow-bandwidth bearers and than use of wide-bandwidth
bearers for IP voice services and ongoing control. Alternative
implementations which only use the packet based service are possible, but
these are less desirable in terms of costs of bearers, battery power
consumption, and network utilisation. A solution using packet based
services only would require the RAS server to be terminating IP based
communications or equivalent from the network infrastructure rather than
using circuit-switched calls.
