Title: Method and apparatus for sustaining conversational services in a packet switched radio access network
Abstract: In a packet switched system used for speech communication, a user entering a silent period, start to transmit silence descriptor information (SID). According to the invention, a speech user is not allocated the same resources for transmission of SID as for the speech communication. Instead, the user is, upon entering a silent period, reallocated to a SID communication channel, shared between a number of different users in silent mode. The resources used for speech may then be more advantageously utilized for other speech users, and is not occupied for SID transmissions only. In an alternate embodiment of the invention, when a mobile station is in silent mode, it receives and transmits associated signaling information on the same transmission resources as those used for the SID transmissions.
Patent Number: 6,898,195 Issued on 05/24/2005 to
| Inventors:
|
Molnö; Johan (Bromma, SE);
Balachandran; Kumar (Cary, NC);
Lindheimer; Christofer (Kista, SE)
|
| Assignee:
|
Telefonaktiebolaget LM Ericsson (publ) (Stockholm, SE)
|
| Appl. No.:
|
568452 |
| Filed:
|
May 9, 2000 |
| Current U.S. Class: |
370/329; 370/437 |
| Intern'l Class: |
H04Q 007/00 |
| Field of Search: |
370/329,330,336,337,341,437,433,435,347,389
|
References Cited [Referenced By]
U.S. Patent Documents
| 4949395 | Aug., 1990 | Rydbeck.
| |
| 5420851 | May., 1995 | Seshadri et al.
| |
| 5475869 | Dec., 1995 | Gomi et al.
| |
| 5519779 | May., 1996 | Proctor et al.
| |
| 5612955 | Mar., 1997 | Fernandes et al.
| |
| 5754537 | May., 1998 | Jamal.
| |
| 5812965 | Sep., 1998 | Massaloux.
| |
| 5940380 | Aug., 1999 | Poon et al.
| |
| 6438115 | Aug., 2002 | Mazur et al.
| |
| 6477176 | Nov., 2002 | Hamalainen et al.
| |
| 6480472 | Nov., 2002 | Jou et al.
| |
| 6519260 | Feb., 2003 | Galyas et al.
| |
| 6577862 | Jun., 2003 | Davidson et al.
| |
| 6631274 | Oct., 2003 | Keshavachar.
| |
| Foreign Patent Documents |
| 0 872 982 | Oct., 1998 | EP.
| |
| 1 006 695 | Jun., 2000 | EP.
| |
| WO 9624200 | Aug., 1996 | WO.
| |
| WO 9632817 | Oct., 1996 | WO.
| |
| WO 9835523 | Aug., 1998 | WO.
| |
| WO 9837706 | Aug., 1998 | WO.
| |
Other References
ETSI EN 300 963 V8.0.1, Digital Cellular Telecommunications System; Full Speech;
Comfort Noise Aspect for Full Rate Speec Traffic Channels. 1999. pp. 5-7.*
Bruhn. S. et al., "Continuous and Discontinuous Power Reduced Transmission of
Speech Inactivity for the GSM System", GLOBECOM 98, The Bridge to Global Integration.
IEEE Global Telecommunications Conference, 1998. Nov. 8-12, 1998. vol. 4 pp. 2091-2096.*
Guo, et al.; "Agressive Packet Reservation Multiple Access Using Signatures"IEEE
International Symposium on Personal, Indoor and Mobile Radio Communications,
Sep. 18, 1994; pp. 1247-1253.
EPO, Standard European Search Report, Nov. 10, 2000.
U.S. Appl. No. 09/568,451, filed May 9, 2000, Védrine.
U.S. Appl. No. 09/527,415, filed Mar. 17, 2000, Védrine.
|
Primary Examiner: Pezzlo; John
Assistant Examiner: Mills; Donald L
Parent Case Text
CROSS REFERENCE TO RELATED APPLICATION
This application is related to an application entitled "Method and System for
Fast Access to an Uplink Channel in a Mobile Communications Network" Ser. No. 09/527,415,
filed Mar. 17, 2000, and an applications entitled "Method and System for Fast Access
to an Uplink Channel in a Mobile communications Network" Ser. No. 09/568,451),
filed May 9, 2000. The applications are incorporated by reference herein.
Claims
1. In a packet switched mobile communication system where delay sensitive information
contains speech information, said delay sensitive information being transmitted
to a first mobile station on a first communication channel, a method for routing
data communications to and receiving data communications from a plurality of mobile
stations, said method comprising:
transmitting a silence descriptor (SID) on a second communication channel to
the first mobile station, wherein the SID is associated with the delay sensitive
information being transmitted on the first communication channel;
detecting the SID on the second communication channel and reallocating said first
mobile station to said second communication channel, whereupon after said reallocation,
said first mobile station then receiving a first block of said delay sensitive
information via said second communication channel; and
switching said first mobile station to said first communication channel for receiving
subsequent blocks of delay sensitive information associated with said first block
of said delay sensitive information received on said second communication channel.
2. The method of claim 1, wherein said first block of said delay sensitive information
includes a reallocation message for reallocating said first mobile station to said
first communication channel for receiving said subsequent blocks of delay sensitive
information associated with said first block of said received delay sensitive information.
3. The method of claim 1 wherein unused frames in said first communication channel,
originally designated for said first mobile station are filled with blocks of delay
sensitive information intended for a second mobile station.
4. The method of claim 1, wherein said second communication channel is primarily
utilized to sequentially transmit the SIDs to and from said plurality of mobile stations.
5. The method of claim 1, wherein said transmitting further comprises repetitively
transmitting SID information to each of said plurality of mobile stations on said
second communication channel.
6. The method of claim 1, wherein the step of transmitting SID information comprises
sending SID information on the second communication channel to each of said plurality
of mobile stations, wherein each of said plurality of mobile stations being in
a silent period triggers SID transmission.
7. The method of claim 1, further comprising:
transmitting Packet Slow Associated Control Channel (PSACCH) related information
on the second communication channel.
8. The method of claim 1, wherein:
the transmitting of the SID comprises transmitting the SID to a mobile station
of the packet switched mobile communication system.
9. The method of claim 1, wherein:
the transmitting of the SID comprises transmitting the SID to a base station
of the packet switched mobile communication system.
10. An apparatus for routing data communications to and receiving data communications
containing delay sensitive information from a plurality of mobile stations on a
first communications channel, said apparatus comprising:
detection means for detecting a silence descriptor (SID) associated with the
delay sensitive information intended for a first mobile station, in a second communication
channel;
means for reallocating the first mobile station to the second communication channel
following detection of the SID;
transmission means for transmitting a first block of delay sensitive information
to the first mobile station on the second communication channel prior to reallocation
of the first mobile station to the first communication channel, whereupon after
reallocation
means for transmitting subsequent blocks of the delay sensitive information associated
with said first block of delay sensitive information to the first mobile station
via the first communication channel.
11. The apparatus of claim 10, wherein the first and second communication channels
are downlink communication channels.
12. The apparatus of claim 10, wherein:
the first and second communication channels are uplink communication channels.
13. The apparatus of claim 10, further comprising:
a means for detecting a transmission of SIDs associated with the first mobile
station on one or more communication channels other than the second communication
channel; and
signaling means for notifying the first mobile station of a reallocation to the
second communication channel following the detection of a SID associated with the
first mobile station.
14. The apparatus of claim 10, further comprising:
a means for detecting transmission of said delay sensitive information on the
second communication channel.
15. The apparatus of claim 14, wherein the apparatus further operates to:
transmit a message to the first mobile station identifying the other communications
channel to which the first mobile station is to be reallocated.
16. The apparatus of claim 10, wherein transmission of the first block of delay
sensitive information, occurs in a block of time reserved for transmitting SID
information to a different mobile station.
17. The apparatus of claim 16, wherein the apparatus further operates to:
transmit SID information to the mobile station on the second communications channel
in a block that is not reserved for SID transmission to any mobile station.
18. The apparatus of claim 10, wherein the apparatus comprises a base station.
19. The apparatus of claim 10, wherein the apparatus further operates to:
periodically transmit Packet Slow Associated Control Channel (PSACCH) Information
and SID information on the second communications channel.
20. A method of communicating information in a cellular telecommunications system comprising:
receiving data communications containing delay sensitive information on a first
communications channel;
detecting transmission of a silence descriptor (SID) in a first communication
channel intended for a first mobile station;
reallocating the first mobile station to the second communication channel following
the detection of the SID associated with the first mobile station; and
sending a first block of delay sensitive information to the first mobile station
via the second communication channel prior to the reallocation of the first mobile
station to the first communications channel; and
sending subsequent delay sensitive information associated with said first block
of delay sensitive information to the first mobile station via the first communication
channel.
21. The method of claim 20, wherein the first and second communication channels
are downlink communication channels.
22. The method of claim 20, wherein:
the first and second communication channels are uplink communication channels.
23. The method of claim 20, further comprising:
detecting a transmission of SIDs associated with a plurality of mobile stations
on one or more communication channels other than the second communication channel;
and
reallocating the plurality of mobile stations to the second communication channel
following detection of the SIDs that are associated with each of the plurality
of mobile stations.
24. The method of claim 20, further comprising:
detecting transmission of delay sensitive information with intended for the first
mobile station, on the second communication channel; and
responsive to the detection of the delay sensitive information, reallocating
the first mobile station to the first communication channel.
25. The method of claim 24, further comprising:
transmitting a message to the first mobile station identifying the other communications
channel to which the first mobile station is reallocated.
26. The method of claim 20, wherein
the transmission of the first block of delay sensitive information to the first
mobile station occurs in a block of time reserved for transmitting SID information
to a second mobile station.
27. The method of claim 26, further comprising:
transmitting SID information to the second mobile station on the second communications
channel in a block that is not reserved for SID transmission to any mobile station.
28. The method of claim 20, further comprising:
periodically transmitting PSACCH information and SID information on the second
communications channel.
Description
BACKGROUND OF THE INVENTION
The present invention generally relates to wireless communication. More specifically,
the invention relates to a method and apparatus for multiplexing real-time users
in a packet switched radio communication system.
There is presently ongoing a paradigm shift in telecommunication. Historically,
the telecommunications industry has been focusing on voice communication over fixed
lines or radio communication links like, e.g., cellular telephony systems like
Global System for Mobile communication (GSM). Communication has typically been
transmitted in a circuit switched manner, i.e., with dedicated connections between
users or end nodes. Circuit switched communication requires continuous allocation
of physical transmission resources, or communication channels, for the whole duration
of a connection, regardless of the actual use of the connection.
With the explosive growth of Internet traffic however, the focus has shifted
towards more efficient ways of transferring data communication in a telecommunication
network. Packet switched communication protocols has been developed, e.g., General
Packet Radio Service (GPRS) to be used together with GSM and the Time Division
Multiple Access (TDMA) system compliant to the TIA/EIA-136 standard. The advantage
with these packet switched communication protocols is that there is no need to
have physical transmission resources reserved for users that are not making use
of it. For example, a user may share a transmission resource with one or several
other users and occupy the resource only when there is user data to send. If there
is no data to send during certain periods, other users may utilize the transmission
resources. This is a more efficient way of allocating users onto physical channels
than the circuit switched strategy, where a user is always a sole owner of a communication channel.
With the identification of packet switched methods as being an efficient way
of transferring data, the next step is basically a step back. The focus is again
on voice, but it is also a step forward in that the aim is now set on voice over
packet switched communication, or more generally, real-time services over packet
switched communication channels. With this and other aims, there will be a large
variety of services carried over packet switched communication channels, services
with completely different demands in terms of delay, delay variations (jitter)
and error rates. For example, a web browsing session may not suffer seriously from
being slightly delayed, it is however important that the transfer is error free.
For voice communication, it is basically the other way around; a voice conversation
is extremely sensitive to delay and delay variations but may perhaps tolerate a
non-zero error rate and still provide acceptable quality.
In the Universal Mobile Telecommunications System (UMTS), there are four proposed
classes defined to further characterize different services and the respective Quality
of Service (QoS) demands: conversational, streaming, interactive and background.
One main distinguishing factor between these classes is delay related. The conversational
class is intended for delay sensitive traffic, such as speech, while the background
class is the most delay insensitive class. Conversational and streaming classes
are intended to be used to carry real-time traffic flows and interactive and background
classes are intended to carry, e.g., Internet applications like WWW-browsing, file
transfer and e-mail services.
As voice communication involves constraints on delay, it does not tolerate the
sharing of a transmission resource, or physical channel, as liberally as the fundamentals
of packet switched communication allow. It is necessary to introduce priority for
voice users over, e.g., a background user on the same channel, such that the real-time
aspects of the voice connection may be maintained.
In an exemplar voice call, there are typically periods of silence in one direction
when the other direction speaks, and vice versa. With circuit switched radio communication
connections, it is possible to utilize these silent periods and decrease the output
power from the transmitter while a voice stream from a speech coder is paused.
This will mean a system gain in terms of less interference. The physical communication
channels, e.g., in terms of frequency, timeslot or code is however still occupied.
There may however be even more to gain if other users could be multiplexed onto
the same physical channel during these speech pauses. By using packet switched
methods for transferring voice communication of the conversational class, it will
become possible to more efficiently make use of the transmission resources while
in a period of speech silence. One way to do this is to allocate the resources
to a best effort user, e.g., of the background or interactive class, while in a
silence period and maintain the high priority for the conversational class user.
Thus, it will be easy to, as soon as a silence period is interrupted by a speech
period, prioritize allocation of the conversational class again. With this flexible
method of allocating shared resources, it will be possible to allocate more users
than the number of available transmission resources or channels. If there is a
high number of transmission resources, it may even be possible to allocate more
voice users than the number of channels, assuming that it is highly unlikely that
all users need transmission resources at the same time. This strategy is usually
referred to as statistical multiplexing.
For the Adaptive Multi-Rate (AMR) speech coder structure of GSM, as in many other
speech coders designed for circuit switched connections, the silent periods discussed
above are not completely transmission free, i.e., the transmission resources are
still utilized. During a silent period, when no speech is processed, the speech
coder generates what can be referred to as a Silence Descriptor (SID). This silence
descriptor is transmitted according to some repetition rate in order to generate
"comfort noise" in the receiving end. It is typically the case in a voice communication
that there is no complete silence, and to "simulate" the noise usually present
in the surroundings of the silent speaker, SIDs are transmitted with a certain
repetition rate. The SIDs defined for circuit switched speech are traditionally
transmitted on the same physical resource as the regular voice communication.
If a packet switched system is considered, the silent periods should optimally
enable allocation of other users onto the physical communication channel, e.g.,
background or interactive class users. It would of course be possible to do this
and still transmit SIDs from a conversational user also. However, if one consider
utilizing a communication channel for more than one conversational user in one
way or another (e.g., statistical multiplexing), the SID transmissions that are
continuously repeated with some repetition pattern will pose a problem, since a
continuous allocation for e.g., another real-time user will be impossible. There
is thus a need to develop and prepare techniques to more efficiently allocate resources
and allow a more flexible scheduling, than what is possible with the presently
used SID techniques.
SUMMARY OF THE INVENTION
In one aspect of the present invention, allocation of conversational users that
are in a silence period is made on a single predetermined communication channel.
All users that are allocated on one of a certain number of predetermined channels
for traffic communication and that are in a silent period are re-allocated to the
single predetermined communication channel for SID transmissions. Thus, the resources
on the channels used for e.g., conversational transmission is not used for SID
transmissions, but may instead be utilized by allocating another user thereon.
As soon as a user enters a silent period in one direction, a reallocation to the
shared SID transmission resources takes the user away from the resources used for
the, e.g., conversational transmission.
In another, aspect of the present invention a Packet Slow Associated Control
Channel
(PSACCH) is allocated to share transmission resources with the SID transmissions.
The PSACCH is allocated in a certain repetition pattern on the same physical communication
channels as the SID transmissions, such that users receiving SID transmissions
also receive PSACCH transmissions with a certain repetition rate.
In another aspect of the present invention, when leaving a downlink (in the direction
from the base station to the mobile station) silence- or pause period, the first
data block, e.g., containing speech data, is transmitted together with a channel
allocation on the same communication channel, utilizing the same transmission resources
as for the previously mentioned SID and PSACCH transmissions. This will advantageously
handle allocation delays in the downlink, that would otherwise be introduced in
the beginning of an active period. The SIDs associated with other users, that would
normally be transmitted on the transmission resources that instead is used for
a first data and allocation block are delayed until transmission resources are
available again on the "SID PSACCH" communication channel.
In yet another aspect of the present invention, downlink transmission resources
are made available, such that, when stealing SID resources for a first data and
allocation block, the SID is displaced and transmitted during a sequence of one
or more periods on the same channel, that are not part of any repetitive SID or
PSACCH transmission.
In yet another aspect of the present invention, uplink allocation of SID transmissions
on a shared resource is made by assigning to a user in a silent period, a periodic
repetition starting from a certain frame number and optionally a frame number offset.
BRIEF DESCRIPTION OF THE DRAWINGS
Features, objects and advantages of the present invention will become apparent
to those skilled in the art by reading the following detailed description where
references will be made to the appended figures in which;
FIG. 1 illustrates an exemplary cellular system where a base station is responsible
for providing radio communication to a certain geographical area.
FIG. 2 illustrates a timeslot and frame divisioning according to GSM specification.
FIG. 3 illustrates how SID blocks from different users in the downlink are multiplexed
onto one common SID communication channel and how users leaving silence periods
are allocated to speech communication resources according to one embodiment of
the present invention.
FIG. 4 illustrates procedures in the mobile station when in a downlink silent
period on the SID communication channel, according to one embodiment of the present invention.
FIG. 5 illustrates procedures in the network side for the downlink SID communication
channel according to the present invention.
FIG. 6 illustrates how SID blocks from different users in the uplink are multiplexed
onto one common SID communication channel according to one embodiment of the present invention.
FIG. 7 illustrates procedures in the mobile station when in an uplink silent
period on the SID communication channel, according to one embodiment of the present invention.
FIG. 8 illustrates a SID communication channel where PSACCH transmissions are
multiplexed with SID transmissions according to another embodiment of the present invention.
FIG. 9 illustrates an exemplary cellular system and a mobile station including
means for supporting SID and PSACCH allocations according to the present invention.
DETAILED DESCRIPTION
The invention will now be described making references to a GPRS/EGPRS based cellular
packet data communication system and extensions and variants thereof, as briefly
described in the background. It should be understood that the invention is not
limited to these types of systems. In general, all TDMA based systems where real-time
applications are transmitted over a packet switched channel may be considered.
FIG. 1 illustrates an exemplary cellular system where a base station,
12,
is responsible for providing radio communication possibilities to a certain geographical
area. A mobile station,
10, may move between areas served by different base
stations and communicate with different base stations,
14, dependent upon
its position. In FIG. 1, a cellular pattern based on 3 frequencies or 3 frequency
groups are illustrated. F
1, F
2 and F
3. This is usually referred
to as a 3 reuse pattern and indicate that there is a 3-repetition of frequencies,
or, that a frequency is only used once every third cell. There are of course other
ways to arrange the distribution of a number of frequencies for a group of cells,
e.g., with other repetition patterns like
1,
9,
12 and
21.
FIG. 2 illustrates the time division specified for a GSM frequency. Every GSM
frequency is divided into 8 different timeslots, where each timeslot form a communication
channel for a circuit switched GSM connection. The downlink and uplink are identical
in timeslot divisioning but separated in frequency. At least one timeslot on one
frequency in each cell, e.g. TN
0, is, in the downlink direction, allocated
for common control channel and broadcast transmissions. The corresponding uplink
timeslot is usually used for random access. i.e., a way to make the system aware
of a mobile user's request for a communication channel. This is illustrated in
FIG. 2, where timeslot
0, TN
0, is shaded. The different timeslot
transmissions are repeated in what is referred to as frames. There are several
important frame repetition patterns in GSM, one being the 52-frame pattern illustrated
in FIG.
2. The chronological order at which transmission occur is, e.g.,
frame
0 (TN
0-TN
7)-frame
1 (TN
0-TN
7)-frame
2 (TN
0-TN
7).etc. A speech frame in GSM is 20 ms. When the
speech coder and the channel coder have processed the speech frame the resulting
number of bits corresponds to what can be carried over the air in four full bursts
on a time slot.
The packet switched modes designed for GSM, -GPRS and EGPRS, are in many ways
similar to GSM. For example, the timeslot structure and frame divisioning are identical
to GSM for the traffic channels. This means that much of what is specified for
voice transmission over GSM circuit switched communication channels, may also hold
for voice transmission on over GPRS packet switched communication channels. For
example, a speech period may still be 20 ms. A 20 ms period in GPRS corresponds
to the transmission of four frames and for one timeslot in GPRS and EGPRS this
is often referred to as a block period, or block for short.
There are several different speech codecs developed for GSM, the most recent
being the Adaptive MultiRate (AMR) speech codec. This speech codec has an adaptive
output of speech information bits. When these bits are combined with an adaptive
portion of channel coding the total number of bits adds up to a constant number,
i.e., constant gross rate. For example, with a very good channel quality, there
is no need to use a large amount of channel coding bits. The transmitted bits may
instead comprise a larger amount of speech information. For a poor quality, more
protection is needed, and channel coding bits are required, at the cost of speech
codec information. However, the gross bit rate over the air, i.e., speech codes
information+channel coding bits, does not vary with the AMR codec in GSM. This
quality adaptation may be performed on speech frame basis and results in higher
perceived quality.
A similar strategy to the GSM AMR is also being considered for voice communication
over a packet switched channel, like a GPRS or EGPRS channel.
Considering now the SID transmissions for voice connections in silent
mode for GSM AMR as an exemplary case. In the AMR coder, the SID occupies every
eighth block. This means that, for circuit switched voice communication, 7/8 of
a communication channel (i.e., timeslot) is unused during silent periods. In a
packet switched communication system, it would be advantageous to utilize the resources
for other users during this 7/8 of the time. A user, U
1, engaged in a voice
connection, is allocated downlink transmission resources on timeslot number
3,
TN
3, for transmission of speech information. When user U
1 enters
a period of silence in the speech flow, for example after a sentence, awaiting
an answer in the other direction, a Voice Activity Detector (VAD) detects the start
of a silent period. In GSM, this VAD triggers a decrease of output transmission
power on TN
3, since there is no speech information to send. The transmission
of SIDs also starts with the VAD indication of a silent period. For GSM AMR, the
SID blocks are transmitted with a 160 ms repetition cycle.
According to one aspect of the present invention, in order to enable, e.g.,
a GSM AMR like approach a so for GPRS, a packet switched SID communication channel
is defined. The packet switched SID communication channel has similar functionality
to the circuit switched SID channel in GSM, although with a multiplexing method
that intelligently utilizes the more liberal multiplexing techniques allowed in
a packet switched system.
The downlink direction of the SID communication channel is illustrated as TN
0
in FIG.
3. According to the invention, the SID transmissions are not transmitted
on the same channel as the speech blocks. Instead, SID transmissions, e.g., SID(
1)
for user U
1, are transmitted on a separate SID communication channel. This
is illustrated in FIG.
3. Thus, upon entering a silent period by receiving
a first SID(
1) on TN
3, user U
1 is reallocated to the SID communication
channel on TN
0 and may there start to receive SID blocks periodically. The
advantage of re-allocating a silent user, is that TN
3 in the downlink becomes
available for other continuous communication flows. For example, if a user U
2
is about to leave the downlink SID channel upon entering a speech period, user
U
2 may be allocated to TN
3 for its next speech period. This is illustrated
in FIG. 3, showing that user U
1, after receiving the first SID(
1)
on TN
3, switches to the SID communication channel TN
0 while user
U
2, ending a silent period, receives a first speech block on the SID communication
channel containing a reallocation indication to TN
3.
FIG. 3 illustrates an exemplary SID communication channel that handles N-m different
users in silent periods simultaneously. N corresponds to the repetition rate of
SID transmissions, e.g., in GSM with AMR speech codec, 160 ms. m corresponds to
a number of blocks that are not allocated in any repetitive SID transmissions.
The m spare blocks are reserved for use when a user leaves a silent period.
For example, when user U
2 starts to receive speech data in the downlink
again after a silent period, there is a need to allocate resources for speech information
transmission. According to one embodiment of the present invention, a base station
transmits the first speech block after a silent period on the SID communication
channel, as illustrated in FIG.
3. An MS in a downlink silent mode always
listens to the SID communication channel and thus receives the first speech block
intended for it. The first speech block after a silent period may be coded with
less speech information bits, if necessary, to allow transmission of an allocation
indication included in the first speech block transmitted on the SID communication
channel. To be able to allocate resources for the second and following speech blocks,
the first speech block may contain a message, such that a user U
2 is informed
where the following downlink speech block will be sent. After having received the
first speech block on the SID communication channel, the user leaves this channel
and moves to the channel indicated in the allocation message, in this example TN
3,
where subsequent blocks may be received.
By transmitting the first speech block, along with an allocation indication,
reserved
resources for SID transmissions may be "stolen". Since a first speech block advantageously
is transmitted without any delay, it is possible that it "steals" resources from
soma other users SID transmission now. For example, user U
2, listening to
SIDs SID(
2) may receive a first speech block in a position usually used
for SID transmissions to a user U
3, receiving SID (
3). This is handled
such that any of the last m spare blocks in the repetition pattern may take care
of the SID (
3) transmission instead, since this is not as delay sensitive
as the first speech block to user U
2. The last m blocks are thus used for
transmitting "stolen" SIDs. Occasionally, the first speech blocks will be transmitted
in one of these m blocks, if the downlink communication resumes during that period.
Note that when a speech user is reallocated from a speech communication channel,
it may also receive its first SID block on the speech communication channel.
The allocation messages, although illustrated for the downlink in FIG. 3, may
also reallocate uplink communication.
FIG. 4 illustrates the procedure to be followed on the mobile terminal side
when the downlink is in a silent period, and is therefore allocated on a downlink
SID channel. In
42, the mobile decodes all the SID blocks, even those intended
for other users. The mobile only updates its SID information when blocks with a
matching address are decoded. As long as the decoded blocks are identified as SID
blocks, this procedure continues,
44. At one point, when the silent period
for a connection is terminated in the downlink, the system will send, in the first
occurring SID position, a first data (e.g. speech) block in the downlink, together
with a channel allocation description,
46. A mobile will make a check in
an address field to identify if the speech block has a matching address or if it
is intended for another user. If it is intended for another user, the mobile ignores
the block and the procedure continues. If the address is matching with that of
the mobile, it will leave the silent mode, decode the speech block and move to
a new communication channel, e.g., timeslot/frequency as indicated in a channel
allocation description, and continue to receive downlink speech blocks on the new
channel,
48.
FIG. 5 illustrates the network side of the downlink SID allocation according
to the present invention. A scheduler continuously receives SID information,
50,
and schedules this on the SID communication channel according to a predetermined
repetition pattern,
52. If there is no "first speech block" to any of the
users in silent mode on the SID channel, the scheduler will not map any information
to the last m spare blocks in the repetition cycle,
54. If there is a "first
speech block" arriving, this will be immediately scheduled and transmitted and,
if an outstanding SID is deprioritized in this procedure, this SID will be sent
in one of the m spare blocks,
56.
With the above described handling of the SID transmissions, it is possible to
multiplex speech users in the downlink in a more efficient way than if the SID
transmissions were allocated in the same way as for circuit switched voice communication,
-on the same communication channel as the speech blocks. Utilizing the present
invention, implementation of statistical multiplexing for real time users will
be facilitated.
Turning now to FIG. 6, an exemplary uplink SID allocation corresponding to
the downlink described for FIG. 3 is illustrated. The uplink SID communication
channel does however not include any scheduled spare blocks, corresponding to the
m spare blocks described for the downlink. This is explained by the fact that the
uplink resources are allocated to a number of distributed mobile users and it is
difficult to control and redirect uplink SID transmissions to other blocks than
those reserved for a certain user. After having ended a speech period in the uplink,
a user U
1 moves to the common uplink SID communication channel where the
periodic transmission of SID (
1) starts. The periodicity of the SID transmissions
can be indicated in an allocation message at call setup, or in the SID allocation
message transmitted in the beginning of a silent period. Note that the allocation
messages, although indicated for the uplink in FIG. 6, may also reallocate downlink communication.
FIG. 7 illustrates the MS procedure for uplink handling of silent periods on
the SID communication channel. The SID number indicates which SID block the user
is allocated when in a silent period. When the silent period has started, the transmitter
part of the MS analyzes the content,
70, of the output from the speech codec
and if it contains SID, the SID is sent on the allocated SID channel block,
72.
(Note that the first SID in the uplink is transmitted together with a SID reallocation
request on the speech communication channel, e.g., TN
3 as illustrated in
FIG. 6.) if the content is a "first speech block" after a silent period, the MS
may immediately request and uplink allocation
74 via, e.g., a random access
channel. After the uplink allocation is given, the MS initiates uplink transmission
on the new uplink communication channel,
76.
For the uplink SID communication channel, the network receives SID transmissions
according to the scheduled allocation and forwards them to respective packet switched
connection end users or nodes. Should the network receive an access request from
a user in a silent period, top priority is given to an uplink allocation for that
user. The access burst may be sent on a random access channel which may either
be allocated on another physical channel than the uplink SID communication channel,
or alternatively share resources with the SID communication channel. It should
be noted that since it is the base station that determines the allocation even
for the uplink, there may be some additional delay experienced when leaving the
uplink silent period.
According to another aspect of the present invention, the SID transmission
resources may be shared with transmission of signalling, or control, information.
In the present GSM system, a signalling channel called Slow Associated Control
Channel, SACCH, is defined. This channel is typically used for, e.g. transmission
of measurement reports and system information messages. It is defined for both
the up and the downlink. A variant of this circuit switched SACCH is also needed
for a real-time session over a packet switched connection. This packet switched
SACCH is hereafter referred to as PSACCH.
In a spent period, it is possible to allow the SID transmissions to share transmission
resources with the PSACCH transmissions to or from a mobile station. For example,
in GSM AMR, the normal rate for SID blocks is 160 ms, i.e., SID information is
updated each 160 ms period. This high repetition rate is important since the SID
blocks also contain adaptation information for the AMR codec. The normal period
for a SACCH transmission is 480 ms, and this could also be used for a PSACCH. With
the described SID communication channel. It is possible to allow for a maintained
PSACCH transmission period and lower the update rate of SID information to alternating
periods of 160 and 320 ms. This is illustrated in FIG. 8 where each 3
rd SID
transmission is replaced with an PSACCH transmission (It should be noted that other
ways of allocating PSACCH every 3
rd block are possible and FIG. 8 only
shows an exemplary alternating method). Thus, according to this aspect of the invention,
the PSACCH transmissions does not need to be allocated any additional transmission
resources than those used for SID transmissions. The blocks carrying the PSACCH
information can also hold AMR adaptation information for the purpose of a continuous
AMR update.
FIG. 9 illustrates a GPRS communication system according to several embodiments
of the present invention. A real time user
90 may communicate with the fixed
part of the cellular system via a serving base station
92. The serving base
station may host a scheduler,
94, for the physical channels utilized in
the base station's coverage area. The base station is connected to a a Base Station
Control node, BSC,
96, which in turn is connected to a Serving GPRS Support
Node, SGSN,
98, serving one or several BSCs. The SGSN is typically the node
controlling the packet flow to and from the different base stations, via the BSCs.
Another GPRS support node is a Gateway GPRS Support Node,
99, connected
to e.g., internet or other external networks (not illustrated). The scheduler,
94, is in this example located in the base station, but may alternatively
be housed in any other network node. Scheduling functionality may also be split
between different nodes in the system. The scheduler, 94, will control the allocation
of SID blocks onto a SID communication channel. Different users, carrying different
temporary identities, will be allocated to repetition patterns according to the
methods described previously. For the downlink, the scheduler,
94, will
schedule SID transmissions and, at occurrence, replace SID transmissions with a
"first speech block" transmission instantly upon arrival. The scheduler,
94
will also allocate resources from the m spare blocks to send outstanding SID transmissions.
Similarly, in mobile station
90, transmission means
91 are included.
The transmission means
91 is responsible for transmitting SID blocks in
the uplink according to the repetition pattern indicated by the network for uplink
SID transmissions. When a SID and an PSACCH information flow is multiplexed according
to one embodiment of the present invention, the transmission means
91 is
responsible for transmitting the PSACCH and SID transmissions according to the
allocation scheme as indicated by the network.
Additionally, according to one aspect of the present invention, the
scheduler,
94, include means for allocating resources to PSACCH transmissions
on the SID communication channel at a certain repetition rate, and replace SID
transmissions to users in silent periods with PSACCH transmissions.
Although the present invention has been described with examples from a packet
switched communication system compliant to the GPRS/GSM specifications, it should
be noted that the solutions presented is equally well applicable to any other packet
switched data communication system with the same or similar structure and functionality.
The specific embodiments should therefore be considered exemplary rather than limiting
the scope of the invention. The invention should rather be defined by the following claims:
*
