1.Dual Transfer Mode within GSM and GPRS
“Dual
Transfer Mode” (DTM) is also known as "BSS co-ordination of Radio Resource
allocation for class A GPRS services - GSM Radio Access " and "BSS
co-ordination of Core Network Resource allocation for class A GPRS services
-GSM-UMTS Core Network".
The definition of GPRS class A mode of
operation in Releases 97 and 98 was done assuming a total independence between
the CS and PS domains. Thus the direct implementation of the standards would
have resulted in mobile stations required to operate in two different
frequencies simultaneously. The Dual Transfer Mode overcomes this by allowing a
simultaneous existence of circuit-switched connection and packet-switched
sessions within GSM/GPRS.
This is done by "associating" one
or more packet-switched timeslot(s) to the circuit-switched timeslot, or even
by sending packet-switched data and circuit-switched voice in the same
timeslot, i.e. when half-rate speech is used.
To perform this association, the Base
Station Controller (BSC) needs to know the IMSI of the Mobile Station.
Consequently, in a similar manner to UMTS, the A interface is modified so that
the BSC can be made aware of the IMSI associated with each SCCP connection
towards the MSC. With this information, the BSC can do the co-ordination of
resources allocated to the circuit-switched and packet-switched domains.
2.General Packet Radio Service enhancements
Also called “General Packet Radio Service (GPRS) - radio part (R99)”.
GPRS was introduced in Release 97. Release
99 brings a number of essential corrections e.g. on Downlink Power Control, Timing
Advance procedure, Cell reselection delay time, PBCCH scheduling, PTCCH
mapping, etc.
A number of
specifications were impacted, the most important among them being listed in the
table above3.GSM on 400 MHz Frequency Band
"GSM in the 400 MHz bands" allows
to deploy GSM within frequency bands in the range 450 to 500 MHz,
previously used for analogue cellular systems.
When GSM in the band 1800 MHz was
added to the standard in Release 96, some modifications were done to make the
set of specifications less band-dependent. As a continuation for this process,
the possibilities to further adapt the specifications toward band free format
were considered where appropriate in this work item.
GSM 400 systems are targeted to offer large
coverage in rural areas. The work item “GSM in the 400 MHz bands” includes a
support for cell ranges up to 120 km by enabling two more bits in “Timing
Advance” Information Element.
GSM 400 systems are specified for two
frequency allocations. Primary utilisation will be allocations around 450 MHz.
For some countries allocations around 480 MHz is possible.
The two bands are:
- GSM 450 Band
- 450.5 – 457.5 MHz: mobile transmits, base receives;
- 460.5 – 467.5 MHz: base transmits, mobile receives;
- GSM 480 Band
- 479 – 486 MHz: mobile transmits, base receives;
- 489 – 496 MHz: base transmits, mobile receives;
with a carrier spacing of 200 kHz.
Consequently,
if we call F1(n) the nth carrier frequency in the lower band, and
Fu(n) the nth carrier frequency in the upper band, we have, for the
450 band:
- Fl(n) = 450.6 + 0.2*(n-125) (MHz) (125 £ n £ 159)
- Fu(n) = Fl(n) + 10 (MHz)
The value n is the Absolute Radio Frequency
Channel Number (ARFCN).
It is reasonable to assume that BTS heights
in rural area are higher than in urban area thus minimum coupling loss (MCL) 65
dB is valid assumption in rural areas.
The adoption of GSM 900 or EGSM 900 radio
frequency requirements to GSM 450 systems with minimal changes made it easy to
adapt standard GSM technology.
It is unnecessary to do any changes to
existing GSM 900 modulation mask while it is adapted to GSM 450 systems (both
for coordinated and uncoordinated case).
4.Enhanced Data rates for GSM
Evolution
EGPRS Enhanced GPRS
ECSD Enhanced Circuit Switch Data
COMPACT Deployment of services in spectrum below 1
MHz
Enhanced Data Rates for GSM Evolution
(EDGE) refers to the introduction of new modulation techniques based on 8-PSK
(Phase Shift Keying), both for uplink and downlink, in order to evolve data
services in GSM while reusing as much of the physical layer as possible. In
packet, bitrates approximately three times higher than for standard GPRS are
enabled. In circuit, the user data rates are limited to 64 kbit/s, i.e. not
more than without EDGE, but achievable with smaller number of time slots and
relatively simple MS implementation.
Existing services like GPRS and HSCSD are
enhanced by the new physical layer, but services themselves are not modified;
therefore, EDGE is introduced in the existing specifications without creating
new ones.
Two work items introduce EDGE into GSM,
i.e. EDGE Network sub system (NSS) and EDGE Basestation sub system (BSS). The
EDGE BSS work item provides a platform to employ new modulation techniques,
whereas the EDGE NSS work item defines the network changes to facilitate the
physical layer.
EDGE was developed in two phases, and only
Phase 1 is part of Release 1999. EDGE Phase 1 includes Enhanced GPRS (EGPRS)
Phase 1, Enhanced CSD (ECSD) Phase 1, EGPRS COMPACT and Support for EGPRS in
ANSI-136 networks. The new physical layer based on 8-PSK modulation is
introduced and EGPRS and ECSD facilitate the new modulation in single and multi
slot constellation.
No tight link is defined between 8-PSK and
GMSK classes.
EDGE is independent
of the frequency bands, so it can be deployed e.g. in the 900 and the
1900 MHz bands.
4.1.Enhanced GPRS
With the introduction of EGPRS, bitrates
approximately three times higher than for standard GPRS are enabled.
The architectural impacts of EGPRS are
related to the GPRS Packet Control Unit (PCU) and Channel Codec Unit (CCU). The
GPRS specifications allows the placement of the PCU either in the BTS, BSC or
GSN, whereas the CCU is always placed in the BTS. When the PCU is placed remote
to the BTS, information between the CCU and the PCU is transferred in PCU
frames. The CCU may control some of the functions in the remote PCU and vice
versa through inband signalling using the PCU frames.
The functions of the CCU are channel
coding, including FEC and interleaving, and radio channel measurement
functions, including received quality level, received signal level and information
related to timing advance measurements. The PCU is responsible for LLC PDU
segmentation and re-assembly, ARQ (Automatic Repeat reQuest) functions
(including RLC block ACK/NACK), PDCH scheduling, channel access control, and
radio channel management functions.
GPRS and EGPRS employ the same physical
layer, except for the PDTCH. In the case of EGPRS, the modulation format is
inherently signalled by the rotation factor of the training sequences as
specified in GSM 05.04 and GSM 05.02, enabling blind detection in the receiver.
The Radio Block structure for data transfer is different for GPRS and EGPRS,
whereas the same Radio Block structure is used for control messages. For
detailed definition of radio block structure, see GSM 04.60. Other changes
included the introduction of ARQ, Incremental Redundancy mechanism, new MCS
(Modulation and Coding Schemes).
Nine new modulation and coding schemes,
MCS-1 to MCS-9, are defined for the EGPRS packet data traffic channels. For all
EGPRS packet control channels, the corresponding GPRS control channel coding is
used. Coding schemes MCS-1 to MCS-9 are mandatory for MSs supporting EGPRS. A
network supporting EGPRS may only support some of the MCSs. The selection of
MCS is controlled by the network. The coding is based on a punctured
convolutional code with rate 1/3, and for each MCS, there are 2 or 3 puncturing
schemes. For incremental redundancy, an incorrectly received RLC/MAC block is
retransmitted using a different puncturing scheme. In the case of MCS-5-7 and
MCS-6-9, retransmissions with incremental redundancy can even use a different
MCS. Transmission and reception data flows are same for GPRS and EGPRS, except
for EGPRS MCS-7 to MCS-9, where four normal bursts carry two RLC blocks (two
RLC/MAC blocks per radio block, one RLC block within two bursts).
Type II hybrid ARQ is mandatory in EGPRS MS
receivers and the associated performance requirements are specified in GSM
05.05.
EGPRS supports both a pure link adaptation
mode and a combined link adaptation and incremental redundancy mode.
The incremental redundancy mode requires no
extra signalling between the CCU and PCU compared to the link adaptation mode.
The incremental redundancy mode however requires an enhanced CCU functionality.
RLC/MAC protocol is the most affected by EDGE.
EGPRS also impacts the Radio Resource
Management (RR), Mobility Management (MM), and Session Management (SM), leading
to changes on GSM 04.08 and 04.60.
An extended Channel Coding Command (CCC) is
required for EGPRS since there are 8 different MCSs. In addition to this, an
extra 'retransmission resegment' field (1 bit) should be sent along with this
command. This field tells the MS whether it should resegment retransmissions to
an MCS close to the one indicated in the CCC (e.g. used in LA mode or when
memory shortage in BS in IR mode), or stick to the initially used MCS (used in
IR mode when memory is available).
EGPRS also requires a modified/extended
link quality measurement report.
For services where delay is the most
crucial quality parameter but some errors are acceptable, e.g. for some real
time services, the retransmissions associated with the acknowledged mode are
unacceptable, since the delay would be too large. Therefore EGPRS is also able
to operate in unacknowledged mode, i.e. without retransmissions. To achieve an
acceptable BLER or BER without ARQ, the link adaptation switching points in the
proposed scheme are moved to yield the desired robustness in the
non-acknowledged mode.
There are two modulation modes for PDCH in
EGPRS: Linear 8-PSK and GMSK. GMSK is used as fallback when 8-PSK is not
appropriate for the current channel conditions.
Two classes of EDGE-capable mobiles are
defined:
·
Class 1: The two modulation
modes can be used in the downlink, while only GMSK is used in the uplink.
·
Class 2: The two modulation
modes can be used both on uplink and downlink.
Other channels than PDCH always use GMSK
modulation.
The first downlink block in each multiframe
must be modulated with GMSK if a GPRS terminal is allocated on the same timeslot.
Channel coding,
puncturing, interleaving, CRC and burst mapping lead to changes on GSM 05.03.
4.2.Enhanced Circuit Switched Data
ECSD supports both transparent and
non-transparent services, up to 64 kbit/s and 57.6 kbit/s respectively.
The user data rates are limited in Phase 1
ECSD specifications to 64 kbit/s. This means that the maximum data rates are
not increased from the rates supported in current GSM. On the other hand, the
same services are achievable with smaller number of time slots and relatively
simple MS implementation making them more attractive to various data
applications.
ECSD supports interworking with audio
modems and ISDN services on various rates. Group 3 fax services are not
supported with ESCD, because existing GSM channels are adequate for G3 fax
support. If the digital extension of G3 and G4 fax is more widely adopted in
the future, the fax services can be reconsidered.
One of the main applications for ECSD are
the video applications. Besides data rates, there are other requirements for
video transmission: video services usually include several components, each of
them with varying QoS requirements. Figures below show two concepts in terms of
QoS of providing video telephony service. The differences between the concepts
are in the way of ensuring the QoS for different components: audio, video and
data.
For non-transparent (NT) data, BER < 10-3
requirement is assumed. Corresponding value for transparent (T) data is BER
< 10-4…10-5.
The ECSD architecture is largely based on
HSCSD transmission and signalling. This ensures a minimum impact on existing
specifications.
New parameters are required in signalling
messages due to new modulation and channel coding schemes, but the signalling
mechanism is the same as in HSCSD. Also fall back to existing HSCSD and single
slot data services is supported in case the network/MS does not support ECSD.
Link adaptation (LA) between channel coding
schemes in 8-PSK modulation and between GMSK and 8-PSK coding schemes require
new algorithms in BSC. Link adaptation is not a mandatory feature in current
GSM, but particularly in high data rate call using transparent mode connection
LA becomes essential in order to provide the good enough service over large
coverage areas.
Link adaptation algorithms used depend on
the service data rates supported in the network and they are not specified in
GSM specifications.
The required C/I for both GSMK and 8-PSK
modulation with different RX_QUAL values (0-7) shows that the step size between
quality classes is about 2 to 4 dB and dynamic range is about 20 dB for both
modulations. Link simulations show that even C/I of 25 dB is enough in 38.8
kbit/s service for coded BER of 10-5. In practice it is very difficult to
achieve higher C/I values than 30 dB. On the other hand, in the lower end C/I
of 10 dB does not provide good enough service for 29 kbit/s. This would imply
that the dynamical range in current measurements is sufficient for 8-PSK and
link adaptation to GSMK should be performed well before the quality of the
connection falls as low as this.
In practice, the granularity with BER based
reporting and use of three bits for RX_QUAL levels, as in current GSM, is
enough for covering the operative quality range.
It is possible to improve the link
adaptation performance by replacing BER with C/I based reporting and
introducing new measurement metrics like C/I variance used in GPRS. Including
RX_LEV reports from more than six neighbouring cells also improves the
performance of the link adaptation particularly for multi-band mobiles. These
enhancements, however, require more drastic changes in signalling and are seen
as a general improvement to circuit switched measurements rather than EDGE
specific change.
4.3.Support for EGPRS in ANSI-136 networks” and “EDGE Compact” (for
ANSI-136 networks only)
4.3.1.Framework
In January 1998, the ANSI-136 TDMA community,
through the UWCC and TIA TR45.3, evaluated and adopted EGPRS as a key part of
its high speed data evolution. Consequently, a large part of EGPRS was
incorporated as "136 High Speed (136HS)" into the TDMA IMT-2000
proposal called "UWC-136". There were two key characteristics that
136HS allowed: data rates up to 384 kbit/s and initial deployment in less than
1 MHz of spectrum. The TDMA Community has studied how to further enhance 136HS
such that it would be closer to ETSI-EGPRS to better facilitate global roaming,
while also keeping the desire for initial deployment in less than 1 MHz of
spectrum. The result of this effort is this Work Item, which intends to
harmonise the SMG work with the one in Universal Wireless Communication
Consortium (UWCC) to introduce the support of EDGE and GPRS in ANSI-136
networks.
It includes all necessary changes for the
support of the EDGE Compact concept described below (which is an EDGE
implementation in a limited frequency band), and the interaction with the
ANSI-41/136 network. It also includes the changes for the support of roaming
between EDGE Compact and classical EDGE (EDGE Classic) implementations.
The work item includes in total the work on
four different areas:
- Support of GSM on 850 MHz, re-using the work of the work item
“GSM on 400 MHz Frequency Band” described above.
- EDGE Compact, developed in the next section.
- Signalling support for interaction with the ANSI-41/136
network.
- Support for roaming between E-GPRS and GPRS136HS EDGE .
4.3.2.EDGE Compact
COMPACT can be deployed in 600 kHz (plus
guard) of spectrum, and looks as an overlay system to an existing ANSI-136
network. As such, COMPACT is independent of the ANSI-136 system, which
facilitates roaming of EGPRS only mobile stations. It also allows operators to
deploy different infrastructure vendors for their data solution from their
voice network.
For operators having more available
bandwidth, the TDMA Community also supports the development of ETSI-EGPRS
(referred to as "EGPRS Classic"), requiring 2.4 MHz of initial
spectrum. The support of both COMPACT and EGPRS Classic under what is called
EGPRS-136 eases the convergence of GSM and ANSI-136 systems.
Integration of GPRS with ANSI-136 is
logically accomplished by the addition of the GPRS network nodes SGSN and GGSN
to the ANSI-41 circuit-switched network.
EGPRS (i.e. EGPRS Classic) terminals on the
850 or 1900 MHz band will also support COMPACT to facilitate roaming.
COMPACT terminals would also support EGPRS
Classic to facilitate roaming.
The idea is that a mobile supporting TDMA
band (850 and 1900) shall support both Compact and Classic EDGE (in USA
EDGE Classic uses the same as EDGE with
4x12 except BCCH on TN0.
EDGE Compact uses BS sync and modified
control channel. In EDGE Compact, the PSCH is different due to 52 multiframe
structure, and the cell reselection is modified and a new measurement method is
used. New broadcast information info on 200 kHz for 30 kHz voice page is also
used.
Estimations mention to save 20 to 25% load
with Edge compact compared to Edge classic because of DTX.
Note that EDGE Compact was developed to
meet the UWCC requirements but can be implemented in other frequency bands with
limited spectrum, e.g. GSM400.
Continue to read, 3G Charging and Testing , Please Click Here
0 comments:
Post a Comment
If there is any comments,Please leave a comment at here.