Professional Documents
Culture Documents
EDGE: Enhanced Data Rates For GSM and TDMA/136 Evolution: IEEE Personal Communications July 1999
EDGE: Enhanced Data Rates For GSM and TDMA/136 Evolution: IEEE Personal Communications July 1999
EDGE: Enhanced Data Rates For GSM and TDMA/136 Evolution: IEEE Personal Communications July 1999
net/publication/3344055
CITATIONS READS
322 3,401
4 authors, including:
Anders Furuskar
Ericsson
56 PUBLICATIONS 3,636 CITATIONS
SEE PROFILE
All content following this page was uploaded by Anders Furuskar on 18 September 2013.
The use of high-level modulation to provide enhanced data rates for GSM Each of these functions addresses the need
evolution is currently being standardized for GSM by the European for increased data capabilities. HSCSD in-
Telecommunications Standards Institute. Likewise, Edge is being adopted troduces multislot operation, and the com-
for TDMA/136 evolution by the Universal Wireless Communications Con- bination of HSCSD and GPRS adds a new
sortium. mechanism through the air interface, where-
by users can remain connected to the net-
The authors describe the architecture and equipment-related modifica-
work but only use radio capacity when ac-
tions that are needed for introducing Edge into a GSM or TDMA system, tually transmitting or receiving data. In ad-
and explain how the Edge concept affects radio-network planning. They dition to these functions, new core network
also describe an evaluation of system performance. parts—which will be positioned in parallel
to the mobile switching center (MSC)—will
provide direct access to the Internet/in-
tranet.
HSCSD and GPRS achieve high bit rates
through multislot operation. But because
Background these techniques are based on original
GSM and TDMA/136 are second-generation Gaussian minimum-shift keying (GMSK)
cellular standards with worldwide success. modulation, they yield only a moderate in-
GSM is currently used by more than 135 crease in bit rates per time slot.
million subscribers in over 100 countries. For TDMA/136 evolution, similar stan-
The TDMA/136 system family (which in- dardization activities have been started. In
cludes EIA-553 and IS-54) serves approxi- 136+, for example, the combination of
mately 95 million subscribers in over 95 multislot operation and the new modulation
countries. scheme, 8-PSK (based on the 30 kHz carri-
While speech is still the main service of er bandwidth), will increase data rates by ap-
these mobile systems, support for data com- proximately four times.3
munications over the air interface is rapid- The main driver for third-generation
ly improving. Standard products currently wireless communication and IMT-2000 is
provide data services with bit rates up to the ability to supplement standardized ser-
9.6 kbit/s, but forthcoming steps in the evo- vices currently available in GSM and
lution of GSM, known as Phases 2 and 2+, TDMA/136 with wideband services. In
are already in the final stages of standard- brief, third-generation systems will provide
ization, which among other things, defines wide-area coverage at 384 kbit/s and local-
• high bit-rate, circuit-switched modes area coverage at approximately 2 Mbit/s.4, 5
(high-speed circuit-switched data, The new 2 GHz frequency band for wide-
HSCSD1); band code-division multiple access
• packet services (general packet radio ser- (WCDMA) is being supported and stan-
vice, GPRS2). dardized by both the European Telecom-
BOX A, ABBREVIATIONS
Radio-network planning
An important prerequisite (and to a large
Figure 3
extent, one that will determine the success GSM network architecture.
of Edge) is that network operators should be
able to introduce Edge gradually. The ini-
tial deployment of Edge-capable trans-
ceivers will supplement standard GSM GMSC PSTN
transceivers in a subset of cells where Edge
coverage is desired. An integrated mixture A MSC /
of circuit-switched, GPRS and Edge users VLR
will thus coexist in the same frequency band. Abis
To minimize operator efforts and costs, BSC HLR
Edge-related implementation must not re- GPRS
register
quire extensive modification of the radio-
network plan (including cell planning, fre- SGSN
Gb
quency planning, the setting of power lev-
els and other cell parameters).
Coverage planning
Backbone
One characteristic of non-transparent radio- External
IP network GGSN IP network
link protocols (for example, protocols that
include automatic repeat request, ARQ) is that
poor radio-link quality only results in a
lower bit rate. Unlike speech traffic, a low Affected by Edge introduction
carrier-to-noise ratio does not cause data ses-
60
Link adaptation ment must be automated, to avoid the split-
ting of channels into static groups. Other-
50
wise, trunking efficiency would diminish.
40
Link adaptation
30 The dynamic selection of modulation and
coding scheme to suit radio link quality is
20 referred to as link adaptation. The Edge stan-
dard supports a dynamic-selection algo-
10
rithm that includes
• the measurement of, and reports on,
0
0 5 10 15 20 25 30 35 40 45 50 downlink quality;
C/I (or C/N) (dB) • a means of ordering new modulation and
coding for the uplink.
Figure 4 Link adaptation is meant to be fully auto-
Principle of link adaptation: based on measurements of link quality, it is possible to select mated; that is, it will not require the net-
the modulation and coding scheme that yield the highest data rate. work operator to do any addition planning
(Figure 4). Possibilities of enhancing ARQ
performance through incremental redun-
dancy (Hybrid II/III ARQ18) are also under
investigation; for example, as proposed for
transmission (DTX), a reuse factor as low as TDMA/136. A scheme of this kind could
three becomes feasible.17 A frequency reuse reduce the need to use link adaptation when
factor of three implies that each frequency selecting modulation. Similarly, the ARQ
is reused every third base station. scheme might be enhanced to handle the
Edge supports this trend. In fact, thanks “selection” of channel coding. Link adapta-
to link-adaptation techniques, Edge can be tion, incremental redundancy and combi-
introduced into an arbitrary frequency plan, nations of the two are commonly referred to
thereby benefiting from a high carrier-to- as link quality control.
interference ratio (C/I) near base stations. In
summary, we see that Edge can be intro- Power control
duced into an existing GSM frequency plan, Current GSM systems use dynamic power
and that doing so provides support for fu- control to increase equity in the system and
ture, high-capacity solutions. to extend the life of batteries in mobile ter-
minals. Similar strategies will be employed
Channel management for GPRS, although the actual signaling
After Edge has been introduced, a cell will procedure will be different.2 Edge support
typically include two types of transceiver: for power control is anticipated to be more
standard GSM transceivers and Edge trans- or less identical to that of GSM/GPRS.
ceivers. Each physical channel (time slot) in Thus, network operators will still only be
the cell can be viewed as being one of at least able to affect parameter settings. But be-
four channel types: cause Edge users can benefit from a much
1. GSM speech and GSM circuit-switched higher carrier-to-interference ratio than
data (CSD); standard GSM users, Edge power-control
2. GPRS packet data; parameters will be different.
3. circuit-switched data, enhanced circuit-
switched data (ECSD), and GSM speech;
4. Edge packet data (EGPRS), which allows
Edge in TDMA/136
a mix of GPRS and EGPRS users simul- systems
taneously.
While standard GSM transceivers only sup- 136HS requirements
port channel types 1 and 2, Edge trans- Some of the requirements imposed on
ceivers support all four channel types. Phys- 136HS include considerations that exceed
ical channels are dynamically defined ac- the ITU requirements for IMT-2000 but
cording to terminal capabilities and needs are crucial to TDMA/136 operators. Such
in the cell. For example, if several speech considerations include flexible spectrum
users are active, the number of type-1 chan- allocation, high spectral efficiency, com-
Coverage planning
To be introduced into current cell plans and
use existing base stations, 136HS must pro-
vide the same or better coverage as we conclude that EGPRS and 136HS pro-
TDMA/136 and 136+. EGPRS with link vide good packet-data performance, there-
quality control satisfies this requirement, by allowing for an initial deployment in less
providing coverage that is at least as good than 1 MHz of spectrum.
as TDMA/136. Thanks to link quality con-
trol, poor radio-link quality does not cause Control channel aspects
cells to be dropped, but only reduces the user Given a 1/3 frequency reuse pattern and
bit rate. fractional loading in 136HS, Edge carriers
cannot transmit continuously at constant
Frequency planning power. Therefore, the cell that is most suit-
To meet the requirements for deploying able for service cannot be determined
136HS within 1 MHz of spectrum (initial using the standard GSM/GPRS procedure
deployment), it has been proposed that of measuring signal strength on the GSM
EGPRS be deployed using a 1/3 frequency carrier that transmits the broadcast con-
reuse pattern. Thanks to link quality con- trol channel. Instead, to select and reselect
trol, EGPRS can be introduced in a tight cells, 136HS will measure signal strength
frequency plan and still provide high data on the TDMA/136 digital control channel
rates for packet data services. (DCCH). However, all packet channels—
For example, by means of fractional load- that is, traffic and control channels—can
ing, a 1/3 frequency reuse pattern can offer be transmitted on the Edge carrier. This
a C/I that is sufficient for an average system arrangement permits tight frequency
data rate of 384 kbit/s. Based on studies of reuse planning and facilitates a natural in-
EGPRS performance from a 1/3 frequency tegration into existing TDMA/136 sys-
reuse pattern8, 9 and on simulation results, tems.
C.D.F. [%]
to 30%. comes intolerable. 60
dropping user sessions must study not only spectral efficiency per 30
No admission control algorithm is used in se but also the spectral efficiency that can 20
the simulations. Instead, every user that gen- be achieved for maximum levels of delay. 10
based, first-in, first-out (FIFO) fashion. User define. For example, because they take
sessions with poor link quality are dropped longer to transmit, long packets are more Figure 6
according to a leaky-bucket algorithm: each likely to exceed an absolute limit than short Distribution of normalized delay for differ-
user’s counter is initially set at its maximum packets—even without queuing and re- ent numbers of users per sector,
Edge/GSM case.
value of 32. A negatively acknowledged transmission. Even so, some recipients are
block (NACK) decreases the counter by one, certain to be satisfied at having received
whereas each acknowledged block (ACK) large amounts of data, despite higher ab-
increases it by two. If the counter reaches solute delays. This line of reasoning leads
zero, the session is dropped. The drop rate to the introduction of a measure of normal-
is less than 1%, even at maximum load. ized delay. Further, assuming that the ac-
More sophisticated algorithms are expected ceptable delay can be doubled for packets
to improve system performance. that are twice the size of ordinary packets,
tripled for packets thrice the size, and so on,
Coverage simulations in noise-limited then we can project the normalization in a
systems linear fashion by dividing absolute delay by
Where coverage is limited, performance is the size of the packet. We thus define the
independent of interference or traffic dy- measurement of normalized delay as being
namics. Therefore a static simulation tech- the total absolute delay (queuing time plus
nique can be used. Snapshots are taken of transmission time) divided by packet size in
the system, in which stationary mobile ter- kbits. Note: The normalized packet delay is
minals are placed randomly according to a the inverse of the bit rate measured per
uniform distribution. GSM and packet. Accordingly, the maximum ac-
TDMA/136 systems with 95% speech cov- ceptable normalized delay for a packet cor-
erage are used as a reference, to determine responds to the minimum acceptable bit
what coverage can be achieved from Edge in rate for that packet. In summary, this line
existing cell plans. Edge performance is then of reasoning leads to a performance measure
analyzed assuming the same carrier output that shows the spectral efficiency that can be
power as was recorded in these reference sys- achieved for a given maximum normalized delay
tems. For GSM, assuming a full-rate speech requirement.
coder, speech coverage requires a signal-to- This measurement, however, does not in-
noise ratio (Eb/N0) of 6 dB; for TDMA/136, dicate how fair the system is to its users. To Figure 7
Spectral efficiency versus normalized
the requirement is 15.7 dB. These are the determine fairness, we measure the average delay, Edge/GSM and standard GSM. The
values found at the 5th percentile of the packet bit rate per user by averaging the bit average number of users per sector is
Eb/N0 distributions in the cell. When 8- rate of each of its packets, where packet bit given for each simulated value.
PSK is introduced (the Edge modulation rate is defined as packet size divided by time for
0.35
scheme), the Eb/N0 distributions are lower, queuing and transmission.
due to the higher gross bit rate. Therefore, 0.3
Normalized delay at 90th percentile [s/kbit]
Standard GSM
assuming the same carrier output power, we Simulation results 0.25
can calculate the difference in Eb/N0 for
Edge, compared to standard GSM and Capacity, GSM scenario 0.2
Edge
TDMA/136 modulations (Box B, Equations Beginning with GSM, Figure 6 shows the 0.15
1 and 2). distribution of normalized delay among
transmitted packets at different loads. Load 0.1
Unlike circuit-switched systems, packet- users per sector. As expected, delay increas-
0
data systems have no fixed capacity limita- es as load increases. Different loads also re- 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35
Spectral efficiency [bit/s/Hz/site]
tions; that is, packets that cannot be trans- sult in different levels of spectral efficiency.
30
Coverage simulations
20 The coverage simulations gave Eb/N0 dis-
10 tributions for both GSM and TDMA/136
0 scenarios. From the original Eb/N0 distrib-
0 10(80) 20(160) 30(240) 40(320) 50(400) 60(480) 70(560)
Average packet bit rate per user per time slot
utions we can calculate the 8-PSK distrib-
(bit rate per 8 time slots in brackets) [kbit/s] utions (Box B, Equations 1 and 2). From
Conclusion REFERENCES
Operators who introduce Edge into GSM
and TDMA/136 systems can efficiently
reuse existing infrastructure: 1 ETSI, GSM 03.34, “High Speed Circuit GSM and IS-136 Evolution,” in Proceedings
Switched Data (HSCSD),” version 5.0.1, of NRS’98.
• Edge increases GSM data rates to 384 1997-04. 11 H. Olofsson, A. Furuskär, “Aspects of intro-
kbits/s. 2 ETSI, GSM 03.64, “Overall Description of the ducing EDGE in existing GSM Systems,” in
• The introduction need not affect operator GPRS Radio Interface,” version 5.1.0, 1997-11. Proceedings of IEEE ICUPC´98.
coverage and frequency plans. 3 S. Labonte “A proposal for the evolution of 12 ETSI, Tdoc SMG2 WPB 108/98, “ EDGE -
IS-136,” in Proc. IEEE VTC´98. Evaluation of 8-PSK,” 20-24 Apr., 1998.
• Because Edge-capable physical channels 4 ETSI, UMTS 21.01, “Requirements for the 13 J. Dunlop, J. Irvine and P. Cosimini, “Esti-
may be used for standard GSM services, UMTS Terrestrial Radio Access System mation of the Performance of Link Adapta-
fixed channel allocation between services (UTRA),” version 3.0.1, 1997-10. tion in Mobile Radio,” in proceedings of IEEE
is not needed. Thus, from an operator’s 5 Recommendation ITU-R M.1225, “Guide- VTC-95.
point of view, Edge services may be in- lines for Evaluation of Radio Transmission 14 ETSI, GSM 05.02, “Multiplexing and multi-
Technologies for IMT-2000.” ple access on the radio path,” version 5.3.0,
troduced in a smooth fashion. 6 ETSI. Tdoc SMG2 95/97. “EDGE Feasibility March 1997.
• The results of simulated packet-data per- Study, Work Item 184; Improved Data Rates 15 H. Nakamura, H. Tsuboya, M. Nakano and A
formance indicate that Edge enables sig- through Optimized Modulation,” version 1.0, Nakajima, “Applying ATM to Mobile Infra-
nificantly higher peak rates and approxi- December 1997. structure Networks,” in IEEE Communica-
7 P. Schramm et al, “Radio Interface Perfor- tions Magazine, Vol. 36, No.1, Jan 98.
mately triples the spectral efficiency asso- mance of EDGE, a Proposal for Enhanced 16 ETSI, Tdoc EDGE 110/98, “Comparison of
ciated with standard GSM and Data Rates in Existing Digital Cellular Sys- the proposed modulation schemes,” Apr.
TDMA/136. tems,” in Proc. IEEE VTC’98. 20-24, 1998.
• Edge increases packet bit-rate coverage, 8 A. Furuskär, M. Frodigh, H. Olofsson and J. 17 S. Engström et al, “Multiple Reuse Patterns
which means that operators of GSM or Sköld, “System Performance of EDGE, a for Frequency Planning in GSM Networks,”
Proposal for Enhanced Data rates in Exist- in Proc. IEEE VTC’98.
TDMA/136 systems can reuse existing ing Digital Cellular Systems,” in Proc. IEEE 18 R. v. Nobelen, “Towards Higher Data Rates
sites. Edge achieves good spectral effi- VTC’98. in IS-136,” in Proc. IEEE VTC’98.
ciency using only a limited amount of 9 K. C. Zangi, A. Furuskär and M. Höök, 19 K. Blomquist, J.-A. Kjellberg, “A study of
spectrum. “EDGE: Enhanced Data for Global Evolution Self-Similar Data Traffic and Development of
of GSM and IS-136,” Proc. MDMC’98. a WWW Traffic Model,” MSc Thesis,
Edge serves as an enabler of forthcoming 10 A. Furuskär et al, “EDGE: Enhanced Data for Linköping University, Sweden, June 1997.
wideband services in GSM and TDMA/136.