Nokia Siemens Networks

Doubling GSM voice capacity

with the Orthogonal Sub Channel

Technology brief
 A powerful software Contents
upgrade leverages 03 Executive summary
quaternary modulation 04 1. Introduction
05 2. Key cost-efficiency benefits
and MIMO techniques
for network operators
to improve network 05 2.1 Improved BTS hardware utilization
efficiency with existing 05 2.2 Efficient spectrum utilization
handsets. 06 2.3 Minimized site density
06 2.4 Improved energy efficiency
06 2.5 Fast deployment with existing handsets
07 3. Technical concepts
07 3.1 Training sequences differentiate channels
07 3.2 Quaternary modulation in the downlink
08 3.3 Optimized pulse shaping filter
maximizes DL coverage
08 3.4 Multi-user MIMO in the uplink
08 3.5 Matchmaking and resource
optimization by DFCA
09 3.6 Further gains with upcoming 3GPP
VAMOS handsets
10 4. Conclusion
11 Glossary
11 References

2 Doubling GSM voice capacity with the Orthogonal Sub Channel

Executive summary
Driven by new subscriptions, extended
coverage, and lower tariffs, conventional
voice traffic is expected to grow nearly
threefold by 2012 [1]. However, revenues
per delivered minute are declining,
compelling mobile operators to cut
costs by improving GSM networks
efficiency.

Yet another innovative feature from

Nokia Siemens Networks, the
Orthogonal Sub Channel (OSC)
enables operators to double voice
channel capacity in GSM radio networks
by upgrading software. This upgrade
bases on quaternary modulation and
multi-user MIMO techniques in the
downlink and uplink, respectively.
To ensure immediate gains, both
techniques can be applied with typical
GSM handsets that support Single
Antenna Interference Cancellation
(SAIC). 3GPP aims to standardize
these techniques in a work item called
VAMOS, providing further gains in
the near future.

Expecting OSC to reduce operators

CAPEX and OPEX for voice services
by up to 50%, Nokia Siemens Networks
actively contributes to standardization
efforts and aims to release the OSC
in 2010.

Doubling GSM voice capacity with the Orthogonal Sub Channel 3

 1. Introduction
This paper presents a pioneering
feature from Nokia Siemens Networks A
A
called Orthogonal Sub Channel. QPSK Encoding
The OSC aims to improve cost efficiency Transmitter
B
Encoding
for GSM voice services and support
the vision of connecting five billion A
MIMO Decoding
people by 2015. Efforts are underway B
Receiver
to standardize properties associated B
Decoding
with this feature in 3GPP Release 9 and
in a related work item called VAMOS.

The technical method of doubling voice Figure 1: Overview of the OSC with two simultaneous calls in the same radio channel
channel capacity entails adopting the
quaternary modulation scheme in the
As far as radio link performance goes,
downlink and spatially orthogonal sub
receiving an orthogonal sub channel
channels in the uplink, also called a
requires about the same signal energy
multi-user MIMO. With these two key
per user as for a conventional GMSK
techniques linked to adaptive multi-rate
channel. System simulations have
(AMR) half-rate, four legacy SAIC
shown significant capacity gain when
handsets can be served at the same
the available spectrum is exploited
time within a radio time slot.
using fewer TRXs. Each TRXs capacity
increases to maintain maximum
In addition to new 3GPP VAMOS
coverage and reduce per-user energy
handsets, these radio techniques are
consumption. This means the OSC
also applicable to existing GSM SAIC
is able to minimize CAPEX and
handsets. Consequently, upgrading
OPEX, paving a profitable path to
the software in GSM radio networks
growth for operators. Specifically,
brings immediate benefits to network
OSC helps operators:
operators.
Improve BTS hardware utilization
Exploit available spectrum efficiently
Maintain coverage in capacity
extensions
Improve energy efficiency
Enjoy immediate gains with
existing handsets

4 Doubling GSM voice capacity with the Orthogonal Sub Channel

2. Key cost-efficiency benefits
for network operators
2.1 Improved BTS
hardware utilization Erlangs per TRX (2% blocking)
32
FR
The OSC features greatest asset is its 28 HR
ability to double radio capacity, enabling OSC
24
operators to efficiently exploit and even
reduce installed hardware. For example, 20
system simulations have shown that

Erlangs
16
served traffic could exceed 20 Erlangs
per carrier. And with the trunking effect, 12
the OSC could bring even greater
8
gains to smaller BTS configurations.
4
As the OSC increases voice capacity,
0
it also frees up capacity for data traffic. 1 2 3 4
For instance, one carrier could be #TRX

allocated to a downlink dual carrier

without compromising voice quality or Figure 2: Channel capacity
capacity in a typical four TRX per cell
configuration with OSC.

Introducing the OSC may also free

up antennas when another radio
technology or operator needs to
share the same site.

2.2 Efficient spectrum

Spectral Efficiency (TCH, 3 sectors, 3.6MHz)
utilization
40 OSC
AMR
The OSC also increases spectral
30 EFR efficiency measured in Erl/MHz/cell,
Erl / MHz / Site

especially when the number of carriers

20
is restricted in the available spectrum.
As figure 3 shows, the OSC can fully
10
exploit available spectrum with less
TRX hardware than is necessary for
0
2 3 4 AMR. The mean C/I value is generally
#TRXs per sector greater than 12 dB, so the OSC may
be used typically for over 50% of calls.
Figure 3: The OSC exploits the spectrum with high hardware efficiency
Dynamic Frequency and Channel
Allocation (DFCA) excels at pairing
and allocating two users in the optimal
channel, and can further improve
spectral and hardware efficiency.

Doubling GSM voice capacity with the Orthogonal Sub Channel 5

 2.3 Minimized site density
Site density

In the past, mobile operators seeking

to extend capacity to satisfy growing
traffic demand generally augmented 100%
BTS hardware at cell sites with

Site / Erl / km
antenna combiners to add more
transceivers. Wideband transmitter
combiners, for example, introduce 50%
about 3.5dB loss. Compensating
for the decrease in coverage may
necessitate new sites. The OSC,
in turn, maintains the coverage area 0
Legacy BTS Flexi EDGE Flexi EDGE
in capacity extensions. This can BTS BTS with OSC
cut the site count by up to 50%,
and the operators costs accordingly.
Figure 4: BTS site density for providing coverage & capacity
The OSC offers an efficient method of
adding channel capacity while leaving
energy consumption. Whats more,
the site hardware configuration
a country-wide network based on
untouched.
fewer sites requires less power.

2.4 Improved energy Conserving energy is crucial to cutting

efficiency
Less hardware means greater energy
efficiency. Simulations with a typical
traffic pattern show that the OSC reduces
per-Erlang energy consumption by 20
to 50%. By extension, fewer or smaller
rectifiers, cooling units, and battery
backups will further reduce each sites
global carbon emissions. Aiming
to do our part for the global cause,
Nokia Siemens Networks offers a
growing range of green products
designed to maximize network
infrastructures energy efficiency.
2.5 Fast deployment with
existing handsets

Deploying the OSC successfully

requires mobile handsets that support
AMR and SAIC. Estimates put the total
number of such devices on the market
today at about one billion, so operators
can benefit immediately by introducing
the OSC.

6 Doubling GSM voice capacity with the Orthogonal Sub Channel

 3. Technical concepts
3.1 Training sequences
differentiate channels
Various training sequences used
in a legacy burst structure can serve
to separate two simultaneous users in
the downlink and uplink. These training
sequences must exhibit low cross-
correlation and good auto-correlation
in the presence of another sub channel.
The table to the right shows training
sequences that provide this combination
today, enabling multiplexing of widely
used SAIC handsets in both sub
channels.
3GPP seeks to standardize an optimized
set of training sequences to achieve
best performance and compatibility
with upcoming VAMOS handsets.
Q (B)
(0,1) (1,1)
I (A)
(0,0) (1,0)
Figure 5: Bit mapping for two users (A, B) with QPSK,
a subset of 8PSK
TRS-A/TRS-B 0 1 2 3 4 5 6 7
0 - Yes Yes Yes
1 - Yes Yes Yes
2 Yes Yes - Yes
3 Yes Yes - Yes
4 Yes - Yes
5 Yes - Yes
6 Yes Yes -
7 Yes Yes -
Table 1: Suitable training sequence pairs for use with existing handsets
3.2 Quaternary modulation
in the downlink
The BTS multiplexes two users in
the downlink, employing quaternary
modulation as a combination of two
orthogonal and binary sub channels
with different training sequences. The
sub channels are mutually orthogonal,
and use /2 symbol rotation to imitate
GMSK, so legacy GMSK SAIC handsets
can receive them separately.
The figure below shows how users
A and B are mapped in a quaternary
constellation known as QPSK or
4-QAM, which is also a subset of an
8-PSK constellation. User A is mapped
to the first bit and user B to the second
bit, with the resulting signals appearing
along the I or Q axis, respectively.
QPSK is switched to GMSK when
only one sub channel is active to also
facilitate DTX gains. The transmitter
switches off when both sub channels
are inactive. Switching from QPSK to
GMSK is possible to make signaling
more robust. This switching between
modulations is transparent to the
legacy SAIC handset because it
expects to receive GMSK all the time.
Doubling GSM voice capacity with the Orthogonal Sub Channel 7
 3.3 Optimized pulse shaping
filter maximizes DL coverage
In the downlink, with the benefit of an
added, optimized pulse shaping filter,
QPSK requires about the same amount
of energy per user as the related GMSK
channel. As the graph below shows,
the modulation shaping filter currently
used in EGPRS requires about 2dB
more power.
This indicates the measured performance
with a legacy handset for 5.9k half-rate
AMR codec and the OSC with current
and optimized pulse shaping filters.
3.4 Multi-user MIMO
in the uplink
In the uplink, two handsets use
conventional GMSK transmission
differentiated by the training sequence
and spatially orthogonal sub channels
provided by independent multipath
propagations. Base station hardware
requires antenna diversity to receive
two users who are transmitting at the
same time. This rule of one antenna
per user applies to multi-user MIMO
systems in general. Indeed, the power
received in the BTS should be aligned
by means of power control and proper
user pairing within a 15dB window.
3GPPs LTE standard has adopted
similar multiple-input-multiple-output
uplink technology, first demonstrated
by Nokia Siemens Networks in 2007
[2].
8 Doubling GSM voice capacity with the Orthogonal Sub Channel
Measured sensitivity (AHS5.9, TU50, 1800MHz)
10
HR AMR
OSC
OSC
(current pulse)
3GPP limit
(AHS5.9)
FER [%]
1
0.1
-106 -104 -102 -100 -98 -96
RX Level [dBm]
Figure 6: Optimized pulse shaping filter maximizes the OSCs coverage
3.5 Matchmaking and resources, for example, frequency
resource optimization hopping parameters and training
sequences. DFCA excels when there
by DFCA
are more resources to choose from
and when users or in this case,
The pairing of two users in the same
couples have specific C/I demands.
radio channel is one of the most critical
of all the aspects radio resource
So DFCA excels at optimizing
management is tasked optimize.
the tasks of:
The BSC can take advantage of
the software-based pairing capability
Matching pairs for the OSC
of DFCA to search for the best pair
Searching for the best training
among several candidates. Although
sequence for the OSC
the BSC uses part of the DFCA
Searching for the best radio
algorithm for pairing, it is not necessary
resources; that is, frequencies,
to synchronize BSS, as is usually the
time slot, and hopping parameters
case when introducing DFCA.
Optimizing system capacity for voice
users with different C/I demands
The great advantage of a synchronized
network is that the DFCA feature can
also search for the best radio
 Q(B) Q(B)
(0,1) (1,1)

(0,1) (1,1)

I(A) I(A)

(0,0) (1,0)

(0,0) (1,0)

Figure 7: Quaternary constellations with unequal power,

left-hand diagram: user A has larger power than user B, right-hand diagram: vice versa

3.6 Further gains with This improved receiver performance The table below lists properties subject
upcoming 3GPP VAMOS also enables sub channel-specific to standardization, and indicates which
power control, where the square may be applied to the two types of
handsets
quaternary modulation constellation in handsets.
the downlink changes to a rectangular
3GPP aims to standardize this new
constellation, boosting power by Advanced receivers may also be used
channel type in Release 9 as part of
almost 2dB for another sub channel. with the OSC in handsets, for example,
a work item called VAMOS, to include
Constellations providing unequal receiver diversity that could double
support for legacy and new VAMOS
power to sub channels can be derived spectral efficiency.
handsets.
from the existing 8PSK constellation.
This enables precise blind detection
New VAMOS handsets will be aware
of the given modulation constellation
of quaternary modulation and training
in VAMOS handsets on a burst basis
sequences in both sub channels,
to ensure optimum performance.
providing about 1dB link gain
over legacy SAIC handsets.
Handset type: Legacy SAIC VAMOS

New set of training sequences No Yes

Optimized pulse shaping filter in DL Yes Yes

QPSK modulation in DL Yes Yes

Lower power sub channel in DL No Yes

Higher power sub channel in DL Yes Yes

Table 2: VAMOS Rel9 properties (subject to standardization)

and their suitability for legacy SAIC handsets

Doubling GSM voice capacity with the Orthogonal Sub Channel 9

 4. Conclusion
The OSC efficiently exploits radio
spectrum and hardware to rise to
operators biggest challenges cost,
capacity, and coverage. With fewer
sites and more efficient hardware,
the network consumes less power to
deliver the same traffic, and ensures
the operator leaves a much smaller
environmental footprint.
The OSC factors affordability into the
business equation, enabling operators
to cater to new and existing subscribers
with more competitive prices.
10 Doubling GSM voice capacity with the Orthogonal Sub Channel
Optimized pulse shaping in the BTS
transmitter and matchmaking in the
BSC are key capabilities for maximizing
gains brought about by the OSC feature.
Doubling voice capacity and enhancing
overall efficiency requires merely a
software upgrade in the radio network.
The technique works with existing
GSM SAIC handsets, providing
immediate benefits for operators
who have chosen Flexi EDGE BTS.
Glossary
3GPP 3rd Generation Partnership Project
AHS5.9 AMR Half Rate, 5.9 kbps
AMR Adaptive Multi-Rate speech codec
BSC Base Station Controller
BTS Base Station
C/I Carrier to Interference ratio
CAPEX Capital expenditure
DFCA Dynamic Frequency and Channel Allocation
DL Downlink
DTX Discontinuous Transmission
EDGE Enhanced Data Rates for GSM Evolution
EFR Enhanced Full Rate
EGPRS Enhanced general packet radio service
FR Full Rate
GMSK Gaussian minimum shift keying
GSM Global System for Mobile Communications
EGPRS Enhanced General Packet Radio Service
HR Half Rate
LTE Long Term Evolution
MIMO Multiple Input Multiple Output
OPEX Operating expenditure
OSC Orthogonal Sub Channel
QAM Quadrature Amplitude Modulation
QPSK Quaternary Phase Shift Keying
PSK Phase Shift Keying
RF Radio frequency
SAIC Single Antenna Interference Cancellation
TCH Traffic Channel
TRS Training Sequence
TRX Transceiver
TU50 Typical Urban fading profile, 50 km/h
UL Uplink
VAMOS Voice services over Adaptive Multi-user Orthogonal Sub channels

References:
[1] Source: Informa 2008
[2] Nokia Siemens Networks Press release of May 2, 2007:
Researchers at Nokia Siemens Networks double the
capacity in the uplink using Virtual MIMO in LTE networks

Doubling GSM voice capacity with the Orthogonal Sub Channel 11

Nokia Siemens Networks
P.O. Box 1
FI-02022 NOKIA SIEMENS NETWORKS
Finland
Visiting address:
Karaportti 3, ESPOO, Finland

Switchboard +358 71 400 4000 (Finland)

Switchboard +49 89 5159 01 (Germany)
Order No. C401-00357-WP-200902-1-EN
Copyright 2009 Nokia Siemens Networks.
All rights reserved.

Nokia is a registered trademark of Nokia Corporation,

Siemens is a registered trademark of Siemens AG.
The wave logo is a trademark of Nokia Siemens Networks Oy.
Other company and product names mentioned in this document
may be trademarks of their respective owners, and they are
mentioned for identification purposes only.

This publication is issued to provide information only and is not

to form part of any order contract. The products and services
described herein are subject to availability and change without
notice.

www.nokiasiemensnetworks.com