Enhanced Data rates for GSM Evolution - EDGE

EDGE, a new radio interface technology with enhanced modulation, increases the HSCSD and GPRS data rates by up to three fold. EDGE modulation will increase the data throughput provided by the packet switched service even over 400 kbit/s per carrier. Similarly, the data rates of circuit switched data can be increased, or existing data rates can be achieved using fewer timeslots, saving capacity. Accordingly, these higher speed data services are referred to as EGPRS (Enhanced GPRS) and ECSD (Enhanced Circuit Switched Data).

EDGE, expected to be deployed in 2000–2001, is a major improvement in GSM phase 2+. As a modification to existing GSM networks, EDGE does not require new network elements.

EDGE is especially attractive to GSM 900, GSM 1800 and GSM 1900 operators that do not have a licence for UMTS, but still wish to offer competitive personal multimedia applications utilising the existing band allocation. Also, EDGE can co-exist with UMTS, for instance to provide high speed services for wide-area coverage while UMTS is deployed in urban hot spots.

In the US, EDGE is part of the IS-136 High Speed concept which is one of the third generation RTT (Radio Transmission Technology) proposals from TR45. EDGE will be also standardised in US which makes possible to achieve a global mobile radio system with many services characteristic to third generation systems.

Nokia is dedicated to supporting GSM operators with wireless data solutions that help them create value in the market place, both now and in the future. Wireless data is steady evolution, not revolution. With Nokia’s experience, the operator starting today with wireless data can accumulate the skills and know-how to build a strong market position, all the way to third generation systems and the personal multimedia era. This White Paper describes Nokia’s understanding of the role and benefits of EDGE as wireless data evolves towards personal multimedia.

EDGE

The GSM standard is being developed to support mobile services with radio interface data rates even over 400 kbit/s. This work is being performed under the ETSI work item EDGE (Enhanced Data Rates for GSM Evolution).

The major change in the GSM standard to support higher data rates is the new modulation system, known as 8PSK (Phase Shift Keying). This will not replace but rather co-exist with the existing GMSK (Gaussian Minimum Shift Keying) modulation. With 8PSK, it is possible to provide higher data rates with a somewhat reduced coverage, whereas GMSK will be used as a robust mode for a wide area coverage.
EDGE brings more speed and capacity when needed

In mature GSM markets, cellular data penetration is forecast to increase exponentially during the early 2000’s. New wireless data applications and innovative
terminal types will generate completely new markets: aggressive GSM operators can expect to obtain up to 30 % of their airtime and revenue from wireless data by year 2000.

HSCSD (High Speed Circuit Switched Data) and GPRS (General Packet Radio Service), introduced to GSM in 1998 and 1999 respectively, will enable cellular operators to offer higher than 9.6 kbit/s data rates to their subscribers for new data applications.

Cellular operators that have invested in HSCSD and GPRS expect to be able to offer higher data rates without building too many new sites. The ECSD (Enhanced Circuit Switched Data) and EGPRS (Enhanced General Packet Radio System) solutions offer data services comparable to 3rd generation levels with considerably fewer radio resources than in standard GSM. This means that EDGE TRXs (transceivers) carry more data per time slot, decreasing the need for new TRXs/frequencies. In addition, end user response times decrease, ensuring good service levels as data usage increases.

It could be possible for EDGE Phase 2 to provide a voice service using AMR (Adaptive Multirate Codec) type of solution. EDGE TRXs would then be capable of carrying multiple speech calls per time slot, increasing voice capacity. Also, high quality codecs, e.g 32 kbit/s, would be feasible. EDGE as a voice solution looks especially
interesting for indoor systems because of it’s scalable capacity.

Enhanced Uplink DedicatedChannel - Mobile Technology -II

Due to the non-orthogonal uplink transmission in W-CDMA the principles applied for the newly defined
transport channel Enhanced Dedicated Channel (E-DCH) are fundamentally different from HSDPA.
The shared resource in the system is the received interference at the Node B and a transmission at a single
UE can impact the raise over thermal noise as received by different Node B. Continuous uplink power
control is still an essential means of link adaptation due to the well known near-far problem. Consequently
it was decided to support soft handover for E-DCH to minimize intercell interference. Unlike HSDPA
the scheduler is not aware of the transmission buffer status, channel state and the UE transmission capabilities.
Partly this information will be signalled to the Node B via control signalling.
For the support of the new functionality several new physical channels were introduced.
• E-DPDCH: E-DCH Dedicated Physical Data Channel for dedicated uplink data transmission.
During data transmission so-called Scheduling Information such as buffer status, data priority
and power headroom can be piggybacked.
• E-DPCCH: E-DCH Dedicated Physical Control Channel with the associated control data
for E-DPDCH detection and decoding. For the support of the scheduler there is a Happy Bit that
informs if the UE has sufficient resources for transmission.
• E-HICH: E-DCH HARQ Acknowledgement Indicator Channel to transmit HARQ feedback
information (ACK/NACK)
• E-RGCH: E-DCH Relative Grant Channel to grant dedicated resources (up, down, hold) to a
UE
• E-AGCH: E-DCH Absolute Grant Channel is a shared channel that allocates an absolute resource
for one or several UE.
In Figure 3 the E-DCH data and signalling flow is illustrated. Based on the rate request (Scheduling Information
or Happy Bit) the Node B may respond with a resource allocation via the absolute or a relative
grant. The UE will use the grant for data transmission and the Node B will acknowledge the received
packets.

The HARQ protocol defined for HSDPA and for E-DCH is based on an n-channel stop-and-wait protocol.
Since out of sequence delivery is a regular event for this protocol, there is a reordering function in
place to provide in-sequence delivery to higher layer protocols. Unlike in HSDPA this function is contained
in a separate sub-layer called MAC-es. MAC-es is located in the RNC since E-DCH supports soft
handover and the packets can be received by different Node Bs. It must also be noted that the
ACK/NACK reception is not reliable and there may be unwanted repetitions or even packet losses
caused by ACK/NACK misinterpretations at the sender. In that case RLC can recover the packets if configured
in acknowledged mode (AM).

BENEFITS FOR END-TO-END PERFORMANCE
Besides an increase in radio and transport network efficiency for packet based services, HSPA improves
user perception by significantly increased peak data rates and a reduced overall latency. Peak data rates
depend on the supported reference classes. Typically the operator will upgrade the network successively.
The first terminals will be a data cards enabling 1.8 Mbit/s peak data rate.

At the final state of HSPA Release 6 deployment a maximum of 14.4 Mbps will be supported in the
downlink and 5.76 Mbps in the uplink. However it should be emphasized that the peak data rates are
temporary rates at the physical layer and neglect protocol overhead at the different layer. Furthermore an
optimistically high channel code rate at the physical layer is assumed. HSPA networks are not expected to
be deployed before 2007.
In terms of end-to-end delay significant enhancements can be expected due to fast Node-B HARQ retransmission
as well as reduced transmission time interval. Fast HARQ by the Node B will save at least
two times Iub transmission delay compared to RLC ARQ retransmission. Note that the Iub is susceptible
to congestion due to missing statistical multiplexing on the low capacity last mile. HARQ uses on synchronous
ACK/NACK feedback and does not rely on infrequent event based RLC status reports. Furthermore
the interleaving delay decreases proportionally to the TTI reduction. On the other hand the
HARQ generally operates at higher block error rate and will thus have a higher number of retransmissions.
In general there is no easy calculation of the system throughput and latency reduction due to various functions
performed at the different layer. All protocol functions must be modelled realistically to take into
account the impact of encapsulation, segmentation, retransmission, reordering etc. Results will be highly
dynamic and depend on the selected scenario and parameters. Furthermore the gain for a single link may
also not necessarily turn into improvement of overall system performance. Simulations on system level
considering multiple cells and multiple users are well established as means to evaluate system performance
in today’s complex mobile communication systems. Due to the high complexity those simulations are very
time consuming and generally run offline.
Nevertheless, in our research effort Nomor Research has implemented a standard compliant UMTS system
with the enhanced features of HSPDA and E-DCH in our RealNeS platform. The Real-time Network
Simulation (RealNeS) tool with our HSPA implementation as described above allows applications to be
tested live and even provide means to perform measurements and parameter reconfigurations in real-time.

HSPA - Mobile Broadband Today

Intro
Today’s mobile communication systems have been enhanced recently to more efficiently support packet
switched services. In UMTS HSDPA and E-DCH have been specified in downlink and uplink respectively.
By now UMTS is a well-established technology with manifold networks running globally and competitive
terminals on the market. A significant shift from traditional circuit-switched, often constant bit-rate services
to IP packet switched services is expected in the near future. UMTS Release 99, based on dedicated
resource allocation per user, is not well suited for IP packet data traffic. Therefore High Speed Packet
Downlink Access (HSDPA) and Enhanced Dedicated Channel (E-DCH) have been introduced as new
features of UMTS for Downlink and Uplink in UMTS Release 5 and Release 6, respectively. This technology
called High Speed Packet Access (HSPA) claims significant enhancements in end-to-end service provisioning
for IP based services. This introduces these future technology enhancements and assesses the
potential gains for future applications and in term user perception.
In addition to the paradigm change from using dedicated resources to making use of shared radio resources,
the main technology changes introduced are:
• Fast Node B scheduling with adaptive coding and modulation (only downlink) to exploit the varying
radio channel and interference variations and accommodate bursty IP traffic,
• Node B based Hybrid ARQ to reduce retransmission round trip times and add robustness to the
system by allowing soft combining of retransmissions,
• Reduced transmission time interval (TTI) for latency reduction and to support fast scheduler decisions
and quick HARQ retransmissions.
These added functionalities have been specified in several new MAC sub-layers and modifications of the
physical layer as is depicted


In general retransmissions are now performed directly between Node B and the User equipment (UE).
This reduces latency and saves resources on the Iub interface. The distributed scheduling performed by
RNC and Node B requires an additional scheduling buffer in the Node B as well as having an additional
flow control on the Iub interface. Furthermore the Node B needs to be made aware of certain QoS parameter
to ensure that the data transmission complies with the traffic requirements. Nevertheless HSDPA
and HSUPA can be implemented in the standard 5 MHz carrier of UMTS networks and can co-exist with
the existing 3GPP Release 99 networks. In the following sections the principles of HSDPA and E-DCH
are explained in more detail.

HIGH SPEED DOWNLINK PACKET ACCESS

In downlink a new entity called MAC-hs contains the new HSDPA functionality as seen in Figure 1. Instead
of a fixed code allocation with fast power control, the code and power resource is now shared
amongst all active HSDPA users. For this purpose a new transport channel, the High Speed Downlink
Shared Channel (HS-DSCH), has been defined that supports adaptive coding and modulation, whereby
every 2ms the transmission format can change dynamically.
In good radio channel condition 16QAM modulation can be used instead of QPSK and the rate 1/3 turbo
code may be punctured down to enable higher data rates. Depending on the UE capabilities up to 15
codes with a fixed spreading factor of 16 can be received if all codes are allocated to a single UE. Since
power control is replaced by rate control with adaptive coding and modulation the maximum data rate as
received by the user directly depends on the channel and interference conditions as well as the user position
in the cell.
The Node B scheduler must take care that fairness is maintained. The dynamic resource allocation by the
scheduler (per 2ms TTI) is signalled to the users on a new downlink control channel called High Speed
Signalling Control Channel (HS-SCCH). The following information is carried on the HS-SCCH:
• UE Identity (UE ID) via a UE specific CRC which allows addressing specific UEs on the shared
control channel.
• Transport Format and Resource Indicator (TFRI) which identifies the scheduled resource and its
transmission format.
• Hybrid-ARQ-related information to identify redundancy versions for the combining process.
Each user can monitor up to 4 HS-SCCHs. For the support of channel based scheduling and HARQ the
following feedback signalling is transmitted on the High Speed Dedicated Physical Control Channel
(HS-DPCCH) in the uplink:
• Channel Quality Information (CQI) to inform the scheduler about the instantaneous channel
condition.
• HARQ ACK/NACK information to let the sender know the outcome of the decoding process
and to request retransmissions.
Figure 2 depicts the data and signalling flow during HSDPA transmission. Based on the UE channel quality
report the Node B scheduler sends data on the shared downlink channel to the user. The UE will then
reply with an ACK or NACK message based on the outcome of the decoding. Note that the standard
does not specify scheduling and resource allocation which leaves significant freedom to Node-B implementations

Resource ReSerVation Protocol - RSVP

The RSVP protocol is part of a larger effort to enhance the current Internet architecture with support for Quality of Service flows. The RSVP protocol is used by a host to request specific qualities of service from the network for particular application data streams or flows. RSVP is also used by routers to deliver quality-of-service (QoS) requests to all nodes along the path(s) of the flows and to establish and maintain state to provide the requested service. RSVP requests will generally result in resources being reserved in each node along the data path.

A host uses RSVP to request a specific Quality of Service (QoS) from the network, on behalf of an application data stream.


RSVP carries the request through the network, visiting each node the network uses to carry the stream. At each node, RSVP attempts to make a resource reservation for the stream.

To make a resource reservation at a node, the RSVP daemon communicates with two local decision modules, admission control

and policy control. Admission control determines whether the node has sufficient available resources to supply the requested

QoS. Policy control determines whether the user has administrative permission to make the reservation. If either check fails,

the RSVP program returns an error notification to the application process that originated the request. If both checks

succeed, the RSVP daemon sets parameters in a packet classifier and packet scheduler to obtain the desired QoS. The packet

classifier determines the QoS class for each packet and the scheduler orders packet transmission to achieve the promised QoS

for each stream.

A primary feature of RSVP is its scalability. RSVP scales to very large multicast groups because it uses receiver-oriented

reservation requests that merge as they progress up the multicast tree. The reservation for a single receiver does not need

to travel to the source of a multicast tree; rather it travels only until it reaches a reserved branch of the tree. While the

RSVP protocol is designed specifically for multicast applications, it may also make unicast reservations.

RSVP is also designed to utilize the robustness of current Internet routing algorithms. RSVP does not perform its own

routing; instead it uses underlying routing protocols to determine where it should carry reservation requests. As routing

changes paths to adapt to topology changes, RSVP adapts its reservation to the new paths wherever reservations are in place.

This modularity does not rule out RSVP from using other routing services. Current research within the RSVP project is

focusing on designing RSVP to use routing services that provide alternate paths and fixed paths.

RSVP runs over IP, both IPv4 and IPv6. Among RSVP's other features, it provides opaque transport of traffic control and

policy control messages, and provides transparent operation through non-supporting regions.