GPRS/EDGE FUNDAMENTALS

Prepared by Naveen Bhartiya

(E)GPRS Optimization Network Element and Configuration Assessment

BTS GPRS territory BTS HW considerations (TRX & BB-card) BTS SW (EPCR)

BSS PCU variant & dimensioning PCU strategy in mixed configuration BSS SW and features TC HLR/ AC/ EIR MSC/VLR SGSN

HLR QoS profile GPRS settings

TCSM

Unit capacity (PAPU etc.) BSS Gb Flow control

MS/Client parameters GPRS/EDGE capability and release Multislot support RF Gs IP/MPLS/IPoATM backbone GGSN

BTS

BSC Abis Gb

2G SGSN

Gn Gi Application Applicatio Servers n Servers (co located Server load settings (Linux/Win)

RF interface Coverage C/I Capacity Traffic volume Mobility Abis interface EDAP size / dimensioning # of E1/T1s GPRS/EDGE traffic Gb interface Bearer size IP v.s. FR Dimensioning

Prepared by Naveen Bhartiya

General Packet Radio Service (GPRS) & Enhanced data rate for GPRS evolution(EDGE)
GPRS uses a packet-mode technique to transfer high-speed and low-speed data and signaling in an efficient manner. GPRS optimizes the use of network and radio resources GPRS is designed to support from intermittent and bursty data transfers through to occasional transmission of large volumes of data. GPRS uses GMSK Modulation Scheme. Enhanced Data rates for GSM Evolution (EDGE) (also known as Enhanced GPRS (EGPRS) is a digital mobile phone technology that allows improved data transmission rates as a backwardcompatible extension of GSM. EDGE is the radio technology that allows operators to increase both data speeds and throughout capacity 3 times over GPRS. EDGE uses both GMSK and 8-PSK Modulation Scheme. EDGE produces a 3-bit word for every change in carrier phase. This effectively triples the gross data rate offered by GSM. EDGE. like GPRS uses a rate adaptation algorithm that adapts the modulation and coding scheme (MCS) according to the quality of the radio channel. and thus the bit rate and robustness of data transmission.
3

Prepared by Naveen Bhartiya

DEVICES

Prepared by Naveen Bhartiya

GPRS Mobile Station(MS) modes of operation

Class A - Operates GPRS and other GSM services simultaneously.

Class B - Monitors control channels for GPRS and other GSM services simultaneously. but operates one set of services at a time.

Class C - Are connected to either GPRS service or GSM service. Must be switched manually between one or the other service.

Prepared by Naveen Bhartiya

Multislot Classes for GPRS/EGPRS

Multislot Class 1 2 3 4 5 6 7 8 9 10 11 12 30 31 32 Downlink TS 1 2 2 3 2 3 3 4 3 4 4 4 5 5 5 Uplink TS 1 1 2 1 2 2 3 1 2 2 3 4 1 2 3 Active TS 2 3 3 4 4 4 4 5 5 5 5 5 6 6 6

The multislot class determines the speed of data transfer available in the Uplink and Downlink directions.
A multislot allocation is represented as. for example. 5+2. The first number is the number of downlink timeslots and the second is the number of uplink timeslots allocated for use by the mobile station. A commonly used value is class 10 for many GPRS/EGPRS mobiles which uses a maximum of 4 timeslots in downlink direction and 2 timeslots in uplink direction. However simultaneously a maximum number of 5 simultaneous timeslots can be used in both uplink and downlink. The network will automatically configure the for either 3+2 or 4+1 operation depending on the nature of data transfer.

33
34

5
5

4
5

6
6

Under the best reception conditions. i.e. when the best EDGE modulation and coding scheme can be used. 5 timeslots can carry a bandwidth of 5*59.2 Kbit/s = 296 Kbit/s. In uplink direction. 3 timeslots can carry a bandwidth of 3*59.2 Kbit/s = 177.6 Kbit/s.
Prepared by Naveen Bhartiya 6

RF

Prepared by Naveen Bhartiya

Modulation schemes in GPRS/EDGE

GPRS system is using GMSK (Gaussian Minimum Shift Keying). a constant-envelope modulation scheme. The advantage of the constant envelope modulation is that it allows the transmitter power amplifiers to be operated in a non-linear (saturated) mode. offering high power efficiency. The saturation means that even if the input signal level is increased. no increment will be seen in the output power. as shown. 8-PSK. in the form used in EDGE. has a varying envelope. see the lower part. It means that the amplifier must be operated in the linear region in case of 8-PSK since distortion is to be avoided.(There is an additional 22.5 deg rotation to avoid zero crossing.)

Prepared by Naveen Bhartiya

Modulation schemes in GPRS/EDGE

In GMSK each bit is represented by one symbol.
In 8-PSK transmitted symbols are one of the eight sinusoids which have same amplitude and phase but differ in phase. The digital data are combined in group of 3 Bits. Thus there are 8 possible combinations starting from (0.0.0) to (1.1.1).

Each of the 3-bit patterns is then matched to one of 8-PSK Symbols. The Mapping is done in such way that there is a single bit difference between adjacent symbols. It ensures that if a symbol is received in error as an adjacent symbol only one of the bits will be in error.

(d(3k),d(3k+1),d(3k+2))=

8-PSK (Phase Shift Keying) has been

(0,1,0) (0,1,1)

(0,0,0) (0,0,1) (1,0,1)

selected as the new modulation added in EGPRS

(1,1,1)

3 bits per symbol 22.5 offset to avoid origin crossing

(called 3/8-8-PSK)

Symbol rate and burst length

(1,1,0) (1,0,0)
identical to those of GMSK

Non-constant envelope high

requirements for linearity of the power amplifier

Prepared by Naveen Bhartiya 9

Coding schemes in GPRS

GPRS provides four coding schemes: CS-1. CS-2 .CS-3. CS-4. Increasing data throughput rates

CS - 1

CS - 2

CS - 3

CS - 4

Increasing protection against errors Coding Scheme Nokia GPRS PCU Nokia GPRS PCU2 Payload (bits) per RLC block Data Rate (kbit/s)

CS1
CS2 CS3 CS4

181
268 312 428

9.05
13.4 15.6 21.4

CS1 & CS2 Implemented in all Nokia BTS without HW change CS1 & CS4 S11.5 (with PCU2) and UltraSite BTS SW CX4.1 CD1 (Talk
does not support CS3 and CS4)
Prepared by Naveen Bhartiya

More Data = Less Error Correction

10

Error Correction

Data

Coding schemes in GPRS

CS-1
MAC RLC/MAC Block Size:

CS-1 181 40 3 1/2 456 0

CS-2 268 16 6 ~2/3 588 132

CS-3 312 16 6 ~3/4 676 220

CS-2
USF

Block Check Sequence: BCS +4 Precoded USF:

CS-3

rate a/b convolutional coding

length:

puncturing
interleaving 57 57 57 Data rate (kbit/s):

456 bits 57 57 57 57 57

9.05

13.4

15.6

MAC

RLC/MAC Block Size: BCS Size: Precoded USF: Data rate (Kbit/s):

428 16 12 21.4

CS-4

USF

BCS

20 ms
Prepared by Naveen Bhartiya 11

TypeUnitOrDepartmentHere TypeYourNameHere

Coding schemes in EDGE

TypeDateHere

EDGE provides nine coding schemes: MCS-1 till MCS-9.

Scheme Code rate Header Modulation RLC blocks Raw Data Family BCS Tail HCS Data rate Code per Radio within one payloa kbit/s rate Block Radio Block d (20ms) MCS-9 1.0 0.36 8PSK MCS-8 MCS-7 MCS-6 MCS-5 MCS-4 MCS-3 MCS-2 MCS-1 0.92 0.76 0.49 0.37 1.0 0.80 0.66 0.53 0.36 0.36 1/3 1/3 0.53 0.53 GMSK 0.53 0.53 1 1 224 176 B C 11.2 8.8
12

2x592

2x12

2x6

59.2

8PSK

2 2 1 1 1 1

2x544 2x448 592 448 352 296

A B A B C A 8

54.4 44.8 29.6 22.4 17.6 14.8

12

Prepared by Naveen Bhartiya

EDGE and GPRS Master / Slave Channel Usage

Coding scheme GPRS and EDGE CS-1 CS-2 CS-3 CS-4 MCS-1 MCS-2 MCS-3 MCS-4 MCS-5 MCS-6 MCS-7 MCS-8 MCS-9 Retrans.

Bit rate (Kbps) Abis PCM allocation (fixed + pool/slave) 9.05 13.4 15.6 21.4 8.8 11.2 14.8 17.6 22.4 29.6 44.8 54.4 59.2

Higher data rates dont fit in 16 Kbit/s channels GPRS CS-2 requires 1 slave when EDGE activated (TRX/BTS) 32. 48. 64 or 80 Kbit/s Abis links per RTSL needed

EDGE

Prepared by Naveen Bhartiya

13

GPRS Link Adaptation

The LA algorithm measures the signal quality for each TBF in terms of the received signal quality (RXQUAL).

RXQUAL is measured for each received RLC block. which makes it a more accurate estimate than BLER. The PCU determines the average BLER value separately for each BTS by continuously collecting statistics from all the connections in the territory in question. Based on the estimates. the LA algorithm determines which coding scheme will give the best performance. The new LA algorithm can be used in both RLC acknowledged and un-acknowledged modes in both uplink and downlink direction.
Link Adaptation algorithm for PCU1 The GPRS Link Adaptation (LA) algorithm selects the optimum channel coding scheme (CS-1 or CS2) for a particular RLC connection and is based on detecting the occurred RLC block errors and calculating the block error rate (BLER).

Link Adaptation algorithm for PCU2 A new Link Adaptation algorithm is introduced with PCU2. which replaces the previous GPRS LA algorithm and covers the following coding schemes: CS-1 and CS-2 if CS-3 and CS-4 support is disabled in the territory in question CS-1. CS-2. CS-3. and CS-4 if CS-3 and CS-4 support is enabled in the territory .
Prepared by Naveen Bhartiya 14

EGPRS Link Adaptation

The task of the LA algorithm is to select the optimal MCS for each radio condition to maximize RLC/MAC data rate. so the LA algorithm is used to adapt to situations where signal strength and or C/I level is low and changing slowly with time. Ideal LA would follow the envelope of the throughput of different MCSs.
The PCU selects the data block and additionally selects the MCS depending on radio link quality and amount of available dynamic Abis channels. LA is done independently for each UL and DL TBF on RLC/MAC block level. but the LA algorithm is same for uplink and downlink . The MCS selection is not the same in case of initial transmission and retransmission. LA algorithm works differently for RLC acknowledged mode and unacknowledged mode.

-In Acknowledged mode. the algorithm is designed to optimize channel throughput in different radio
conditions. -In Unacknowledged mode. the algorithm tries to keep below a specified Block Error Rate (BLER) limit.

Prepared by Naveen Bhartiya

15

Logical Channels

Prepared by Naveen Bhartiya

16

Logical Channels
Common Control channel(CCH) are bidirectional , point-to multipoint ,signaling channels that are used to establish dedicated channels.
Packet Broadcast control channel (PBCCH): is a downlink-only channel for broadcasting packet data (GPRS) specific system information messages to all GPRS enabled MS in cell.

Packet paging channel(PPCH) : is a downlink only paging channel used to page the MS prior to downlink packet transfer.
Packet access grant channel(PAGCH) : is a downlink only channel used for resource assignment during the packet transfer establishment phase.

Packet random access channel(PRACH) : is an uplink only channel , which MS uses for uplink traffic channel request and for obtaining the Timing advance.
Packet data traffic channel(PDTCH) is reserved for GPRS packet data transfer. Packet associated control channel(PACCH) : is a bi-directional signaling channel dedicated for a certain MS . Packet timing advance control channel(PTCCH) : is used in uplink direction for the transmission of random access bursts to estimate the timing advance for one mobile
Prepared by Naveen Bhartiya 17

Temporary Block Flow

Temporary Block Flow (TBF): Physical connection where multiple mobile stations can share one or more traffic channels each MS has own TFI The traffic channel is dedicated to one mobile station at a time (one mobile station is transmitting or receiving at a time) Is a one-way session for packet data transfer between MS and BSC (PCU) Uses either uplink or downlink but not both (except for associated signaling) Can use one or more TSLs Comparison with circuit-switched: normally one connection uses both the uplink and the downlink timeslot(s) for traffic In two-way data transfer: uplink and downlink data are sent in separate TBFs - as below

Uplink TBF (+ PACCH for downlink TBF)

Downlink TBF (+ PACCH for uplink TBF)

BSC

PACCH (Packet Associated Control Channel): Similar to GSM CSW SACCH

Prepared by Naveen Bhartiya

Timeslot sharing by TBF

Prepared by Naveen Bhartiya

(E)GPRS Resource Allocation

Territory method is used to divide the CS and PS resources Timeslots within a cell are dynamically divided into the CS and (E)GPRS territories. Number of consecutive traffic timeslots in (E)GPRS territory are reserved (or initially available) for (E)GPRS traffic. the remaining timeslots are available for GSM voice . The dynamic variation of the territory boundary are controlled by territory parameters. The system is able to adapt to different load levels and traffic proportions. offering an optimized performance under a variety of load conditions. The PS territory can contain dedicated. default and additional capacity Dedicated capacity: number of timeslots are allocated to (E)GPRS on a permanent basis i.e. are always configured for (E)GPRS and cannot be used by the circuit switched traffic. This ensures that the (E)GPRS capacity is always available in a cell Default capacity: the (E)GPRS territory is an area that always is included in the instantaneous (E)GPRS territory. provided that the current CS traffic levels permit this Additional capacity= Additional (E)GPRS capacity means the extra time slots beyond the default capacity which are assigned due to a load demand.
Prepared by Naveen Bhartiya 20

Territory Method in (E)GPRS

TRX 1

BCCH

SD

TS

TS

TS

TS

TS

TS

Circuit Switched Territory

TRX 2

TS

TS

TS

TS

TS

TS

TS

TS

Packet Switched Territory

Max GPRS Capacity CMAX

Additional GPRS
territory

Default GPRS Capacity CDEF

Dedicated GPRS Capacity

Territory border moves based on Circuit Switched and GPRS traffic load

CDED

Prepared by Naveen Bhartiya

21

Territory Method Load Examples

Case 1: - Many (E)GPRS users - Low CS traffic

TRX 1
TRX 2

Case 2: - High CS traffic - (E)GPRS user have

TRX 1
TRX 2 Dedicated capacity

to take the rest

Case 3: - No (E)GPRS user

TRX 1

- Zero CS traffic

TRX 2

Default capacity

= CSW Territory

Prepared by Naveen Bhartiya

= (E)GPRS Territory

22

Abis Basic Concepts PCM frame (E1)

BTS BSC

Abis

One 64 kbit/s (8 bits) channel in PCM frame is called timeslot (TSL) One 16 kbit/s (2bits) channel timeslot is Sub-TSL PCM frame has 32 (E1) or 26 (E1) TSLs One Radio timeslot corresponds one 16 kbit/s SubTSL (BCCH. TCH/F etc.) and one TRX takes two TSLs from Abis One TRX has dedicated TRXsig of 16. 32 or 64 kbit/s

One BCF has dedicated BCFsig (16 or 64 kbit/s) for O&M

Q1-management needed if TRS management under BSC

MCB/LCB required if loop topology is used

Prepared by Naveen Bhartiya

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

MCB

LCB

TCH 0 TCH 4

TCH 1 TCH 5

TCH 2 TCH 6

TCH 3 TCH 7

TRX1

TRXsig BCFsig

Q1-management

(E)GPRS Dynamic Abis Pool EDAP Introduction

The resources for signaling and voice are fixed. Dynamic Abis pool (DAP) for data Predefined size 1-12 PCM TSL per DAP (24 with Flexi EDGE BTS possible). Typically used range from 4 to 8 TSL. DAP can be shared by several TRXs in the same BCF (and same E1/T1) DAP + TRXsig + TCHs have to be in same PCM UL and DL EDAP is used independently DAP schedule rounds for each active Radio Block (20 ms) Different users/RTSLs can use same EDAP Sub-TSL

0 1 2 3 4 TCH 0 5 TCH 4 6 TCH 0 7 TCH 4 8 TCH 0 9 TCH 4 10 11 12 13 14 15 EDAP 16 EDAP 17 EDAP 18 EDAP 19 EDAP 20 EDAP 21 EDAP 22 EDAP 23 24 25 TRXsig1 26 TRXsig3 27 BCFsig 28 29 30 31

TCH 1 TCH 5 TCH 1 TCH 5 TCH 1 TCH 5

TCH 2 TCH 6 TCH 2 TCH 6 TCH 2 TCH 6

TCH 3 TCH 7 TCH 3 TCH 7 TCH 3 TCH 7

TRX1 TRX2 TRX3

EDAP EDAP EDAP EDAP EDAP EDAP EDAP EDAP

EDAP EDAP EDAP EDAP EDAP EDAP EDAP EDAP

TRXsig2

EDAP EDAP EDAP EDAP EDAP EDAP EDAP EDAP

EGPRS pool

Prepared by Naveen Bhartiya

Packet Abis Solution

Packet Abis means introduction of a new transport concept:
Abis frames conveying traffic and signaling information between BTS and BSC are subject to packetization process prior to sending them to the transmission path as a result of packetization the incoming TRAU/PCU/LAPD frames are converted to new (Packet Abis specific) formats which are encapsulated and form IP packets eventually transmitted over Abis . Bandwidth is pooled for all types of traffic (no dedicated allocation anymore) e.g. PS data traffic can utilize all bandwidth available.

Benefits of Packet Abis :

Bandwidth Savings in terms of E1 saving. Reduction in congestion due to EDAP as bandwidth is pooled for all types of traffic. Reduction in PCU Utilization as resource allocation are dynamic and not hardcoded. Improvement in Latency.

Prepared by Naveen Bhartiya

25

BSS

Prepared by Naveen Bhartiya

26

Packet Control Unit PCU (Introduction)

The PCU is the BSC plug-in unit that controls the (E)GPRS radio resources, receives and transmits PCU frames to the BTSs and Frame Relay (or IP packets) to the SGSN It handles both the Gb interface and RLC/MAC protocols in the BSS and acts as the key unit in the following procedures: (E)GPRS radio resource allocation and management (E)GPRS radio connection establishment and management Data transfer (Modulation and) Coding scheme selection PCU statistics The first generation PCUs (PCU or PCU1) are optimized to meet (E)GPRS requirements, i.e. non real time solutions (QoS classes "Background" and "Interactive). EGPRS, NCCR, NACC are supported by both PCUs. The second generation PCU (PCU2) is a high capacity embedded plug-in unit that provides additional (E)GPRS processing power and extended functionality from BSS11.5 onwards. Second Generation PCUs have a new architecture. There are 3 PCU2 plug-in unit variants for the different NSN GSM/EDGE BSC variants: PCU2-U: BSCi and BSC2i PCU2-D: BSC3i 660, BSC3i 1000 and BSC3i 2000 PCU2-E for BSC3i versions including the Flexi BSC
Prepared by Naveen Bhartiya 27

Packet Control Unit (PCU) - Variants and Connectivity Limits

PCU types and capacity limits The relations between PCU and BSC types as well as the connectivity limits of BTSs, TRXs, TSLs, Abis and Gb TSLs are shown in the table:
BSCi type
BSCi, BSC2i
BSCi, BSC2i BSCi, BSC2i BSC3i

PCU type
PCU
PCU-S PCU-T PCU-B

#logical PCU per PIU

1
1 1 2

#Abis/BTS/TRX/RTSL/Gb per logical PCU

256 / 064 / 128 / 128 / 32*
256 / 064 / 128 / 128 / 32* 256 / 064 / 128 / 256 / 32* 256 / 064 / 128 / 256 / 32*

CPU/memory
166MHz / 128MB
200MHz / 128MB 300MHz / 256MB 300MHz / 256MB

PCU2-U
PCU2-D PCU2-E**

BSC2i
BSC3i BSC3i

1
2 1

256 / 128 / 256 / 256 / 32*

256 / 128 / 256 / 256 / 32* 1024 / 384 / 1024 / 1024 / 128*

450MHz / 256MB
450MHz / 256MB 1.33GHz / 1GB

* Maximum capacity of one FR link is 31 PCM TSL (31 x 64 kbps) in case of Gb over IP value gives the maximum processing capacity of one logical PCU (example 32 means 32 x 64kbps = 2048 kbps) * Each BCSU (Base Control Signaling Unit) can be equipped with 1-2 PCU for BSCi, BSC2i and BSC3i 660* 1-5 PCU for BSC3i 1000/2000* and Flexi BSC
Prepared by Naveen Bhartiya 28

Packet Control Unit (PCU) Dimensioning

Consider site with 3+3+3 configuration : Dedicated TSLS(CDED) 2+2+2 = 6 TSL Default TSLS(CDEF) 2+2+2 = 6 TSL EDAP 3 64 Kbps TSLS = 12 TSL 16 Kbps CMAX 100% Considering additional 30% TSLS (((9*8)-6)*0.3) 20 TSLS based on CMAX With 85% PCU loading and PCU2D we can have 217 TSLS Thus we can map 5 sites (44*5= 220 TSLS) in 1 NSEI or logical PCU

Prepared by Naveen Bhartiya

29

PCU2 Pooling Introduction

With PCU2 Pooling feature the operator can easily take a new PCU PIU into use in live BSC PCU2 Pooling feature introduces the Packet Service Entity (PSE) concept. PSE is a logical concept, which covers several PCU PIUs in BSC. All PCUs in PSE are serving the same NSE When the operator adds a new PCU into PSE, the system automatically configures Gb interface to new PCU and then allocates DAPs and Segments to new PCU based on each Segments (E)GPRS load. The operator does not need to do any changes to the logical network configuration in BSC and either in SGSN

Benefits for the customer: Easy way to increase (E)GPRS capacity in the network, in BSC. All PCUs are effectively in use Savings in configuration costs (no modification to logical network configuration in the BSC and neither in the SGSN) Abis and Gb resources are shared more efficiently and dynamically between PCUs in NSE (E)GPRS load is shared dynamically between PCUs within the PSE when the operator adds a new PCU into the PSE or reallocates PSE configuration Instead of configuring multiple NSEI we configured single NSEI and mapped all PCU IPs to that NSEI Post PCU Pooling Gb traffic doesn't come NS-VCI wise but is generated based on PCU IP.

Gb Interface - Introduction
The Gb interface is the interface between the BSS and the Serving GPRS Support Node (SGSN) It allows the exchange of signaling information and user data between one Packet Control Unit (PCU) or PSE (Packet Service Entity) at the BSS side and one Packet Processing Unit (PAPU) at the SGSN side Each PCU/PSE has its own separate Gb interface to the SGSN. Many users share the same physical resource. Resources are given to a user upon activity (sending/receiving) Signaling and user data are sent in the same transmission plane and no dedicated physical resources are required to be allocated for signaling purposes . Access rates per user may vary without restriction from zero data to the maximum possible line rate. One Gb interface can be implemented using the Frame Relay or IP.

Prepared by Naveen Bhartiya

31

Establishing a DL TBF and Sending Data

Paging Packet Channel Request UL TBF for MS location Immediate Assignment for UL TBF Packet Paging Response (LLC Frame)

PCH RACH AGCH PDTCH

Immediate Assignment for DL TBF

Packet Polling

AGCH PACCH

Packet Control Ack (for TA) PACCH Packet Downlink Assignment Data / Signalling BTS

PACCH
PDTCH PACCH PACCH

Ack / Nack

Prepared by Naveen Bhartiya

32

Multiple Mobiles and Downlink Transmission

The TFI included in the Downlink RLC Block header indicates which Mobile will open the RLC Block associated with its TBF

TFI3

RLC Data Block

TFI2 TFI5

BTS

TFI2

Prepared by Naveen Bhartiya

33

Establishing an UL TBF and Sending Data

Packet Channel Request RACH Immediate Assignment for UL TBF AGCH UL Data PDTCH

Signaling + Ack/Nack PACCH Final UL Data PDTCH Final Ack/Nack PACCH BTS

Packet control Ack

PACCH
Prepared by Naveen Bhartiya 34

Multiple Mobiles and Uplink Transmission

Several mobiles can share one timeslot Maximum of 7 Mobiles are queued in the Uplink Mobile transmissions controlled by USF (Uplink State Flag) sent on DL (dynamic allocation)

Uplink State Flag TS 1

TS 2 New MS TS 3

Mobile with correct USF will transmit in following Uplink block Timeslot selected to give maximum throughput
Prepared by Naveen Bhartiya 35

Multiple Mobiles and Uplink Transmission

The USF included in the Downlink RLC Block header identifies which Mobile will transmit in the following Uplink RLC Block

USF = 3

RLC Data Block

USF = 3 USF = 2

BTS

USF = 1

Prepared by Naveen Bhartiya

36

Packet Data Protocol stack

Prepared by Naveen Bhartiya

37

3G RAN Data Fundamentals

Sections:
WCDMA Overview Architecture and Call Flow RAN Dimensioning Concept of LA, RA and URA

WCDMA Overview
Multiple Access Technology for 3G is wideband CDMA (WCDMA)
All Cells Use Same Carrier Frequency Spreading Codes are used to separate Cells and Users Signal Bandwidth is 3.84 MHz

Multiple carriers can be used to increase capacity Inter-System Functionality to support mobility between GSM and WCDMA Initial version of 3G is known as Release 99 HSDPA in Downlink and HSUPA in Uplink are newer versions supporting higher data rates with the help of Fast Link Adaptation, Effective Power Control and Higher order Modulation and Coding Schemes.

WCDMA Benefits
Wideband CDMA is the Access Technology of UMTS. It offers some key benefits against GSM systems. Soft Handover
Make before Break Connection, unlike GSM. Provides greater reliability.

Processing Gain
Basic CDMA benefit => the wider the transmitted bandwidth compared to the user data rate the less power is needed for the transmission

Advanced Radio Resource Management (RRM)

RRM will control call admission and packet scheduling and all RRM building blocks are closely related to each other

Multipath Signal Processing

Combines power for increased signal integrity => RAKE receiver

Soft Handover
Soft Handover provides greater reliability to the links. It works as follows: UE is simultaneously connected to 2 to 3 cells during soft handover Soft Handover is performed based on UE Cell Pilot Power measurements and Handover thresholds set by radio network planning parameters. Soft Handover consumes both base station and transmission resources.

Concept of Spreading The Processing Gain

CDMA uses a concept of spreading the actual Information signal, which can be of different bandwidth, over a final bandwidth of 3.84 Mchips /sec. This makes it more robust and less prone to external noise, which generally is narrow band and can affect only a portion of the spread signal. This gain in robustness is termed as Processing Gain. So, more the extent of spreading, less would be the actual data rate and more would be the processing gain. Mathematically, processing gain is represented as: Processing Gain (dB) = 10* log (W/R). where W= Final Bandwidth, which is 3.84 Mcps and R = The actual Information Rate. Extent of spreading depends upon the multiplication factor used to increase the bandwidth. This is known as the spreading Factor and is explained further in the next slide

Bits, Chips and The Spreading Factor

Spreading is done by multiplying the Baseband signal with a specific spreading sequence. In the example about, each bit is multiplied by 8 chips to generate a spread signal. So, the spreading factor of the spreading code is 8. At the receivers side, the spread signal, when multiplied with the same spreading code results in the original signal. This is known as De-spreading.

Concept of Code Tree

C4(0)=[11 11] C2(0)=[11 ] C4(1)=[111-1] C0(0) =[1] C4(2)=[111-1] C2(1)=[11] C16(0)=[.... ........] C16(1)=[.... ........] C16(2)=[.... C8(1)=[1111-1-1........] C16(3)=[.... 1-1] ........] C16(4)=[.... C8(2)=[11-1-111........] C16(5)=[.... 1-1] ........] C16(6)=[.... C8(3)=[11-1-1-1-111] ........] C (7)=[.... C8(0)=[11111 111]
16

Channelization Codes used to spread information signal. These codes are orthogonal to each other. Spreading provides gain in the form of robustness and is known as Processing gain More the spreading, less the data rate. Lower SF codes are used to generate higher SF codes, which result into a code tree. If a lower SF code is being used, the codes in the branches below it are blocked.

........] C16(10)=[.... C8(5)=[1-11-1-11- .......] C16(11)=[..... 11] ......] C16(12)=[.. C8(6)=[1-1-111-1- .........] C16(13=[... C4(3)=[1-1- 11] ........] C16(14)=[.. 11] C8(7)=[1-1-11.........] C16(15)=[.. 111-1] .........]

C8(0)=[1-11-1111-1]

........] C16(8)=[.... ........] C (9)=[....

16

SF=1

SF=2

SF=4

SF=8

Channelization codes are used to distinguish between users.

Modulation & Coding Schemes Used

WCDMA offers 3 Modulation schemes viz. QPSK, 16 QAM and 64 QAM. It supports adaptive modulation, meaning the modulation to be used is selected based on the capability of the User Equipment, as well as the RF conditions that it is in. Each Modulation has its own benefits and limitations. QPSK: Known as Quadrature Phase Shift Keying 2 Bits per Symbol Can support double the data rate of BPSK with same bandwidth Is more robust in nature as compared to 16 QAM and 64 QAM Constellation Diagram of QPSK

Modulation & Coding Schemes Used contd..

16 QAM: Known as Quadrature Amplitude Modulation Combines Phase Shifting with Amplitude Modulation to support more data rates. For example symbols 0011 and 0001 have same phase but different amplitude. Similarly symbols 0000 and 1000 have different phase but same amplitude. Provides 4 bits per Symbol Is not as robust as QPSK. Needs better RF conditions, i.e. a better SNR. 64 QAM: Higher order Modulation using combination of PSK and Amplitude Modulation like 16 QAM Provides 6 bits per symbol and hence can support quite higher data rates. Needs very good RF conditions as it is more prone to errors than 16QAM or QPSK Next slide provides details of theoretical Data rates in 3G using these Modulation Schemes with different coding

Constellation Diagram of 16QAM

3G Data Rates
Release 99:
Bandwidth 3840000 3840000 3840000 3840000 3840000 3840000 3840000 3840000 Spreading Factor Channel Symbol Rate (Ksps) 512 7.5 256 15 128 30 64 60 32 120 16 240 8 480 4 960 Channel Bit Rate with QPSK(kbps) 15 30 60 120 240 480 960 1920

HSDPA
Bandwidth 3840000 3840000 3840000 3840000 3840000 3840000 3840000 Spreading Factor 16 16 16 16 16 16 16 Channel Symbol Rate (Ksps) 240 240 240 240 240 240 240 Modulation QPSK QPSK QPSK 16QAM 16QAM 16QAM 64QAM Coding Rate 1/4 2/4 3/4 2/4 3/4 4/4 Channel Bit Rate (kbps) 120 240 360 480 720 960 1440 Rate with 5 Codes(kbps) 600 1200 1800 2400 3600 4800 7200 Rate with 10 Codes(kbps) 1200 2400 3600 4800 7200 9600 14400 Rate with 15 Codes(kbps) 1800 3600 5400 7200 10800 14400 21600

HSDPA +

HSUPA
Bandwidth Coding Rate 1xSF4 (kbps) 2xSF4(kbps) 2xSF2(kbps) 2xSF2 + 2xSF4(Mbps) 3840000 1/2 960 1920 3840 2.88 3840000 3/4 960 1920 3840 4.32 3840000 4/4 960 1920 3840 5.76

HSDPA UE Categories
3GPP Release Release 5 Release 5 Release 5 Release 5 Release 5 Release 5 Release 5 Release 5 Release 5 Release 5 Release 5 Release 5 Release 7 Release 7 Category 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Max No. of HSDSCH codes 5 5 5 5 5 5 10 10 15 15 5 5 15 15 Modulation 16-QAM 16-QAM 16-QAM 16-QAM 16-QAM 16-QAM 16-QAM 16-QAM 16-QAM 16-QAM QPSK QPSK 64-QAM 64-QAM Coding Rate 0.76 0.76 0.76 0.76 0.76 0.76 0.75 0.76 0.7 0.97 0.76 0.76 0.82 0.98 Max Data Rate (Mbps) 1.2 1.2 1.8 1.8 3.6 3.6 7.2 7.2 10.1 14 0.9 1.8 17.6 21.1

Sections:
WCDMA Overview Architecture and Call Flow RAN Dimensioning Concept of LA, RA and URA

3G Network Architecture
CS domain
GSM /GPRS BSS BSC BTS PCU A HLR/AUC MSC/VLR GMSC

PSTN/other PLMN

Gb
SS 7

SCE SMS SCP

RNC NodeB

Iu-CS Iu-PS
PS backbone

SGSN

Internet, Intranet

GGSN
UTRAN CG BG Other PLMN

PS domain

3G System introduces some new Network Elements viz Radio Network Controller (RNC) and Node-B. Combined together it is known as UMTS Terrestrial Radio Access Network.

3G Call Setup Phases

RRC and RAB phases
Phase:

Call Setup divided into 3 phases:

Setup

Access

Active
Active Complete

1. 2.

Success

Setup Complete

Access Complete

3.

Setup Phase: Resource is reserved by the System Access Phase: UE confirms the Setup back to the System Active Phase: Ready for communication These 3 phases are applicable to both RRC and RAB stage RRC is about radio connection, the owner of which is RAN RAB is about the actual Bearer, which is owned by the Core Network.

Access

Active Release Active Failures

Attempts

RRC Drop

Access failures

Setup failures (blocking)

RRC Connection Setup and Access and Active Phase

UE BTS RNC CN

RRC: RRC connection Request RRC SETUP phase RRC Setup time (Resource Reservation in RNC, BTS, Transport) RRC: RRC connection Setup RRC ACCESS phase (RNC waits for Reply from UE) RRC: RRC connection Setup Complete RRC ACTIVE phase RRC: Initial Direct Transfer RANAP: Initial UE Message UE-CN Signalling (E.g. RAB Establishment and Release) RANAP: Iu Release Command

RRC: RRC connection Release

RRC: RRC connection Release Complete Release RRC resources in RNC, BTS, Transport

RAB Setup & Access and Active Phases

UE BTS

RNC

CN

RRC Connection Active Phase, UE-CN Signalling RANAP: RAB Assignment Request RAB SETUP phase RAB Setup time RAB Holding Time (Resource Reservation in RNC, BTS, Transport) RRC: Radio Bearer Setup RAB ACCESS phase (RNC waits for Reply from UE) RRC: RB Setup Complete

RANAP: RAB Assignment Response

RAB ACTIVE phase (User Plane Data Transfer)

RANAP: RAB Assignment Request with IE: RAB reconfiguration RAB Reconfiguration Actions (Reconfigure RAB resources in RNC, BTS, Transport) RRC: RB Reconfiguration RRC: RB Reconfiguration Complete RANAP: RAB Assignment Response RANAP: RAB Assignment Request with IE: RAB Release RRC: Radio Bearer Release RRC: Radio Bearer Release Complete RANAP: RAB Assignment Response Release RAB resources in RNC, BTS, Transmission

UMTS QoS Classes

When RAB is being setup, the core network (CN) provides to RNC bearer attributes like: Traffic QoS class, Maximum bit rate, Guaranteed bit rate, Residual BER, Transfer delay and so on. There are four different QoS classes defined for UMTS Conversational Class Conversational RT Preserve Time variation between information entities of the stream Conversation pattern (stringent and low delay) Voice falls under Conversational Class Streaming Class Streaming RT Preserve Time variation between information entities of the stream Streaming Video falls under Streaming Class

UMTS QoS Classes contd..

Interactive Class Interactive Best Effort Request Response Pattern Preserve Payload Content Web Browsing falls under Interactive Class Background Class Background Best Effort Destination is not expecting the data within a certain time Preserve Payload Background Download of emails falls under Background Class

Key Interfaces
Core Network Iu RNS Iur RNC Iub Node B Iub Node B Iub Node B RNC Iub Node B RNS Iu

The UTRAN consists of a set of Radio Network Subsystems connected to the Core Network through the Iu. A RNS consists of a Radio Network Controller and one or more Node Bs. A Node B is connected to the RNC through the IuB interface. Inside the UTRAN, the RNCs of the Radio Network Subsystems can be interconnected together through the IuR. Iu(s) and IuR are logical interfaces. IuR can be conveyed over direct physical connection between RNCs or virtual networks using any suitable transport network.

Key Interfaces contd..

IuB Interface: Logical Interface between NodeB and RNC Manages Transport Resources Handles Logical O&M of the NodeB Manages traffic of Common & Dedicated Channels IuR Interface: Logical Interface between two RNCs Manages Transport Network Manages traffic of Dedicated Channels and reporting Measurements

Iu Interface: Logical Interface between RAN and Core Network. Responsible for establishing , maintaining and releasing RABs. Also responsible for performing handovers and serving RNC relocations. Paging is also handled by Iu interface. It has two instances:
Iu-CS Interface: Between RAN and Circuit Switched domain in the Core Network Carries communication between RAN & MSC and also UE & MSC Iu-PS Interface: Between RAN and Circuit Switched domain in the Core Network Carries communication between RAN & SGSN and also UE & SGSN

Sections:
WCDMA Overview Architecture and Call Flow RAN Dimensioning Concept of LA, RA and URA

Radio Dimensioning Data Flow

Radio Dimensioning follows the process as shown in the flow chart below, which is based on both Coverage and Capacity Requirements.

Capacity Dimensioning of Node-B ,RNC and the interface between them, i.e. IuB is covered in subsequent slides

Node-B Overview
Specifications:
Up to 3 RF Modules per system Module Up to 2 System Modules 1 Tx sub module per System Module Up to 2 AC/DC Power Modules Optional Outdoor Cabinet

Main functions of RF Module:

Antenna Filtering Power Amplification (Transmitter) Low Noise Amplification (Receiver) Combiner (Carriers)

Node-B Dimensioning
Channel Elements of a Node-B reside inside the System Module. So, Capacity Dimensioning from a HW point of view is done considering the CE capacity of a System Module, which has three variants System Module No. of RF Modules HW Support for No. of Cells CCH Max HW CE Capacity Max SW Capacity (traffic) Total Max SW Capacity with 2 SM

FSMB
FSMC FSMD

3
3 3

6
6 12

240
250 500

240
180 396

480
360 792

Number of Channel Elements Required during a call depend upon the RAB type. Following slide provides the details.

RAB wise CE Requirements

RAB Traffic Class CS/PS Max SF UL rates for each RAB
12.2 7.95 5.9 4.75 12.65 8.85 6.65 16 32 64 128 256 64 64 64 64 64 64 64 64 32 16 8 4

SF DL

CEs UL

CEs DL

AMR Speech AMR Speech AMR Speech AMR Speech AMR Speech AMR Speech AMR Speech Packet Packet Packet Packet Packet

Conversational Conversational Conversational Conversational Conversational Conversational Conversational INT / BG INT / BG INT / BG INT / BG INT / BG

CS CS CS CS CS CS CS PS PS PS PS PS

128 128 128 128 128 128 128 128 64 32 16 8

1 1 1 1 1 1 1 1 2 4 4 8

1 1 1 1 1 1 1 1 2 4 4 8

Packet
UDI Streaming Streaming

INT / BG
Conversational Streaming Streaming

PS
CS CS CS

384
64 57.6 14.4

4
16 16 64

8
32 32 128

16
4 4 1

16
4 4 1

RNC: Overview
Radio Network Controller performs the following Key Functions: Radio Resource Management Telecom Transmission and Transport O&M WCDMA Radio Resource Management can be broken down into following functions: Resource Manager Admission Control Load Control Power Control Handover Control Packet Scheduler

Telecom Functions can be broken into: Security Functions: Integrity Checking, Ciphering User Plane Processing towards CS and PS (e.g. Management of RABs) Radio Network Layer Control Processing Service Area Broadcast Location Services

RNC Dimensioning
RNC Capacity is licensed as:
RNC Model

IuB PS Data Throughput AMR Capacity Number of carriers, BTS or cells

NodeB Equipped Capacity 900 900 600

IuB Throughput Capacity (Mbps) 900 900 600

AMR erlangs Capacity (Erl) 11700 11700 7800

MODEL3 MODEL3 MODEL2

Generally RNC Dimensioning is done considering the factors as given in the below flowchart.

IuB Dimensioning - ATM

Each Tellabs 8660 caters 50% of the total NodeBs MSP 1+1 between STM cards in RNC, Tellabs and ALU MUXs All STM cards have optical STM1 interfaces Each NodeB connected to 4 E1s (17960 cps) to Tellabs 8660. 15 NodeBs (60 E1s) on one ch. STM-1 link to Tellabs. Tellabs RNC interface are overbooked with 19 - 25 NodeBs on 1 VC4 link to RNC. In RU10/RU20, Overbooking is supported only for UBR+ VPs carrying NRT traffic.

IuB Dimensioning Dual IuB Ethernet

Each Tellabs 8660 caters 50% of the total NodeBs 2N redundancy for NPGE cards Each GE Module has 8 x 1000 opt. interfaces ELP redundancy for GE modules b/w RNC and Tellabs. Each Dual IuB NodeB is planned with 17Mbps Ethernet bandwidth for HSPA and R99 traffic. All RT traffics along with Signaling & DCN are kept on ATM. All Dual IuB sites have HSPA fallback feature enabled

Sections:
WCDMA Overview Architecture and Call Flow RAN Dimensioning Concept of LA, RA and URA

Location Area & Routing Area

Location Area (LA) and Routing Area (RA) are used by the Core Network to track UEs LA are used by CS Domain whereas RA are used by PS Domain The main CS Service States are CS-Detached, CS Idle and CS-Connected The main PS Service States are PS-Detached, PS-Idle and PS-Connected

Location Areas: A UE in CS Idle Mode does not have to update the CS Core of its location when moving within an LA. LA consists of one or more RNCs connected to the same CN, i.e. MSC / VLR. The mapping between a LA and its associated RNCs is handled by the MSC/VLR The mapping between LA and its Cells is handled by the RNC. A LA is identified using a Location Area Identity
Routing Areas: A UE in PS Idle Mode does not have to update the PS Core of its location when moving within an RA. RA consists of one or more RNCs connected to the same CN, i.e. SGSN. A RA is always contained within a single LA. The mapping between a RA and its associated RNCs is handled by the SGSN The mapping between a RA and its cells is handled by the RNC A RA is identified using a Routing Area Identity

UTRAN Registration Area

RU 10 onwards, RNC supports URA_PCH State The purpose of this state is to decrease the cell update signaling due to cell reselection , which saves RNC and UE resources When the UE is in Cell_FACH or Cell_PCH state, its location is known at the cell level. Cell updates are sent by the UE when Cell reselection occurs. If too many cell updates are received in a pre-defined window, the UE is ordered to transfer to URA_PCH state in order to reduce cell update signaling between the UE and the RNC. In URA_PCH state, UE sends the URA update to RNC after reselection to new URA. Planning of URA involves a balance between paging load and signaling load. Large URA would increase the signaling load whereas small URA will lead to frequent URA updates which increases signaling load as well as UE power consumption.

THANKS

Prepared by Naveen Bhartiya

71