:.,: PRE,,JTICE
`o e HALL
`
`Fundamentals of LTE
`
`Arunabha Ghosh • Jun Zhang
`Jeffrey G. Andrews • Rias Muhamed
`
`Foreword by Rajiv Laroia
`
`Samsung Ex. 1007
`
`

`

`Many of the designations used by manufacturers and sellers to distinguish their products
`are claimed as trademarks. Where those designations appear in this book, and the pub(cid:173)
`lisher was aware of a trademark claim, the designations have been printed with initial
`capital letters or in all capitals.
`
`The authors and publisher have taken care in the preparation of this book, but make
`no expressed or implied warranty of any kind and assume no responsibility for errors or
`omissions. No liability is assumed for incidental or consequential damages in connection
`with or arising out of the use of the information or programs contained herein.
`
`Credits and permissions appear on pages 417 and 418.
`
`The publisher offers excellent discounts on this book when ordered in quantity for bulk
`purchases or special sales, which may include electronic versions and/or custom covers
`and content particular to your business, training goals, marketing focus, and branding
`interests. For more information, please contact:
`U.S. Corporate and Government Sales
`(800) 382-3419
`corpsales@pearsontechgroup.com
`For sales outside the United States please contact:
`International Sales
`international@pearson.com
`Visit us on the Web: informit.com/ph
`Library of Congress Cataloging-in-Publication Data
`
`Fundamentals of LTE / Arunabha Ghosh ... [et al.].
`p. cm.
`Includes bibliographical references and index.
`ISBN-10: 0-13-703311-7 (hardcover: alk. paper)
`ISBN-13: 978-0-13-703311-9 (hardcover: alk. paper) 1. Long-Term Evolution
`(Telecommunications) I. Ghosh, Arunabha, 1969-
`TK5103.48325.F86 2010
`621.3845'6-dc22
`Copyright © 2011 Pearson Education, Inc.
`All rights reserved. Printed in the United States of America. This publication is protected
`by copyright, and permission must be obtained from the publisher prior to any prohibited
`reproduction, storage in a retrieval system, or transmission in any form or by any means,
`electronic, mechanical, photocopying, recording, or likewise. For infonnation regarding;
`permissions, write to:
`Pearson Education, Inc
`Rights and Contracts Department
`501 Boylston Street, Suite 900
`Boston, MA 02116
`Fax: (617) 671-3447
`ISBN-13: 978-0-13-703311-9
`ISBN-10:
`0-13-703311-7
`
`2010021369
`
`Text printed in the United States on recycled paper at Courier in Westford, Massachusetts.
`First printing, August 2010
`
`Samsung Ex. 1007
`
`

`

`9.6 Scheduling and Resource Allocation
`
`339
`
`9.6 Scheduling and Resource Allocation
`
`The main purpose of scheduling and resource allocation is to efficiently allocate the
`available radio resources to UEs to optimize a certain performance metric with QoS
`requirement constraints. Scheduling algorithms for LTE can be divided into two
`categories:
`
`• Channel-dependent scheduling: The allocation of resource blocks to a UE is
`based on the channel condition, e.g., proportional fairness scheduler, max CI (Car•
`rier to Interference) scheduler, etc.
`
`• Channel-independent scheduling: The allocation of resource blocks to a UE is
`random and not based on channel condition, e.g., round-robin scheduler.
`
`In a multicarrier system such as LTE, channel-dependent scheduling can be further di·
`vided into two categories:
`
`• Frequency diverse scheduling: The UE selection is based on wideband CQI.
`However, the PRB allocation in the frequency domain is random. It can exploit
`time selectivity and frequency diversity of the channel.
`
`• Frequency selective scheduling: The UE selection is based on both wideband
`and subband CQI, and the PRB allocation is based on the subband CQI. This can
`exploit both time and frequency selectivity of the channel.
`
`In this section, we mainly focus on the frequency selective scheduling.
`Dynamic channel-dependent scheduling is one of the key features to provide high
`spectrum efficiency in LTE. To better exploit the channel selectivity, the packet scheduler
`is located in the eNode•B, which allocates physical layer resources for both the DL(cid:173)
`SCH and UL-SCH transport channels every TTI. Resource assignment consists of PRBs
`and MCS. Such scheduling depends heavily on the channel information available at the
`eNode-B, which is provided by the uplink CQI reporting for the downlink channel and
`by channel sounding for the uplink channel, as discussed in Section 9.2 and Section 9.4,
`respectively. The scheduler should also take account of the traffic volume and the QoS
`requirement of each UE and associated radio bearers. Due to the implementation of
`OFDMA/SC-FDMA, LTE is able to exploit the channel variation in both the time and
`frequency domain, which is a major advantage compared to HSPA, which is able to
`exploit channel variation only in the time domain.
`The objective of channel-dependent scheduling, as discussed in Chapter 4, is to ex(cid:173)
`ploit multiuser diversity to improve the spectrum efficiency. Meanwhile, it should also
`consider such issues as fairness and QoS requirements. In addition, scheduling is tightly
`integrated with link adaptation and the H-ARQ process. The scheduling algorithm is not
`standardized and is eNode-B vendor specific. See [10-13] for the investigation of different
`scheduling schemes in LTE, and refer to Chapter 4 for related discussion. In this section,
`we focus on the signaling for both downlink and uplink scheduling. Dynamic scheduling
`is mainly applied on the data traffic, which is the focus in this section. The scheduling
`of VoIP services will be discussed in Section 9.7.
`
`Samsung Ex. 1007
`
`

`

`340
`
`Chapter 9 111 Physical Layer Procedures and Scheduling
`
`9.6.1 Signaling for Scheduling in Downlink and Uplink
`For both downlink and uplink, the eNode-B scheduler dynamically controls which time(cid:173)
`frequency resources are allocated to a certain UE. The resource assignments, including
`the assigned time/frequency resources and respective transmission formats, are conveyed
`through downlink control signaling. The minimum size of radio resource that can be
`allocated to a UE corresponds to two resource blocks, which is 1 ms duration in the
`time domain and 180kHz in the frequency domain. Both downlink and uplink employ
`orthogonal transmission, so each resource block is allocated to a single UE except in
`the MU-1UMO mode. Both localized and distributed resource allocations are supported
`in the downlink, while in the uplink UEs are always assigned contiguous resources, i.e.,
`only localized allocation is supported. In addition, there is a strict constraint on the UE
`transmit power in the uplink, which is subject to the uplink power control that will be
`discussed in Section 9.10.
`
`Signaling for Downlink Scheduling
`
`The channel state information at the eNode-B for the downlink scheduling is obtained
`through CQI reporting from UEs, as discussed in Section 9.2. To enable frequency selec(cid:173)
`tive scheduling, subband CQI reporting is required. The eNode-B dynamically allocates
`resources to UEs at each TTI. A UE always monitors the PDCCH for possible alloca(cid:173)
`tions. For dynamically scheduled data traffic, the UE is configured by the higher layers
`to decode the PDCCH with CRC scrambled by the C-RNTL2 The UE shall decode the
`PDCCH and any corresponding PDSCH according to the respective combinations de(cid:173)
`fined in Table 9.16. For example, when a UE configured in transmission mode 3 or 4 (OL
`and CL spatial multiplexing) receives a DCI format lA assignment, it shall assume that
`the PDSCH transmission is associated with transport block 1 and that transport block
`2 is disabled, and transmit diversity is applied. The DCI carries the downlink scheduling
`assignment and other information necessary to decode and demodulate data symbols.
`The transport channel processing of DCI was described in Section 7.3.
`As shown in Section 6.3.3, in the downlink, while the two distributed allocation types
`(resource allocation type O and type 1) provide better performance with a high overhead,
`the localized allocation type (resource allocation type 2) provides a low overhead alter(cid:173)
`native at the cost of limited scheduling flexibility. The UE shall interpret the resource
`allocation field depending on the PDCCH DCI format detected. PDCCH DCI formats 1,
`2, and 2A with type O and with type 1 resource allocation have the same format and are
`distinguished via the single bit resource allocation header field, where type O is indicated
`by O value and type 1 is indicated otherwise. PDCCH with DCI format lA, lB, IC,
`and lD have a type 2 resource allocation while PDCCH with DCI format 1, 2, and 2A
`have type O or type 1 resource allocation. The details of the resource assignment can be
`interpreted from DCI for different formats.
`To determine the modulation order and transport block size, the UE shall first read
`the 5-bit "modulation and coding scheme" field (IMcs) in the DCI, based on which a
`Transport Block Size (TBS) index can also be determined. The mapping between the
`MCS index 1Mcs, the modulation order, and TBS index ITBS for PDSCH is shown in
`
`2 This is the unique identification used for identifying RRC connection and scheduling.
`
`Samsung Ex. 1007
`
`

`

`9.6 Scheduling and Resource Allocation
`
`341
`
`Mode 1
`
`l'vlode 2
`
`Mode 3
`
`Mode 4
`
`Mode 5
`
`Mode 6
`
`Table 9.16 PDCCH and PDSCH Configured by C-RNTI
`UE DL Transmission
`DCI
`Transmission Scheme
`Format
`of PDSCH
`Mode
`DCI format lA Single-antenna port, port 0
`Single-antenna port, port 0
`DCI format I
`DCI format IA Transmit diversity
`Transmit diversity
`DCI format I
`DCI format IA Transmit diversity
`DCI format 2A 01 spatial multiplexing
`or transmit diversity
`DCI format IA Transmit diversity
`DCI format 2 CL spatial multiplexing
`or transmit diversity
`DCI format IA Transmit diversity
`DCI format 1D Multiuser IvIIMO
`DCI format IA Transmit diversity
`DCI format 1B Closed-loop Rank= 1
`precoding
`If the number of PBCH
`antenna ports is one,
`single-antenna port (port O);
`otherwise, transmit diversity
`Single-antenna port, port 5
`
`Mode 7
`
`DCI format IA
`
`DCI format 1
`
`Table 9.17. The TBS can then be determined based on ITBS and the total number of
`allocated PRBs. Note that in Table 9.17 different MCS indices may be mapped to the
`same TBS, e.g., hICs = 9, 10 are mapped to IrBs = 9, resulting in the same data rate.
`Such modulation overlap is adopted to improve the performance around the modulation
`switching points, as different combinations of modulation and coding with the same rate
`may provide different performance in different scenarios. For 29 5 11\1cs 5 31, the TBS
`is determined from the previous scheduling grant for the same transport block using
`O 5 hws 5 28.
`
`Signaling for Uplink Scheduling
`
`In the uplink, the channel state information is estimated at the eNode-B with the help of
`sounding reference signals, as discussed in Section 9.4. AUE always monitors the PDCCH
`in order to find possible allocation for uplink transmission. Only contiguous resource
`blocks can be allocated to a UE due to the SCFDMA nature of the UL transmission'.
`Frequency hopping can be applied to provide additional diversity. The UE obtains the
`uplink resource allocation as well as frequency hopping information from the uplink
`scheduling grant received four subframes earlier, i.e., if the UE detects a PDCCH with
`DCI format O in subframe n intended for this UE, it will adjust the corresponding PUSCH
`transmission in subframe n + 4 accordingly.
`
`Samsung Ex. 1007
`
`

`

`342
`
`Chapter 9 (cid:127) Physical Layer Procedures and Scheduling
`
`Table 9.17 Modulation and TBS Index for PDSCH
`MCS Index IMcs Modulation Order TBS Index lTBs
`0
`2
`0
`1
`2
`1
`2
`2
`2
`3
`2
`3
`4
`2
`4
`5
`2
`5
`2
`6
`6
`2
`7
`7
`2
`8
`8
`2
`9
`9
`10
`4
`9
`4
`11
`10
`12
`4
`11
`13
`4
`12
`14
`13
`4
`15
`4
`14
`4
`16
`15
`17
`15
`6
`18
`16
`6
`17
`6
`19
`20
`6
`18
`21
`6
`19'
`22
`20
`6
`23
`21
`6
`24
`22
`6
`25
`23
`6
`26
`6
`24
`27
`6
`25
`28
`26
`6
`29
`2
`30
`4
`31
`6
`
`reserved
`
`To determine the modulation order, redundancy version, and transport block size
`for the PUSCH, the UE shall first read the 5-bit "modulation and coding scheme and
`redundancy version" field (h1es) in the DCI., The mapping between I Mes, modulation
`order, and IrBs for the PUSCH is shown in Table 9.18. Note that lp.,JCS also indi(cid:173)
`cates the H-ARQ redundancy version. The redundancy version 1, 2, or 3 is indicated by
`I MCS = 29, 30, 31, respectively, in which case the modulation order is assumed to be
`
`Samsung Ex. 1007
`
`

`

`9.6 Scheduling and Resource Allocation
`
`343
`
`Table 9.18 Modulation, TBS Index, and Redundancy Version for PUSCH
`MCS Index IMcs Modulation Order TBS Index ITBs Redundancy Version
`2
`0
`0
`0
`1
`2
`1
`0
`2
`2
`2
`0
`2
`3
`0
`3
`2
`0
`4
`4
`2
`5
`5
`0
`2
`6
`6
`0
`2
`7
`0
`7
`2
`8
`0
`8
`g
`g
`D
`2
`0
`2
`10
`10
`11
`4
`0
`10
`11
`12
`4
`0
`4
`12
`0
`13
`4
`13
`0
`14
`4
`14
`0
`15
`15
`0
`16
`4
`0
`17
`16
`4
`0
`17
`18
`4
`0
`18
`4
`19
`20
`4
`19
`0
`21
`19
`0
`6
`20
`22
`0
`6
`21
`0
`23
`6
`22
`0
`24
`6
`23
`0
`25
`6
`6
`24
`0
`26
`25
`27
`6
`0
`26
`28
`6
`0
`reserved
`reserved
`1
`29
`reserved
`reserved
`30
`2
`reserved
`reserved
`3
`31
`
`the one indicated in the initial grant. Similar to the downlink, ther"l is also modulation
`overlap around the switching points, e.g., I Mes= 10, 11 are both mapped to IrBs = 10.
`The transport block size can be determined from !ivJCs and IrBS• For 20 :'.S; I Mes :'.S; 31,
`the transport block size is assumed to be as determined from DCI transport in the initial
`PDCCH for the same transport block using O :'.S; Iucs :'.S; 28.
`
`Samsung Ex. 1007
`
`

`

`344
`
`Chapter 9 (cid:127) Physical Layer Procedures and Scheduling
`
`9.6.2 Multiuser MIMD Signaling
`If MU-MIMO is used in the uplink, then it is transparent to the UE with the exception
`that two UEs should transmit orthogonal reference signals in order for the eNode-B
`to separate them. The uplink resource allocation is indicated on PDCCH using DCI
`format 0, which contains a 3-bit field to indicate the cyclic shift in the reference signal
`to be used by each UE.
`When MU-MIMO is used in the downlink, two rank-1 UEs are multiplexed on the
`same physical resource. Unlike SU-MIMO, in this case the power for each UE is reduced
`by 3 dB. This is indicated by the power offset field in DCI format lD, which is used for
`MU-MIMO scheduling.
`
`9. 7 Semi-persistent Scheduling for VoIP
`
`Although the revenue from data services is increasing, voice services still provide the ma(cid:173)
`jority of operators' revenue. In current cellular systems, including WCDMA and HSPA,
`there is a Circuit Switched (CS) domain to provide efficient CS voice service. On the
`contrary, LTE is based on all-IP packet switching, and there is no CS domain. For the
`commercial success of LTE, it should be able to provide voice quality comparable to that
`based on CS. In LTE, VoIP can be used to provide voice service. In this section, we dis(cid:173)
`cuss the challenge of scheduling VoIP packets, and present the semi-persistent scheduling
`that is included in LTE [4, 5].
`
`9.7.1 Motivation for Semi-persistent Scheduling
`VoIP is characterized by the small packet size with very stringent delay and jitter re(cid:173)
`quirements, and a large number of simultaneous users. In LTE, the scheduling of the data
`traffic is based on the shared-channel transmission and the channel-dependent resource
`allocation. The eNode-B allocates radio resources in units of resource blocks to UEs,
`and informs the scheduling information to each UE on PDCCH. Such dynamic channel(cid:173)
`dependent scheduling can exploit the selectivity in both time and frequency domain, and
`significantly improve the system throughput. However, to apply dynamic scheduling di(cid:173)
`rectly for VoIP packets is not desirable. This is mainly due to the small packet size and
`the constant inter-arrival time of VoIP packets, which results in large control overhead
`as control information should be transmitted per TT!. Many UEs may be scheduled si(cid:173)
`multaneously, so the limited resources available on PDCCH for control signaling become
`a bottleneck for the VoIP capacity.
`To reduce the amount of control signaling, persistent scheduling can be applied, where
`a fixed allocation is valid until the next allocation is informed. With persistent schedul(cid:173)
`ing, the H-ARQ operation can be implemented by sending each packet a fixed number
`of times, and then no H-ARQ ACK/NAK is required. However, the inflexibility of such
`resource allocation degrades the performance compared to the dynamic scheduling, espe(cid:173)
`cially due to the mismatch between the allocated resource and the actual radio channel
`for multiple H-ARQ processes.
`Semi-persistent scheduling is proposed to exploit the benefits of both dynamic and
`persistent scheduling: the initial transmissions are based on persistent scheduling, while
`
`Samsung Ex. 1007
`
`