`Exhibit 1014
`
`ZTE Corporation and ZTE (USA) Inc.
`ZTE Corporation and ZTE (USA) Inc.
`
`
`
`Inv. No. 337-TA-868
`RX-3033
`
`High Speed Downlink Packet Access: WCDMA Evolution
`
`Troels Emil Kolding, Klaus Ingemann Pedersen, Jeroen Wigard, Frank Fredefiksen
`& Preben Elgaard Mogensen, Nokia Networks
`
`This article gives an overview of the high speed downlink
`packet access (tISDPA) concept; a new feature which is com-
`ing to the Release 5 specifications of the 3GPP
`WCDMA/ UTRA-FDD standard. To support an evolution to-
`wards more sophisticated network and multimedia services,
`the main target of HSDPA is to increase user peak data rates,
`quality of seruice, and to generally improve spectral efficiency
`for downlink asymmetrical and bu~ty packet data services.
`This is accomplished by introducing a fast and complex chan-
`nel control mechanism based on a short and fixed packet
`transmission time interval (TTI), adaptive nmd~lation and
`coding (AMC), and fast physical layer (L1) hybrid ARQ. To fa-
`cititate fast scheduling with a per- TTI resolution in coherence
`with the instantaneous air interface toad, the HSDPA-related
`MAC functionality is moved to the Node-B. The HSDPA con-
`cept [~cilitates peak dgtta rates exceeding 2 Mbps (theoreti-
`cally up to and exceeding 10 Mbps), and the cell throughput
`gain over previous UTRA-FDD releases has been evaluated to
`be in the order of 50-100% or even more, highly dependent on
`factors such as the radio environment and the service provi-
`sion strategy of the network ~perator.
`
`Introduction
`Data services are anticipated to have an enourmous rate of
`growth over the next years (the so-called data tornado) and
`will likely become the dominating source of traffic load in 3G
`mobile cellular networks. Example applications to supple-
`ment speech services include multiplayer games, instant
`messaging, online shopping, face-to-face videoconferences,
`movies, music, as well as personal]public database access.
`As more sophisticated services evolve, a major challenge of
`cellular systems design is to achieve a high system capacity
`and simultaneously facilitate a mixture of diverse services
`with ve~ different quality of service (QoS) requirements.
`Various traffic classes exhibit very different traffic symme-
`try a~d bandwidth requirements, l~or example, two-way
`speech services (conversational class) require strict adher-
`ence to channel symmetry and veI3.~ tight latency, while
`lnternet download services (background class) are often
`asymmetrical and are tolerant to latency. The streaming
`class, an the other hand, typically exhibits tight latency re-
`quirements with most of the traffic carried in the downlink
`direction.
`Already in Release 99 of" the WCDMA/UTRA specifica-
`tions, there exist several types of downlink radio bearers to
`facilitate efficient transportation of the different service
`classes. The forward access channel (FACH) is a common
`channel offering low latency. However, as it does not apply
`fast closed loop power control it exhibits limited spectral el-
`ficiency and is in practice limited to carrying only small data
`amounts. The dedicated channel (DCH) is the "basic" ra-
`
`die-bearer in WCDMA/UTRA and supports all traffic
`classes due to high parameter flexibility. The data rate is up-
`dated by means of variable spreading factor (VSF) while the
`block error rate (BLER) is controlled by inner and outer loop
`power control mechanisms. However, the power and hard-
`ware efficiency of the DCH is limited for bursty and high
`data rate services since channel reconfiguration is a rather
`slow process (in the range of 500 ms). Hence, for certain
`Internet services with high maximum bit rate allocation the
`DCI-I channel utilization can be rather low. To enhance
`trunking efficiency, the downlink shared channel (DSCH)
`provides the possibility to time-multiplex different users (as
`opposed to code multiplexing) [1]. The benefit of the DSCH
`over the DCH is a fast channel reconfiguration time and
`packet scheduling procedure (in the order of 10 ms inter-
`vals). The efficiency of the DSCH can be significantly higher
`than for the DCH for bursty high data rate traffic [2].
`The HSDPA concept can be seen as a continued evolution
`of the DSCH and the radio bearer is thus denoted the high
`speed DSCH (HS-DSCH) [3]. As will be explained in the fol-
`lowing sections, the HSDPAconcept introduces several adap-
`ration and control mechanismsin order to enhance peak data
`rates, spectral efficiency, as well as QoS control for bursty
`and downlink asymmetrical packet data [4]. In this paper, is-
`sues of importance to radio resource management (RRM) are
`discussed and UE capability implications are introduced.
`Next, the potential pe~ormance of the HSDPA concept is
`evaluated for different environments before the paper is con-
`cluded with a short discussion of further HSDPA enhance-
`ments proposed for future 3GPP standard releases. At the
`time of this writing, the Release 5 specifications have not yet
`been frozen so the specific details may be subject to change.
`
`Concept Description
`The fundamental characteristics o[" ~he HS-DSCH and the
`DSCH are compared in Table 1. On the HS-DSCH, two fun-
`damental CDMA features, namely variable spreading factor
`(VSF) and fast power control, have been deactivated and re-
`placed by AM C, short packet size, multi-code operation, and
`fhst L 1 hybrid ARQ (HARQ). While being more complicated,
`the replacement of fast power control with fast AMC ~elds a
`power efficiency gain due to an elimination of the inherent
`power control overhead. Specifically, the spreading factor
`(SF) has been fixed to 16, which gives a good data rate reso-
`lution with reasonable complexity. In order to increase the
`link adaptation rate and efficiency of the AMC, the packet
`duration has been reduced from normally 10 or 20 ms down
`to a fixed duration of 2 ms. To achieve low delays in the link
`control, the MAC functionality for the HS-DSCH has been
`moved from the RNC to the Node-B. This is a noticeable ar-
`chitectural change compared to the Release 99 architecture.
`
`© 2003 IEEE
`
`IEEE Vehicular Technology Society News ,~ February 2003
`
`N K8681TCO 11070377
`ZTE Corporation and ZTE (USA) Inc.
`Exhibit 1014-00001
`
`
`
`Feature
`
`DSCH
`
`HS-DSCH
`
`Variable spreading factor (VSF)
`
`Yes (4-256)
`
`No (16)
`
`Fast power control
`
`Yes
`
`Adaptive modulation and coding (AMC)
`
`No
`
`Fast L1 HARQ
`
`Multi-code operation
`
`No
`
`Yes
`
`No
`
`Yes
`
`Yes
`
`Yes, extended
`
`Transmission time interval (TTI)
`
`10 or 20 ms
`
`2 ms
`
`Location of MAC
`
`RNC
`
`Node-B
`
`Table 1 Comparison of fundamental propeties of DSCH
`and HS-DSCH.
`
`This leads to significant software and hardware changes
`compared to the existing Release 99 Node-B implementa-
`tion. For the HS-DSCH, only hard handover is supported as
`of the Release 5 specifications.
`
`Modulation and coding options in HSDPA
`To substitute the functionality of fast power control and
`VSF, the modulation, coding, and multi-code part of HSDPA
`must cover a wide dynamic range corresponding to the
`channel quality variations experienced at the UE (including
`fast as well as distance-dependent variations). The means of
`adaptation are the code rate, the modulation scheme, the
`number of multi-codes employed, as well as the transmit
`power per code. The HS-DSCH encoding scheme is based on
`the Release 99 rate-l/3 Turbo encoder but adds rate match-
`ing with puncturing and repetition to obtain a high resolu-
`tion on the effective code rate (approximately from 1/6 to
`1/1). To facilitate very high peak data rates, the HSDPA con-
`cept adds 16QAM on top of the existing QPSK scheme avail-
`able in Release 99. The combination of 16QAM and e.g.
`rate-¾ channel encoding enables a peak data rate of 712
`kbps per code (SF=16). Higher robustness is available with
`a QPSK rate-lA scheme but at the penalty of having only a
`119 kbps data rate per code. A modulation and coding combi-
`nation is sometimes denoted a transport format and re-
`source combination (TFRC). Five example TFRCs available
`on the HS-DSCH are shown in Table 2. Given sufficiently
`good channel conditions, a single user may simultaneously
`receive up to 15 parallel multi-codes leading to very high
`peak data rates up to 10.8 Mbps. This is the maximum peak
`data rate supported by the HSDPA concept, which can only
`be achieved in very favourable environments or with ad-
`vanced transmission and reception technologies. It will be
`shown in a later section that the HSDPA concept defines a
`number of UE capability classes, and that only the high-end
`UE classes will support the very high data rates.
`
`The dynamic range of the AMC for a single code is illus-
`trated in Figure 1 showing the available user data rate ver-
`sus the instantaneous (per-TTI) Es/No. The curve includes
`the gain from fast HARQ based on chase combining which
`significantly improves the throughput at low Es/No values.
`
`TFRC
`
`Modulation Eft. Code Data rate Data rate Data rate
`Rate
`(1 code)
`(5 codes)
`(15 codes)
`
`1
`
`2
`
`3
`
`4
`
`5
`
`QPSK
`
`QPSK
`
`QPSK
`
`16QAM
`
`16QAM
`
`1A
`
`1~
`
`~A
`
`~A
`
`~A
`
`119 kbps
`
`0.6 M?ops
`
`1.8 Mbps
`
`237 kbps
`
`1.2 M?ops
`
`3.6 Mbps
`
`356 kbps
`
`1.8 M?~ps
`
`5.3 Mbps
`
`477 kbps
`
`2.4 M?ops
`
`7.2 Mbps
`
`712 kbps
`
`3.6 M?~ps
`
`10.8 M?Dps
`
`Table 2 Example transport format and resource combinations
`and corresponding user data rates at layer I (including over-
`head), More options for TFRCs are given in [1,3].
`
`February 2003 ÷t÷ IEEE Vehicular Technology Society News
`
`The Es/No range from where the throughput of TFRC1 is
`larger than 32 kbps to where the data rate of TFRC5 satu-
`rates to the maximum throughput of 712 kbps is on the or-
`der of 20 dB. As is also shown in Figure 1, the AMC curve be-
`comes smoother when using multiple codes, i.e. multi-code
`operation provides increased granularity of the AMC. Fur-
`ther, multi-code operation enhances the dynamic range of
`AMC by the number of available codes. Hence, the total dy-
`namic range of for instance AMC with 15 multi-codes is on
`the order of 32 dB. If all the code rate resolution available to
`HSDPA is utilized this will also lead to a smoother AMC
`curve than presented in Figure 1, which only includes the
`five example schemes of Table 2.
`
`L1 retransmission techniques
`The HARQ protocol selected for HSDPA is stop and wait
`(SAW). In SAW, the transmitter persists on the transmis-
`sion of the current block until it has been successfully re-
`ceived by the UE. In order to avoid waiting times for ac-
`knowledgements, N parallel SAW-ARQ processes may be
`set for the UE, so that different processes transmit in sepa-
`rate TTIs. The value for N may maximally be 8 but in prac-
`tice, the delaybetween the original and the first retransmis-
`sion is expected to be on the order of 8-12 ms. The control of
`the L1 HARQ is located in the MAC-hs, so that the storage of
`unacknowledged data packets and the following scheduling
`
`800
`
`600
`
`400
`
`200
`
`0
`-15
`
`Single code
`
`TFRC:
`
`1
`
`2 3 4
`
`5
`
`Dynamic range
`
`-10 -5 0 5 10
`
`15
`
`Per-2ms averaged Es/No [dB]
`
`12000
`
`&lO000
`
`Up to 15 codes
`
`8000
`
`6000
`
`4000
`
`2000
`
`0
`-2O
`
`TFRC: 1 2 3 4 5
`
`Dy~am~ range
`
`~
`
`-15 -10 -5 0 5 10 15 20 25
`
`Per-2ms averaged Es/No [dB]
`
`Figure 1 Dynamic range of the HSDPA AMC.
`Simulation assumptions: RAKE receiver,
`ITU Pedestrian-A profile, 3km/h.
`
`N K8681TCO 11070378
`ZTE Corporation and ZTE (USA) Inc.
`Exhibit 1014-00002
`
`
`
`Flat fading, 3 kmph
`
`2nd transmission
`(chase combining) /
`
`~2nd transmission / ~t
`fst tranSm~g
`
`¯
`
`¯
`Combining loss
`
`(incremental redundancy)
`
`0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
`
`Effective code rate
`
`Figure 2 Performance of different retransmission strategies.
`
`of retransmissions does not involve the RNC. Hence, Iub
`signaling is avoided and the resulting retransmission delay
`of HSDPA becomes much lower than for conventional RNC
`retransmissions. The HSDPA retransmission procedure is
`thus several orders of magnitude faster than the convert-
`tional RNC based ARQ implementation and enables the use
`of advanced retransmission strategies with lower delay jit-
`tering and higher spectral efficiency, even for delay sensi-
`tive services like streaming.
`The HSDPA concept supports both the incremental re-
`dundancy (IR) and chase combining (CC) retransmission
`strategies. The basic idea of the CC scheme is to transmit an
`identical version of an erroneously detected data packet and
`then for the decoder to combine the received copies weighted
`by the SNR prior to decoding. With the IR scheme, addi-
`tional redundant information is incrementally transmitted
`if the decoding fails on the first attempt. The performance of
`CC and IR schemes are compared in the curves of Figure 2
`showing the information bit energy to interference ratio
`(Ei/Io) required to obtain a BLER of 30%. Ei/Io values are
`given as a function of the effective Turbo encoding rate. The
`curve labeled "1st transmission" shows the required Ei/Io if
`successful detection is to be accomplished in a single trans-
`mission with a probability of 30%. The curves labeled "2na
`transmission" indicate the required Ei/Io calculated as the
`linear sum of the Ei/Io of the two individual transmissions,
`still at a 30% probability of correctly detecting the packet.
`As can be seen for the case of chase combining, a slight com-
`bining loss must be expected (loss slightly higher if lower
`BLER target is set after second transmission). This loss is
`mainly attributed to the combining operation itself, which is
`basedon the combining of soft information values. As can be
`noticed for a code rate of 3/4, there is a large advantage ofap-
`plying IR since the resulting code rate after the second
`transmission is close to optimum (1/3 which is the base en-
`coder rate). For code rates of ½ or lower, IR does not provide
`a significant gain over chase combining since almost all code
`information has been sent in the first transmission. The dis-
`advantage of IR over CC is the much higher memory re-
`quirements for the UE. The possibility to utilize IR for a cer-
`tain TFRC and multi-code combination is defined by the UE
`capability class. Depending on the data rate compared to
`the UE capability as well as the code rate of the first trans-
`mission, aspects of both the CC and the IR schemes will be
`utilized in the retransmissions. When 16QAM is used as the
`modulation scheme, two of the feur bits constructing the re-
`ceived symbols will have a higher probability of error than
`the other two bits. In order to compensate for this effect it is
`
`possible to use constellation re-arrangement for
`retransmissions, which provides a swapping of the bit
`streams in a way that all bits experience the same average
`level of error probability after the retransmission
`combining.
`Retransmission utilization for a user depends on whether
`the channel quality is generally in the lower or the upper
`end of the AMC dynamic range and if it exceeds this dy-
`namic range. For optimal spectral efficiency and a simple
`round-robin scheduling scheme (without consideration of
`hardware and code utilization issues), users located at the
`cell edge will experience an average first transmission
`BLER around 30-60%, while users located in the vicinity of
`the Node-B will operate with a first transmission BLER
`around 10-20%. The reason for the higher BLER at the cell
`edge is that a user in bad conditions will more often be in a
`condition where even the most robust TFRC cannot be re-
`ceived without error in the ist transmission.
`
`Spectral and code efficiency
`Before reaching the pole capacity, a synchronous WCDMA
`system may be capacity limited due to either a power short-
`age or a code shortage. One of the major benefits of the
`HSDPA concept is the ability to make a tradeoff among
`power and code efficiency to accommodate the current state
`of the cell. This aspect is illustrated in Figure 3, where the
`five example TFRCs are plotted in a diagram showing both
`their power efficiency (measured as allowed noise power to
`user bit energy ratio for a BLER of 10%, e.g. Io/Ei) and their
`respective code efficiency (measured as supported data rate
`per code). If the Node-B has relatively more power resources
`than code resources available (code limited), the link adap-
`tation algorithm will optimize for a more code efficient
`TFRC while a more robust TFRC with more multi-codes will
`be used when the Node-B is mainly power limited.
`
`Link adaptation and support channels
`The overall concept of the HS-DSCH link adaptation (LA) is
`illustrated in Figure 4. The Node-B tracks the radio channel
`quality in the downlink direction by monitoring the trans-
`mit power on the downlink associated DCH (adjusted via
`control commands available on the uplink associated DCH).
`The UE can also be requested to regularly send a specific
`channel quality indicator (CQI) on the uplink high speed
`
`TFRC1
`........... -’~::’~.ooi ........................ ,. ...............................................
`"i’.. TFRC2 i
`
`0.6
`
`0.4
`
`0.2
`
`o
`0
`
`IncreaDed code e~’ciency
`
`600
`400
`200
`Peak throughput per SF = 16 code (kbps)
`
`800
`
`Figure 3 Power and code effciency for different TFRCs.
`
`IEEE Vehicular Technology Society News ÷=÷ February 2003
`
`N K8681TCO 11070379
`ZTE Corporation and ZTE (USA) Inc.
`Exhibit 1014-00003
`
`
`
`(a)
`
`Channel quality feedback (HS-DPCCH, DCH)
`,..,-.-.’,’,’, :~; :-: ;.: .~ ...............
`
`¯ . (HS-DSCH,
`
`UE1
`
`Fast adaptation is conducted in Node B ~
`based on reported channel quality, QoS
`demands, resource availability, etc.
`
`(b)
`
`Downlink
`
`HS-SCCH
`
`HS-DSCH ~
`
`!Verification time
`
`Uplink i
`
`Ass. DCH
`
`Figure 4 The general HSDPA operating principle is shown
`in (a) and inter-channel operation is illustrated in (b).
`
`dedicated physical control channel (HS-DPCCH). The CQI
`is an indicator of the TFRC and multi-code number cur-
`rently supported by the UE (e.g. the supported data rate).
`The feedback cycle of the HS-DPCCH CQI can be set as a
`network parameter in predefined steps from 2 ms to infinite
`time (i.e. disabled). The power of the HS-DPCCH is set as an
`offset compared to the uplink dedicated physical control
`channel (DPCCH) and to guarantee full cell-coverage a CQI
`repetition scheme can be used. For the Node-B to know ira
`transmitted packet was detected correctly or wrongly in the
`receiver end, the UE is required to send a CRC-based
`ACK/NACK response on the HS-DPCCH. It is up to the
`Node-B (and thus the manufacturer/operator) to decide
`whether it will base its link adaptation strategy primarily
`on the associated DCH power control commands, the
`HS-DPCCH information, or a cembination of the two.
`Depending on packet prioritization and resource avail-
`ability, the Node-B then schedules data to the users on the
`HS-DSCH. In this sense, two users may be both time and/or
`code multiplexed to better utilize the available resources
`under the constraint of having different UE capability
`classes. Prior to sending data on the HS-DSCH, the Node-B
`sends a detailed demodulation message to the active users
`via the high speed shared control channel (HS-SCCH). This
`information describes the employed TFRC, the multi-code
`set, as well as the H-ARQ process control and is transmitted
`2 slots in advance of the HS-DSCH. The UE being active on
`the HS-DSCH musL be capable of receiving up to four paral-
`lel HS-SCCHs in order to determine if data is being trans-
`mitted to the UE in the following time period. Masking the
`CRC field on the HS-SCCH with a unique UE ID facilitates
`the UE identification process. The power of the HS-SCCH is
`controlled by the Node-B and it may have a time-variant
`
`Febru~ury 2003 ÷:÷ IEEE Vehicul~ur Technology Society News
`
`power. The HS-DSCH contains user data as well as a single
`24-bit CRC field that is calculated over all the transmitted
`multi-codes (i.e. one logical transport channel maps into
`several physical channels). This approach yields a Turbo
`coding gain for larger data rates where the encoding block
`size increases.
`
`UE Capabilities
`With the introduction of the HSDPA concept into the Re-
`lease 5 specifications, a new generation of UE capability
`classes will be introduced. Five main parameters are used to
`define the physical layer UE capability level [5]:
`÷ Maximum number of HS-DSCH multi-codes that the
`UE can simultaneously receive. At least five multi-codes
`must be supported in order to facilitate efficient
`multi-code operation.
`÷ Minimum inter-TTI interval, which defines the dis-
`tance from the beginning of a TTI to the beginning of
`the next TTI that can be assigned to the same UE. E.g. if
`the allowed interval is 2 ms, this means that the UE can
`receive HS-DSCH packets every 2 ms.
`÷ Maximum number of HS-D SCH transport channel bits
`that can be received within a single TTI.
`÷ The maximum number of soft channel bits over all the
`HARQ processes.
`÷ If the LIE supports 16QAM (e.g. code efficiency limitation).
`Further, parameters are specified for informing the net-
`work what is the total L2 buffer capability (MAC and RLC) in
`the UE. Examples of UE capability classes proposed in 3GPP
`are listed in Table 3, but more combinations are possible [5].
`Note that a ’low-end’ HSDPA UE will support a maximum of 5
`simultaneous HS-DSCH codes, and the minimum distance
`between the starting points of two successive data packets is
`3 TTIs (i.e. 6 ms). Such a UE will support a maximum of 7300
`bits in each TTI and thus belong to the 1.2 Mbps class. An-
`other important difference is the amount of soft channel bits
`defined for each UE class. The number of soft channel bits
`will impact the UE receiver performance when HARQ is em-
`ployed. A UE with a low number of soft channel bits will not
`be able to support IR for the highest peak data rates and its
`performance will thus be slightly lower than for a UE sup-
`porting a larger number of soft channel bits.
`
`Architecture Issues and RRM
`It is entirely up to the network operator to choose a policy for
`weighting the different offered services, the subscription
`
`Reference combination
`
`1.2 3/Pops 3.6 3/Pops 7 Mbps 10 3/Pops
`class
`class
`class
`class
`
`RLC and MAC-hs parameters
`
`Total RLC AM and MAC-hs
`buffer size (kbytes)
`
`Maximum number of AM RLC
`entities
`
`PHY parameters
`
`FDD HS-DSCH category
`
`Maximum number of bits of
`HS-DSCH codes received
`
`50
`
`50
`
`100
`
`150
`
`6
`
`1
`
`5
`
`6
`
`5
`
`5
`
`8
`
`8
`
`7
`
`10
`
`9
`
`15
`
`Minimum inter-TTI interval
`
`3 (6 ms)
`
`1 (2 ms)
`
`1 (2 ms) 1 (2 ms)
`
`Maximum number of bits ~f an
`HS-DSCH transport block
`received within an HS-DSCH TTI
`
`7300
`
`7300
`
`14600
`
`20432
`
`Total number of soft
`channel bits
`Table 3 Example of HS-DSCH UE capability classes [5]. All
`example catagories support 16QAM.
`
`115200 172800
`
`19200
`
`57600
`
`N K8681TC011070380
`ZTE Corporation and ZTE (USA) Inc.
`Exhibit 1014-00004
`
`
`
`classes, etc. This weighting applies to both the priority in
`the network among other services/users but also to the ser-
`vice quality targets; e.g. in terms of guaranteed data rates,
`minimum delays, etc. One exan~ple is to use a paradigm of
`having different QoS classes; e.g. the premium, gold, and
`silver subscription ~livision [6]. The RRM algorithms are re-
`sponsible for best utilization of the available system re-
`sources to meet the service policies adopted by the network
`provider as well as maximizing the system capacity. Specifi-
`cally, the RRM algorithms are responsible for implementing
`a hardware and power resource sharing between the differ-
`ent channel types, so that a Node-B can convey traffic over
`e.g. the DCHs, DSCH, and HS-DSCH at the same time. The
`admission control (AC) functionality is critical to obtaining
`the best tradeoff among capacity and single-service quality.
`Guaranteeing a negotiated QoS level calls for an efficient
`and QoS-aware AC mechanism, which adjusts its user ad-
`mission criterion according to the service requirements of
`the currently active users as well as the pending new user.
`The cell specific power and code resource allocation
`among different channel types is negotiated in the RNC on a
`rather slow basis compared to the TTI for the HS-DSCH (us-
`ing cell reconfiguration messages). The power and code re-
`sources reserved for HS-DSCH and HS-SCCH are subse-
`quently reported to the Node-B (MAC-hs) over the open Iub
`interface [7]. If no power constraints are specified, the
`Node-B can use all excess power not needed for other traffic
`for the HS-DSCH. The number of channelization codes for
`HS-DSCH (SF--16) and HS-SCCH (SF=128) are explicitly
`dictated by the RNC, while the MAC-hs has the freedom to
`freely distribute the power between the user data and the
`control channels. The packet scheduling (PS) is responsible
`for conducting the scheduling of the users, which have been
`directed to the HS-DSCH. The complicated scheduling oper-
`ation must consider UE capability issues (e.g. use of code
`multiplexing), QoS requirements and priority, pending
`re-transmissions, user’s current channel quality, etc. Subse-
`quently, the link adaptation process and the SAW channel
`selection process are conducted. As the HS-DSCH offers
`per-TTI bit rate modification and time/code multiplexing
`between different users, the MAC-hs, containing the
`HSDFA PS, link adaptation, and HARQ entities, has been
`moved to the Node-B. This is illastrated in Figure 5.
`
`Packet scheduling
`The high scheduling rate combined with the large AMC dy-
`namic range available with the HSDPA concept, makes it
`possible to conduct the packet scheduling according to the
`radio conditions as well as the data amounts to be transmit-
`ted to the different users. Hence, the HSDPA concept opens
`for Waterfilling based packet scheduling strategies for opti-
`
`RNC __
`lub
`
`CQI, ACK/NACK/TPC~
`
`~ ~
`
`Figure 5 HSDPA RRM entities in the Node-B.
`
`mized cell throughput!fairness strategies, see e.g. [8,9]. The
`PS methodologies can generally be characterized by:
`Scheduling period/frequency: The period over which
`users are scheduled ahead in time. If short, the PS may uti-
`lize fast channel variations and track fast fading for low-mo-
`bility users. Shorter periods call for higher computational
`complexity in the Node-B.
`Serve order: The order in which users are served; e.g. ran-
`dom order (round robin) or according to channel quality (C/I or
`throughput based). More advanced order mechanisms require
`higher computational processing at the Node-B.
`Allocation method: The criterion for allocating re-
`sources; e.g. same data amount or same power/code/time re-
`sources for all queued users per allocation interval.
`Some general packet scheduling methods and their char-
`acteristics have been compared in Table 4. The fair through-
`put (FT) scheduler serves users in a random order and ac-
`cording to the same data amount. In theory, all users
`currently active in the system will therefore experience the
`same delay and throughput. With the fair resource (FR)
`scheduler, users receive equal resources in random order
`and will thus experience different data rates according to
`their average channel quality. With the C/I PS method (also
`denoted the thoughput or TP method), the user with the
`best channel quality is served until the queue is emptied.
`This leads to a very different service experience among us-
`ers and to the potential situation where a certain poor-qual-
`ity user will experience excessive service delays. The sched-
`uling rate for these packet schedulers is assumed to be slow
`such that fast channel variations are not incorporated (av-
`eraging still may be faster than shadowing variations,
`though). An available option with the HSDPA concept is to
`make very fast scheduling, which tracks the fast fading
`variations. Ultimately, users are only scheduled when they
`are experiencing constructive fading; thereby improving
`both the user throughput and cell throughput for
`time-shared channels. The Max C/I or throughput (M-TP)
`method is the most drastic method, which only serves the
`best user during the current TTI; e.g. the user who can sus-
`tain the highest throughput. Compared to the TP scheduler,
`this scheduler is fairer to the users since a single user’s fad-
`ing variations typically exceed or are on the order of the av-
`erage C/I difference between different user locations in the
`cell. However, the outage of this method is still significant.
`To obtain a fairer scheduling method, it is possible to define
`and calculate a relative instantaneous channel quality
`(RICQ) measure as a selection and prioritization metric.
`The RICQ measure is often identical to the ratio of the user’s
`instantaneous throughput and the user’s average served
`throughput [9]. In calculation, it utilizes the CQI informa-
`tion as well as the link quality estimation algorithms, which
`are located in the Node-B. This fast scheduling method is re-
`ferred to as the proportional fair resource (P-FR) scheduling
`method as illustrated in Table 4. The proportional schedul-
`ing method results in all users getting approximately an
`equal probability of becoming active even though they may
`experience very different average channel quality.
`The above-mentioned schedulers are "prototype" packet
`schedulers, which use different means to utilize and distrib-
`ute excess capacity of the network. They basically yield a
`very different tradeoffbetween user fairness and cell capac-
`ity. Prioritization based on either QoS constraints or differ-
`ent subscription classes (e.g. premium, gold, and silver us-
`ers) will in general override the scheduling principles
`depicted in Table 4 and the scheduling will then only be ap-
`plied to groups of users/services encompassing the highest
`
`IEEE Vehicular Technology Society News ÷~*÷ February 2003
`
`N K8681TCO 11070381
`ZTE Corporation and ZTE (USA) Inc.
`Exhibit 1014-00005
`
`
`
`PS method
`
`Scheduling rate
`
`Serve order
`
`Allocation method
`
`Fair throughput (FT)
`
`Slow
`
`Round robin in random order
`
`Round robin in random order
`
`Resources according to same data amount (up to
`max. allocation time)
`
`Same resources (time, code, or power) and uneven
`data amount
`
`Fair time (FR)
`
`C/I or throughput
`(TP)
`
`Proportional fair
`resource (P-FR)
`
`Max C/I or
`throughput (M-TP)
`
`Slow
`
`Slow
`
`Fa~t
`
`Fast
`
`Based on highest average C/I (fast enough
`to track shadowing)
`
`Same resources (time, code, or power) and uneven
`data amount
`
`Ba~ed on highest relative in~tantaneou~
`channel quality (tracks fast fading)
`
`Same resource~ (time, code, or power) and uneven
`data amount
`
`Based on highest instantaneous channel
`quality (tracks fast fading)
`
`Same resources (time, code, or power) and uneven
`data amount
`
`Table 4 Comparison of different simpfified packet scheduling methods [9, 11].
`
`600
`
`400
`
`200
`
`o
`
`600
`
`400
`
`200
`
`0
`0 2 4 6 8
`
`Power allocated per code (out of 20 W) [W]
`
`Figure 6 Code throughput versus code power allocation.
`
`been evaluated under the assumptions listed in Table 4. Only
`one user prioritization class is considered and the packet sched-
`uler operation is not limited by QoS constraints. As TCP and
`other higher layer protocols are not considered in the evalua-
`tion, we attempt a "best effort" type simulation assuming no
`degradation from e.g. slow start effects. The average cell capac-
`ity for the different packet scheduling methodologies are com-
`pared in Figure 7 also including reference numbers for Release
`
`Re199 WCDMA capability [12]
`I~ Fair resource (FR)
`
`~ Proportional fair resource (P-FR)
`
`priority level. When QoS requirements dominate the sched-
`uling strategy, the differentiation between different PS
`strategies becomes less significant and the capacity gain of
`the most aggressive schedulers reduces (while becoming
`more fair).
`Performance
`The performance of the HS-DSCH depends on a large number
`of aspects, such as (i) channel conditions including othercetl
`interference and time dispersion, (ii) UE demodulation perfor-
`mance and capability, (iii) nature and accuracy of RRM algo-
`rithms, and (iv) hardware imperfections. The throughput per-
`formance for a single link employing link adaptation is shown
`for different channel profiles and average Iorfioe values in Fig-
`ure 6 versus the code power allocation. In the estimation of the
`UE channel quality (Es/No) at the Node-B some error must be
`expected. In these simulations, a lognormally distributed er-
`ror with a standard deviation of 1 dB a