CC BY
111
LTE/LTE-Advanced PHY Layer Coding Rate Based Performance Verification
Keywords: LTE, LTE-Advanced, comprehensive LTE UE receiver, 3GPP, PHY Layer coding.
Hiang Chuan, Tan,
Keysight Wireless Solution Architect, mobile devices,
Introduction
This paper addresses the measurement needs and test challenges for LTE and LTE-Advanced mobile devices. In particular, it focuses on verifying user equipment (UE) receiver and transmitter implementation and characterizing its performance which is not fully specified in 3GPP test requirements but is expected to be in compliance. 3GPP defines the detailed implementation of the LTE radio technologies and derived UE test specifications that must be met universally by chipset providers and mobile device providers. Protocol implementation, radio resource management and radio conformance test are governed by certification labs to ensure the UE released to the network will perform according to acceptable minimum requirements. However, large portion of implementations of the chipset resource scheduling mathematically specified and implemented changes according to numerous varying factors that are complex and time consuming to be fully tested. Characterizing of receiver performance is generally covered in 3GPP test requirements, with limited pre-defined emulation conditions. This paper recommends comprehensive LTE UE receiver and transmitter performance verification with varying code rate to allow robust verification of the receiver implementation for operation on the real worid.
ceplible lo narrowband interference. With Orthogonal Frequency Division Multiple Access (OFDMA), subcarriers can be allocated dynamically among different users of the channel. OF-DMA also allows non-contiguous allocation of subcarriers for a single user.
Mobile devices, commonly known as UE supporting LTE are becoming common in recent years as network providers race to provide faster data rale to attract ever demanding consumers. Increasing demand to take advantage of these fast networks drove the market adoption of high-end mobile devices with embedded LTE chipsets.
Driven by the ever-increasing hunger for throughput and limitation of carrier frequency spectrum governed by each specific country; the push to aggregate multiple frequencies within and/or across bands with different combinations of allocated bandwidth led to greater demands for power efficiency, sensitivity, spectral purity and process scheduling of the wireless chipset.
The following sections describe high-level LTE radio technology fundamentals required to help understand the importance of verifying the receiver and transmitter resource allocation based oil coding rate which also indirectly, helps verifying all major factors in the UE receiver and transmitter implementation.
LTE radio bandwidth
Initial LTE wireless technology defined by 3GPP introduced a set of frequency bandwidth that allows scalable deployment of LTE with a range of transmission bandwidth. It includes 1.4MHz, 3MHz, 5MHz, 10MHz, 15MHz and 20MHz frequency domain bandwidth which allows scalability according to the radio spectrum availability. With the recent introduction of release 10 3GPP specifications, carrier aggregation between combinations of these transmission bandwidths were made available up to 100MHz aggregated bandwidth with 5 component carriers (CC) to achieve peak data rate of IGbps Although release 10 defined the signaling for up to 5 CC, test requirements are only defined for up to 2 CC so far.
Downlink / uplink modulation
The LTE technology uses Orthogonal Frequency Division Multiplexing (OFDM) modulation with 15 kHz subcarriers spacing on the downlink, 7,5 kHz subcarricrs spacing is also defined in the LTE standards for Multimedia Broadcast Multicast Service Single Frequency Network (MBSFN). With standard OFDM, subcarrier allocations are fixed for each user and are more sus-
Symbols (Time)
Symbols (Time)
Subcarriers
OFDM
OFDM A
□
User 1
User 2
□
User 3
Figure 1. OFDM versus OFDMA allocation
The LTE uplink however uses Single-Carrier Frequency Division Multiple Access (SC-FDMA) modulation to reduce Peak-to-Avcrage Power Ratio and only the subcarrier spacing of 15kHz is used. The detail behind the using SC-FDMA is beyond the scope of this paper.
Downlink resources
A resource element (RE) is the smallest unit in the physical layer that occupies one OFDM symbol in lime domain and one subcarrier in the frequency domain. A physical resource block (RB) consists of A^t, x W™ resource elements, corresponding to
one slot in time domain and x Af in frequency domain. Table 1 below shows the values for the downlink resource parameters.
A normal cyclic prefix configuration has 84 resource elements per slot; in FDD single 10 ms frame (10 subframes or 20 slots) structure has total of 1680 RE/RB. However, not all RE is allocated for physical downlink shared channel (PDSCH). Some RE needs to be allocated to carry control information including physical downlink control channel (PDCCH), physical control format indicator channel (PCFICH) and physical hybrid auto-
Reference signals
3GPP TS 36.211 [1] defined six types of reference signals listed in table 3 below. Reference signals ill general arc used for channel estimation and equalization. The only mandatory downlink reference signal is the CRS.
Table 3
Types of downlink reference signal
Reference signal 3GPP CP and Af
Cell specific (CRS) Rel.8 Normal CP and Extended CP with àf = 15 kHz only
MBSFN RS Rel. 8 Extended CP with Af = 7.5 kHz only
UE Specific (DM-RS) Rel.8 Normal CP and Extended CP
Positioning Reference Signal (PRS) Rel. 9 Normal CP and Extended CP with - kHz only
CSI-RS Rel. 10 Normal CP and Extended CP with A/ = 15 kHz only
However, the number of RE allocated for the RS varies depending on the number of antennas used in the transmission. Figure 4 below shows that 8 REs are allocated for each sub frame for cach antenna when two antennas are used for transmission. Notice that there are 8 REs reserved (L'Not Used") for the first antenna where those 8 REs are used on the same location of the Second antenna and vice versa.
K i1 A
i * *
> .h m Lt.. it
Table 4
DM-RS for PUCCH symbols per slot
PUCCH format Normal cyclic prefix Extended cyclic prefix
1,1a, 1b 3 2
2. 3 2 1
2a. 2b 2 N/A
Figure 4. Downlink CRS mapping in normal CP with two antennas [11
Two uplink reference signals are also defined. The Demodulation Reference Signal (DM-RS) associated to PUSCH and PUCCH and the optional Sounding Reference Signal (SRS). The DM-RS for PUSCH takes up 12 RE/RB on the 4lh symbol of normal CP and on the 3rit symbol of extended CP on each slot. The number of RE/RB occupied by DM-RS for PUCCH depends on the PUCCH formal as shown in the Table 4 below and is defined in 3GPPTS 36.21 i Section 5.5.2.2 [1].
The uplink SRS will occupy 12 RE/RB on the last symbol of each subframe as shown in the Figure 3 above. 12 RE/RB are allocated but the transmission of SRS signal may not occupy all RE/RB.
Downlink transmission mode
There are 9 transmission mode defined as of 3GPP Release 10 as shown in the table 5 below.
Table 5
Transmission Modes {2| ¡4]
TM 3GPP Transmission scheme Description of usage
1 Rel. 8 Single-antenna port Basic RXD
2 Rel.8 Transmit diversity For low SNR condition
3 Rel.8 Transmit diversity Fallback mode
Large delay CDD Increase throughput
4 Rel. 8 Transmit diversity Fallback mode
Closed-loop spatial multiplexing Improve throughput
5 Rel. 8 Transmit diversity Fallback mode
Multi-user MIMO Improve spectral efficiency
6 Rel. 8 Transmit diversity Fallback mode
Closed-loop spatial multiplexing using a single transmission layer (beamsteering) Improve signal robustness
7 Rel. 8 Single-antenna port or Transmit diversity Fallback mode
Single-antenna port (beamforming) Improved signal robustness with non-codebook precoding
8 Rel. 9 Single-antenna port or Transmit diversity Fallback mode
Dual layer transmission (beamforming) Same as TM7 with increased throughput
9 Rel. 10 Non-MBSFN subframe: Single-antenna port or Transmit diversity. M8SFN subframe: Single-antenna port Fallback mode
Up to 8 layer transmission (beamforming) Same as TM8 with improved throughput
With these transmission modes definitions, various antenna configurations and various layers of transmission can be configured to achieve the ideal maximum data throughput operating in different channel conditions, depending on the chipset capabilities. For more explanation of MIMO, multiple layers transmission and codeword, refer to Keysight application note "MIMO in LTE Operation and Measurement - Excerpts on LTE Test"
Digital modulation types
Just like for any other radio technology defined in the 3GPP standards, digital modulation schemes are defined for LTE PDSCH and PUSCH transport channel transmission to carry digital data. QPSK, 16QAM and 64QAM modulation types are defined to transmit 2 bits, 4 bits and 6 bits per symbol respectively. The selection of the modulation type depends greatly on the propagation conditions, resource elements available and the channel quality indicator (CQI), which is part of the channel state information (CSi) reporting.
Downlink PDCCII uses QPSK. modulation type to carry control information; while uplink PUCCH uses BPSK and/or QPSK based on the PUCCH format defined in 3GPP TS36.211 section 5.4 to carry control information.
Downlink channel coding rate
The Physical Downlink Shared Channel (PDSCH) transport block size and modulation order map is defined in 3GPP TS 36.213 Section 7.17 [2J as shown in the Table 6 and 7 below.
Table 6
PDSCH MCS to TBS Index \2]
MCS Index Modulation Order TBS Index
^MCT Q„ -^TBS
0 2 0
1 2 1
2 2 2
3 2 3
4 2 4
5 2 5
6 2 6
7 2 7
8 2 8
9 2 9
10 4 9
11 4 10
12 4 11
13 4 12
14 4 13
15 4 14
16 4 15
17 8 15
18 6 16
19 6 17
20 6 18
21 6 19
22 6 20
23 6 21
24 6 22
25 6 23
26 6 24
27 6 25
28 6 26
29 2
30 4 reserved
31 6
TBS index points to the set of tables' matrix (/TBS, jVPRB) defined in 3GPP TS36.213 Section 7.1.7.2. It is used to calculate and determine the coding rate that can be applied to transport block without causing decoding error. The UE may skip decoding a transport block in an initial transmission if the effective channel code rale is higher than 0.93 and so the limit of 0.93 is used throughout this paper.
Qm = bits per symbol
Physical channel bits = NIm x REmscH x Qm
Downlink information bits = TBS(/TBS, A'PRB) + 24 CRC
bits
DL Coding Rate is the ratio of downlink information bits to be transmitted and the available physical channel bits per subframe.
Uplink channel coding rate
The Physical Uplink Shared Channel (PUSCI-I) transport block size and modulation order map are defined in 3GPP TS 36.213 Section 8.6 |2] as shown in the Table 7 below.
Table 7
PDSCH MCS Index to TBS Index [2]
MCS Index Modulation Order TBS Index
fil ■'res
0 2 0
1 2 1
2 2 2
3 2 3
4 2 4
5 2 5
6 2 6
7 2 7
8 2 8
9 2 9
10 2 10
11 4 10
12 4 11
13 4 12
14 4 13
15 4 14
16 4 15
17 4 16
18 4 17
19 4 18
20 4 19
21 6 19
22 6 20
23 6 21
24 6 22
25 6 23
26 6 24
27 6 25
28 6 26
29
30 reserved
31
Similar to the dow nlink, TBS index points to the set of tables matrix (/I BS, A'PRB) defined in 3GPP TS36.213 Section 7.1.7.2 and it is used to calculate and determine the coding rate that can be applied to transport block without causing decoding error.
Qm - bits per symbol
Physical channel bits = NPRB x REPVSCH x Qm
Uplink information bits = TBS(/TBS, ATRB) + 24 CRC bits
UL Coding Rate is the ratio of uplink information bits to be transmitted and the available physical channel bits per subframe.
The uplink Qm will be limited to 4 bit per symbol (16QAM) from IMCS 21 onwards when the UE is not capable of supporting 64QAM transmission on PUSCH. When TT1 bundling is used, Qm will be clipped to 2 bit per symbol (QPSK).
Putting it all together
In order to help to put the pieces together, let's take a practical approach example of FDD frame structure. A downlink FDD frame structure with normal cyclic prefix configuration has 84 REs per slot, and 168 REs in one subframe (2 slots).
DM-RS, PRS and CSI-RS will also affect downlink PDSCH RE availability.
It may be worth noting that the maximum data throughout for the example above is the sum of TBS values for all 10 subframes multiplied by 100. The throughput is therefore (600f349f}+(4392<8)xl00 and equals 3.9232Mbps for each codeword. In this single-user M1MO case, two codewords yield 7.8464Mbps.
The uplink coding rale uses the same computation algorithm except that the allocation of PUSCI-I, PUCCH and SRS are mapped differently on the resource grid as shown on Figure 9 below.
I 1
u
Sublnm* w rthoul PUCCH orsftS
№t 0 Slot 1
FlritEBODC Sutfca-ne
I_I PUKHPMRS
i " ■ i PMdïHDMftî I - I FUlttt
Lait 3fiEV=N5tbTrame
I ^tMidliuMcicvtslfhilftflS]
) PlrtíK I limned
Figure 9. FDD Uplink Single Normal CP Rcsourcc Grid
As you may have observed, each sublrame is governed by the maximum target coding rate of 0.93 which forces resources allocated for different users to have different coding schemes, modulation types and transport block sizes based on various factors including cyclic prefix, transmission mode, antenna configuration, transmission bandwidth, resource block allocation and number of control channel symbols.
In radio conformance test, target coding rates of 1/6, !/5, 1/3, 1/2, 2/3 and 3/4 are used and without subframe 5 allocations. With target coding rate of 3/4 or 0.75, the error correction in the data transmission prevents some implementation issues from surfacing and furthermore subframe 5 data coding is not even tested. Although in real world implementation, the environmental effect hardly allows constant coding rale of 0.93, it is crucial to use highest possible coding rate with ideal channel condition in the design verification to minimize error correction that hides the implementation issue on the physical layer and transport layer.
CSI based scheduling
In the non-ideal real world condition, CSI based scheduling is used to dynamically reduce or increase coding rate using UE channel estimation. The channel condition is the determining factor for the CQI table mapping. It ensures optimal modulation type and transport block size are used to cany the required data across the transmission medium, adapting to the environment assessment information feedback through the CSI report. The data rate may be reduced in poor signal-to-noise ratio environment in accordance with the CQI reported based on UE receiver assessment and performance. Lower code rate allows successful decoding of data transmitted with higher redundancy and also effectively reduces overhead for retransmission.
3GPP TS 36.213 Table 7.2.3-1 [2J shows the CQI table and its target code rate. It is shown in table 9 beiow.
Table 9
CQI Table |2]
COI index modulation ¡ code rate x 1024 | efficiency
0 out of range
1 QPSK 78 0.1523
2 QPSK 120 0.2344
3 QPSK 193 0.3770
4 QPSK 308 0.6016
5 QPSK 449 0.8770
6 QPSK 602 1.1758
7 16QAM 378 1.4766
8 16QAM 490 1.9141
9 16QAM 616 2 4063
10 64QAM 466 2.7305
11 64QAM 567 33223
12 64QAM 666 3.9023
13 64QAM 772 4.5234
14 64QAM 873 5.1152
15 64 QAM 948 5.5547
This table will be offset based on different resource allocations and on UE wideband or narrow band reporting. The algorithm may be optimized by each chipset modem vendor or manufacturer to provide efficient coding rate in dynamic scheduling in the real world.
In order to measure the performance of the dynamic scheduling, Additive White Gaussian Noise (AWGN), Ortogonal Channel Noise Generator (OCNG) and propagation conditions are applied to verify the algorithm implementation.
Implementing code rate based verification
To effectively perform the various combinations of aggregated carriers with different bandwidth allocations, transmission modes and antenna configurations; creating a complex matrix of each possible permutations will allows completeness in verification of various signaling implementation of transport channels and control channels. Verifying the PDCCH, PCFICH, PHICH, PBCH, PDSCH, PUCCH and PUSCII performance under different antenna configurations, transmission modes and bandwidth allocations by allocating desired transport block size and modulation schemes based on maximum target coding rate allowed for a given allocation can be a complex task. Figure 10 below shows an example of flexible RMC configurations provided by the Keysighi E75Î5A UXM wireless test set that helps engineers to graphically setup the resource allocation.
