Non-Orthogonal Multiple Access for Vehicular

Context Design of NOMA and SIC Software Defined Radio Conclusion and Perspectives

Non-Orthogonal Multiple Access for Vehicular Networks based Software-Defined Radio Rafik Zitouni ‡ ?

‡ ?

and Samir Tohme

VEDECOM Institute, Versailles, France SIC laboratory-ECE Paris, Paris, France [email protected]

Limassol, Cyprus June 28, 2018

1 / 27

Rafik Zitouni

IWCMC2018

‡


About VEDECOM Mobility Program V2X communications, IoT, Smart cities, Big data etc Autonomous car level 4/5

Vehicular Program Electrification, Inductive battery charging, etc

2 / 27

Rafik Zitouni

IWCMC2018


Summary

1

Context IEEE 802.11p standard NOMA and SIC Software Defined Radio: GNU Radio flow graphs

2

Design of NOMA and SIC Software Defined Radio Flow graphs Simulations Results

3

Conclusion and Perspectives

3 / 27

Rafik Zitouni

IWCMC2018


IEEE 802.11p standard NOMA and SIC Software Defined Radio: GNU Radio flow graphs

Context Wireless Vehicular Ad-hoc NETworks (VANETs) allow cars to exchange information about their perception of the surrounding environment in order to avoid emergency situations The direct V2V communication is required through the de-facto standard IEEE 802.11p for VANETs.

Limited radio channel capacity

4 / 27

Rafik Zitouni

IWCMC2018



IEEE 802.11p standard The standard defines CSMA/CA and Orthogonal Frequency Division Multiplexing (OFDM) for MAC and PHY layer, respectively IEEE 802.11a/g (WiFi) BPSK, QPSK, 16QAM, 64QAM 6; 9; 12; 18; 24; 36; 48; 54 2.4 GHz and 5 GHz 20

Number of subcarriers Symbol duration

IEEE 802.11p (ITS-G5) BPSK, QPSK, 16QAM, 64QAM 3; 3.5; 6; 9; 12; 18; 24; 27 5.9 GHz 5, 10 and 20 5 channels for ITS-G5 and 7 channels for DSRC 52 (48 for data and 4 pilots) 8 µs

Exchanged packets

CAM, DENM, etc

Beacons, DATA, ACK, etc

Modulations Data rate (Mb/s) Frequency band Channel bandwidth (MHz) Channels

14 channels, 189 channels 52 (48 for data and 4 pilots) 4 µs

How to increase the throughput or channel capacity of the network without changing standard’s specifications? How to virtualize the IEEE 802.11p radio devices for high reconfigurability? 5 / 27

Rafik Zitouni

IWCMC2018



IEEE 802.11p standard Brief review of OFDM LTE, DVB, WiFi, DAB etc Complex modulations symbols (BPSK, QPSK, 16QAM, etc) Subcarriers: Discrete frequencies on which data are transmitted Pilot symbols: Special symbols known a-priori

6 / 27

Rafik Zitouni

IWCMC2018



NOMA and SIC Non Orthogonal Multiple Access (NOMA) is a 5G radio access technology candidate promising the frequency reuse

x=

√

p1 s1 +

√

p2 s2

p1 , p2 are output powers of superoposed signals s1 and s2 , respectively

We superpose two signals with NOMA, and we extract them through Successive Interference Cancellation (SIC) Downlink: One vehicle superposes two signals and receivers perform SIC Uplink: Two vehicles send signals on the same frequency, and Road Side Unit (RSU) extracts them via SIC 7 / 27

Rafik Zitouni

IWCMC2018



NOMA and SIC Downlink NOMA is able to double the channel capacity (b/s/Hz) of OMA for two users (vehicles) OMA

NOMA

 p1 |h1 |2   ) Cu1 = β log2 (1 + βN0,1 p2 |h2 |2   Cu2 = (1 − β) log2 (1 + ) (1 − β)N0,2

 p1 |h1 |2   ) Cu1 = log2 (1 + N0,1 2 p |h  2 2|  Cu2 = log2 (1 + ) p1 |h2 |2 + N0,2

12 OMA NOMA

10

β channel bandwidth in (Hz) pi output power of signal si N0,i power spectrum density or Gaussian noise on signal si hi complex channel coefficient between vehicles i

Capacity (bit/s/Hz)

Cui channel capacity of user i 8

6

4

2

0 2

4

6

8

10

12

SNR(dB)

8 / 27

Rafik Zitouni

IWCMC2018

14

16

18

20



NOMA and SIC

Successive Interference Cancellation 1

Symbol level SIC receiver

2

Codeword-level SIC s1+s2+noise Decode s2

-

s2

Encode s2

s1+noise

Decode s1

s1

Re-encoding technique would increase the probability of successful recovery of signal at closest receiver

9 / 27

Rafik Zitouni

IWCMC2018



NOMA and SIC If the transmitter superposes two Quadrature Phase Shift Keying (QPSK) for each subcarrier, the receiver decodes 16 Quadrature Amplitude Modulation (16QAM)

=

+

QPSK

QPSK

16QAM

α is the ratio between the two output powers p1 and p2 is where 1 α = p1p+p , α ∈ [0, 1] 2 We implemented a proof of this concept on Software Defined Radio (SDR) 10 / 27

Rafik Zitouni

IWCMC2018



Software Defined Radio: GNU Radio flow graphs Software Defined Radio (SDR) offers reconfigurable solutions Antenna Front End Intermediate Frequency(IF)

ADC

DAC

Software

Baseband Processing

SDR is more reconfigurable compared to hardware transceivers GNU Radio toolkit allows designers to prototype an SDR of IEEE 802.11p PHY 1 We proposed to modify this prototype and introduce NOMA and SIC SDR 1 Bloessl, B., Segata, M., Sommer, C., and Dressler, F. (2017). Performance Assessment of IEEE 802.11p with an Open Source SDR-based Prototype. IEEE Transactions on Mobile Computing, pages 1–1 11 / 27

Rafik Zitouni

IWCMC2018



Software Defined Radio: GNU Radio flow graphs Flow graphs or software radio programs are the result of connecting signal and data processing blocks Developed in C++ and glued through Python scripts

Data or samples are buffered between software processing blocks of a flow graph Example of a flow graph or a Flowgraph Additional latency Write

Read

Processing Block 1

Write

Read Processing Block 3

Processing Block 2 Data Buffer

12 / 27

Additional latency

Data Buffer

Rafik Zitouni

IWCMC2018



Software Defined Radio: GNU Radio flow graphs 802.11p Tx flow graph

2

2 Bloessl, B., Segata, M., Sommer, C., and Dressler, F. (2017). Performance Assessment of IEEE 802.11p with an Open Source SDR-based Prototype. IEEE Transactions on Mobile Computing, pages 1–1 13 / 27

Rafik Zitouni

IWCMC2018



Software Defined Radio: GNU Radio flow graphs 802.11p Rx flow graph

14 / 27

Rafik Zitouni

IWCMC2018



Software Defined Radio: GNU Radio flow graphs Interoperabilty between OBU-201 Autotalks transceiver and USRP B210 driven by GNU Radio receiver

15 / 27

Rafik Zitouni

IWCMC2018


Flow graphs Simulations Results

Design of NOMA and SIC Software Defined Radio Objective Implement NOMA at the transmitter and SIC at the receiver for each subcarrier as in NOMA cellular downlink NOMA transmitter Message Source

MAC

User 1

Addresses

Mapper Channel Encoder

BPSK, QPSK

SIGNAL Field

OFDM Carrier

Encoding

Allocator

Message Source

MAC

User 2

Addresses

Mapper Channel Encoder

Ampli er

BPSK, QPSK

We use two QPSK subcarriers in our simulations 16 / 27

Rafik Zitouni

IWCMC2018

IFFT

OFDM Cyclic Pre xer

Packet Pad



Design of NOMA and SIC Software Defined Radio SIC receiver Autocorrelation

Frame

Frequency

Symbol

Stream to

Detection

Correction

Alignment

Vector

Equalize

SIGNAL Field

Constellation

Symbols

Frame

Symbols

Decoding

Decoder

User 1

Decoding

FFT

Codeword-level SIC Equalize

Signal Field

Constellation

Symbols

Frame

Symbols

Decoding

Decoder

User 1

Decoding

Constellation

Ampli er

Encoder SUB

17 / 27

Rafik Zitouni

IWCMC2018

Symbols

Frame

User 2

Decoding



Design of NOMA and SIC Software Defined Radio

SIC function is a computationally complex due to the re-encoding operation Through simulation, we can calculate the Bit Error Rate (BER) rather than frame error, and we avoid all complex issues of the receiver synchronization Simulate NOMA and SIC with loopback flow graphs on one subcarrier and calculate BER We consider the case of platooning since the distance is fixed and the output power could be estimated

18 / 27

Rafik Zitouni

IWCMC2018



Simulations

Parameters of loopbak flowgraphs OS and SDR CPU

Ubuntu 14.04.5 LTS, GNU Radio 2.7.9 Intel Core i5 at 2.4 GHz, 6 family

NOMA superposition Amplifier p1 , p2

Two QPSK 1 α = p p+p , α ∈ [0, 1]

Channel model

19 / 27

1

2

−

Gaussian Noise PN = 10

Rafik Zitouni

SNR 10 , SNR(dB) ∈ [0, 20]

IWCMC2018



Design of NOMA and SIC Software Defined Radio GNU Radio Flow graphs NOMA Tx

Rx without SIC Rx with SIC

20 / 27

Rafik Zitouni

IWCMC2018



Design of NOMA and SIC Software Defined Radio Superposition of two users

Symbol constellations at SIC receiver

21 / 27

Rafik Zitouni

IWCMC2018



Design of NOMA and SIC Software Defined Radio Receiver with and without SIC

22 / 27

Rafik Zitouni

IWCMC2018



Results BER function of α for user 1 with SIC function NOMA/SIC with an SNR=10dB 1

User 1 BER without SIC User 1 BER with SIC

Bit Error Rate (BER)

0.8

0.6

0.4

0.2

0 0

0.2

0.4

0.6 α amplifier P1/P

23 / 27

Rafik Zitouni

IWCMC2018

0.8

1



Results BER function of α for two users with SIC NOMA/SIC with an SNR=10dB 1

User 1 BER with SIC User 2 BER with SIC

Bit Error Rate (BER)

0.8

0.6

0.4

0.2

0 0

0.2

0.4

0.6 α amplifier P1/P

24 / 27

Rafik Zitouni

IWCMC2018

0.8

1


Conclusion

We proposed a design of NOMA and SIC SDR for IEEE 802.11p without changing the standard NOMA and SIC have been found useful to increase the channel capacity or spectrum efficiency We illustrated the user fairness through a calculated BER versus the ration (α) between the output powers of superposed signals. For two users: The capcity of channel would be improved and BER decreased by up to 20% The SIC receiver gets BER ' 0 when α = 0.9 or α = 0.1

25 / 27

Rafik Zitouni

IWCMC2018


Perspectives

What is the impact of mobility and scalability on the channel capacity? Considerate the processing optimization of the SIC computational complexity on SDR

26 / 27

Rafik Zitouni

IWCMC2018


Thank you for your attention

27 / 27

Rafik Zitouni

IWCMC2018