Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
1/22
Dr. Essam SourourPrince Sattam Bin Abdulaziz UniversityFaculty of Engineering, Wadi Aldawaser
Non‐OrthogonalMultipleAccessasaCandidatefor5G
2/22
• The demand on cellular service and 5G response to it• NOMA and how it works• Advantages of NOMA• Conclusions
PresentationLayout
3/22
CellularMobileCommunications
4/22
• Demand on cellular mobile communications is increasing• 1G, 2G and 3G systems provided voice services 2G deployed in 1990 3G deployed in 2000 4G deployed in 2010
• The 3.5G (HSPA) and the 4G (LTE) focused on data services• 4G LTE provided data rate up to 1 Gbps • Still a dramatic increase in mobile users and wireless devices is anticipated
DemandonCellularMobileCommunications
5/22
• Powerful smartphones, tablet PCs and laptops are popular• Demanding real time video and multimedia capabilities • Number of wireless connected devices increasing• 5G systems are expected to exceed 4G: 1000 times increase in capacity (connected devices/cell) 10 times increase in data rate ( 10 Gbps) 10 times spectral efficiency (data rate per Hz) 25 times cell throughput ( data rate per cell)
• Several candidate technologies are under research
Demandfor5G,Vision2020
6/22
• Reliable connection, anytime, anywhereMachine to machineDevice to deviceInternet of thingsInternet of vehiclesHD real time videoMobile gamingHigh energy efficiencyGreen systemsLower latencyHigher reliabilityWi‐Fi integration
WhatDoes5GLookLike?
7/22
• Dense microcells• Relay stations• Massive number of antennas per base station and devices• Beamforming antenna patterns• Using millimeter waves (frequencies bands above 10 GHz)• Cognitive radio• Spatial modulation• Non‐Orthogonal Multiple Access (NOMA)
HowCan5GDoThat?
8/22
• 2G, 3G and 4G are mainly OMA systems• Users are allocated orthogonal resources Separate time slots, and Separate frequency allocation
• OMA avoids inter‐user interference• Fast growth in mobile devices vs. limited available spectrum• NOMA: new signal design, users share time & frequency
Orthogonalvs.NonOrthogonalMultipleAccess(OMAvs.NOMA)
9/22
• Serve more than one user on the same time and frequency resource
• Higher spectral efficiency (more data rate per Hz)• Benefit from the geographical distribution of users• Better serve cell edge users (users far from the base station)• Some applications need low data rate. A waste of resources to allocate dedicated time and frequency Sensor readings Inter‐vehicle communicationsMachine to machine communications
MotivationsofNOMA
10/22
• Consider two users example• The base station selects two appropriate users to pair Near‐user (strong channel gain) Far‐user (weak channel gain)
• Served on same time and frequency• Tx Power is split between them High power share to Far‐user Low power share to Near‐user
• Typically interference should happen !!
HowNOMAworks(1)?
11/22
NOMASuperimposedSignal,QPSKexample
Conventional QPSK NOMA QPSK
12/22
• Far‐user signal has small interference from the Near‐user signal Far‐user decodes its signal normally Suffers from slight extra interference
• Near‐user signal has large interference from Far‐user Near‐user decodes Far‐user signal first Subtracts this interference from the composite NOMA signal Hence, Far‐user interference is cancelled Near‐user decodes its data from the cleaned signal
• Note that Far‐user is unable to cancel Near‐user interference because it is too weak to be decoded
HowNOMAworks(2)?
13/22
Far‐UserandNear‐UserProcessing
Far‐User Processing
Near‐User Processing
14/22
• Performance is measured by the Shannon channel capacity
• Shannon capacity (bits per second) gives the achievable data rate In a certain channel with bandwidth W , and Certain signal to noise ratio (SNR)
• Note that the bandwidth has a higher impact on capacity than SNR
IsNOMABeneficial?
2log 1 bits/sC W SNR
15/22
• In OMA each user takes half the bandwidth, but no interference
• OMA: and
• NOMA Near‐user can cancel Far‐user signal:
• NOMA Far‐user can't cancel Near‐user signal:
CapacityComparison
16/22
CapacityComparison
2 2log 1 + log 1 bits/s2 2
near farOMA
P PW WCnoise noise
2 2log 1 + log 1 bits/snear farNOMA
near
P PC W Wnoise P noise
= constantfar near totalP P P
17/22
• The trade‐off between Near‐user and Far‐user is better in NOMA
Capacitycomparison,cont.
Yang Liu, Gaofeng Pan, Hongtao Zhang, and Mei Song, “On the capacity comparison between MIMO‐NOMA and MIMO‐OMA,” IEEE Access, May 2016
18/22
Yuya Saito, Anass Benjebbour, Yoshihisa Kishiyama, and Takehiro Nakamura, “System‐level performance evaluation of downlink non‐orthogonal multiple access (NOMA),” PIMRC 2013
OMAandNOMAComparison
In OMA all users below 4 MbpsIn NOMA some users reach 8 Mbps
19/22
• In 2015 NOMA is accepted as a candidate technology for 5G• In 5G its name is changed from NOMA to MUST (Multi‐User Superposition Transmission)
• Research points being considered: User pairing selection Power sharing ratio Using multiple antennasMore realistic scenarios
NOMAin5GStandardization(MUST)
20/22
• A grant of wireless equipment was received from National Instrument last year for project:
Enhancing Non‐Orthogonal Multiple Access (NOMA) Using Interference Alignment
• Research team: Essam Sourour and Mohammed Al‐Ansi• Goal is to implement NOMA on USRP (wireless modems) using LabVIEW)
• Grant: $50 k for 2 USRP, USRP Controller, master clock, and LabVIEW Training and support The college offered lab room, PC, desk, chairs, tables, power cables, etc.
NationalInstrumentGranttoPSAU
21/22
NationalInstrumentGrant
22/22
• NOMA is based on sharing resources between users• Sharing allows higher sum of data rate• Can help increase number of users and higher data rate• Adopted as a candidate for 5G• Considerable research is going on to put it into practice
Conclusions