Spectrum Collaboration Challenge

Similar documents
The world s first collaborative machine-intelligence competition to overcome spectrum scarcity

Spectrum Collaboration Challenge (SC2)

Ultra Wideband Signal Impact on IEEE802.11b and Bluetooth Performances

Investigation on Multiple Antenna Transmission Techniques in Evolved UTRA. OFDM-Based Radio Access in Downlink. Features of Evolved UTRA and UTRAN

Spectrum Collaboration Challenge (SC2)

Long Term Evolution (LTE) and 5th Generation Mobile Networks (5G) CS-539 Mobile Networks and Computing

Outline / Wireless Networks and Applications Lecture 3: Physical Layer Signals, Modulation, Multiplexing. Cartoon View 1 A Wave of Energy

HY448 Sample Problems

Planning Your Wireless Transportation Infrastructure. Presented By: Jeremy Hiebert

Wireless Point to Point Quick Reference Sheet

Measurement of Digital Transmission Systems Operating under Section March 23, 2005

SPREAD SPECTRUM CHANNEL MEASUREMENT INSTRUMENT

802.11ax introduction and measurement solution

Final draft ETSI EN V1.1.1 ( )

Does The Radio Even Matter? - Transceiver Characterization Testing Framework

2.4GHz & 900MHz UNLICENSED SPECTRUM COMPARISON A WHITE PAPER BY INGENU

5G deployment below 6 GHz

HR001118S0020 Millimeter-Wave Digital Arrays (MIDAS) Frequently Asked Questions (FAQ) March 12, 2018

HR001118S0020 Millimeter-Wave Digital Arrays (MIDAS) Frequently Asked Questions (FAQ) February 12, 2018

Using the epmp Link Budget Tool

Interference management Within 3GPP LTE advanced

Department of Computer Science and Engineering. CSE 3213: Communication Networks (Fall 2015) Instructor: N. Vlajic Date: Dec 13, 2015

Technical Aspects of LTE Part I: OFDM

5G - The multi antenna advantage. Bo Göransson, PhD Expert, Multi antenna systems Systems & Technology

Federal Communications Commission Office of Engineering and Technology Laboratory Division

To Fragment or Not To Fragment: Viability of NC OFDMA in Multihop Networks. Muhammad Nazmul Islam WINLAB, Rutgers University

LTE-Advanced and Release 10

Point to Point PTP500

Realization of Peak Frequency Efficiency of 50 Bit/Second/Hz Using OFDM MIMO Multiplexing with MLD Based Signal Detection

Wireless Digital Nodes

Automatic power/channel management in Wi-Fi networks

Increasing Broadcast Reliability for Vehicular Ad Hoc Networks. Nathan Balon and Jinhua Guo University of Michigan - Dearborn

1 UAT Test Procedure and Report

March, 2003 IEEE P /131r0. IEEE P Wireless Personal Area Networks

A-SAT TM Adaptive Satellite Access Technology John Landovskis

OBJECTIVES. Understand the basic of Wi-MAX standards Know the features, applications and advantages of WiMAX

Reconfigurable 6 GHz Vector Signal Transceiver with I/Q Interface

DYNAMIC BANDWIDTH ALLOCATION IN SCPC-BASED SATELLITE NETWORKS

SIGNAL PROCESSING CHALLENGES IN THE DESIGN OF THE HOMEPLUG AV POWERLINE STANDARD TO ENSURE CO-EXISTENCE WITH HOMEPLUG 1.0.1

Earth-Stations. Performance Requirements

RECOMMENDATION ITU-R M.1167 * Framework for the satellite component of International Mobile Telecommunications-2000 (IMT-2000)

Overview. Cognitive Radio: Definitions. Cognitive Radio. Multidimensional Spectrum Awareness: Radio Space

Chapter- 5. Performance Evaluation of Conventional Handoff

Announcements : Wireless Networks Lecture 3: Physical Layer. Bird s Eye View. Outline. Page 1

CROSS-LAYER DESIGN FOR QoS WIRELESS COMMUNICATIONS

Overcoming Key OTA Test Challenges from 4G to 5G

Smart Antenna Techniques and Their Application to Wireless Ad Hoc Networks. Plenary Talk at: Jack H. Winters. September 13, 2005

Wireless replacement for cables in CAN Network Pros and Cons. by Derek Sum

MATERIAL SPECIFICATIONS FOR WIRELESS LINK

Wireless Communication

APPENDIX B. 4. DEFINITIONS, SYMBOLS AND ABBREVIATIONS For the purposes of the present document, the following terms and definitions apply.

Introduction to Communications Part Two: Physical Layer Ch3: Data & Signals

Department of Computer Science and Engineering. CSE 3213: Computer Networks I (Fall 2009) Instructor: N. Vlajic Date: Dec 11, 2009.

5G: Opportunities and Challenges Kate C.-J. Lin Academia Sinica

Motorola Wireless Broadband Technical Brief OFDM & NLOS

Information Theory at the Extremes

Production Test and Spectral Monitoring

ELEC-E7120 Wireless Systems Weekly Exercise Problems 5

Basic Radio Settings on the WAP371

II. LAB. * Open the LabVIEW program (Start > All Programs > National Instruments > LabVIEW 2012 > LabVIEW 2012)

UTILIZATION OF AN IEEE 1588 TIMING REFERENCE SOURCE IN THE inet RF TRANSCEIVER

3GPP: Evolution of Air Interface and IP Network for IMT-Advanced. Francois COURAU TSG RAN Chairman Alcatel-Lucent

Opportunistic Communication in Wireless Networks

Application Note. LTE Measurement. MT8820C Radio Communication Analyzer

Cellular Infrastructure and Standards while deploying an RDA

Datasheet. Licensed Backhaul Radio. Model: AF-4X. Up to 687 Mbps Real Throughput, Up to 200+ km Range

Vulnerability modelling of ad hoc routing protocols a comparison of OLSR and DSR

AN EDUCATIONAL GUIDE HOW RPMA WORKS A WHITE PAPER BY INGENU

TV White Spaces white space device requirements

Minimum requirements related to technical performance for IMT-2020 radio interface(s)

Federal Communications Commission Office of Engineering and Technology Laboratory Division

Sigfox RF & Protocol Test Plan for RC1-UDL-ENC-MONARCH

License Exempt Spectrum and Advanced Technologies. Marianna Goldhammer Director Strategic Technologies

RECOMMENDATION ITU-R M.1652 *

LTE Aida Botonjić. Aida Botonjić Tieto 1

RF Receiver Hardware Design

Introduction to Surface Acoustic Wave (SAW) Devices

RSU-101E Specifica on

Overcoming Interference is Critical to Success in a Wireless IoT World

Sigfox Verified TM. Modem Test Plan for RC5-UDL-ENC. Version August 10, Public Use

PXI WLAN Measurement Suite Data Sheet

Carrier Aggregation and MU-MIMO: outcomes from SAMURAI project

Simulating Mobile Networks Tools and Models. Joachim Sachs

The Impact of Channel Bonding on n Network Management

IEEE c-23. IEEE Broadband Wireless Access Working Group <

Sigfox RF & Protocol Test Plan for RC2-UDL-ENC

Sigfox Verified TM. Modem Test Plan for RC2-UDL-ENC. Version April 24, Public Use

Project: IEEE P Working Group for Wireless Personal Area Networks N

Understanding and Mitigating the Impact of Interference on Networks. By Gulzar Ahmad Sanjay Bhatt Morteza Kheirkhah Adam Kral Jannik Sundø

SYSTEM SENSOR WIRELESS REMOTE INDICATOR PRODUCT SPECIFICATION

ECE 630: Statistical Communication Theory

TSTE17 System Design, CDIO. General project hints. Behavioral Model. General project hints, cont. Lecture 5. Required documents Modulation, cont.

LTE-Unlicensed. Sreekanth Dama, Dr. Kiran Kuchi, Dr. Abhinav Kumar IIT Hyderabad

Using Variable Coding and Modulation to Increase Remote Sensing Downlink Capacity

MICHIGAN DEPARTMENT OF TRANSPORTATION SPECIAL PROVISION FOR TRAFFIC SIGNAL WIRELESS COMMUNICATIONS LINK

RF exposure impact on 5G rollout A technical overview

DraftETSI EN V1.2.1 ( )

Overview: Trends and Implementation Challenges for Multi-Band/Wideband Communication

Radio Interface and Radio Access Techniques for LTE-Advanced

Sirindhorn International Institute of Technology Thammasat University

Transcription:

Spectrum Collaboration Challenge Frequently Asked Questions (FAQ) November 22, 2017 Defense Advanced Research Projects Agency 675 North Randolph Street Arlington, VA 22203

Revision Summary Section Revision Description Date Q50 14 Made question consistent with Q32 11/22/17 Q32 Clarified criteria necessary to achieve throughput Q48 - Q52 13 Added 11/22/17 Q32 Clarified how throughput is calculated Q47 12 Added 11/15/17 Q45 - Q46 11 Added 11/10/17 Q43 10 Updated to reflect antenna pairs 11/3/17 Q43 - Q44 9 Added 11/2/17 Q39 - Q42 8 Added 10/27/17 Q37 - Q38 7 Added 10/19/17 Q32 - Q36 6 Added 10/19/17 Q27-31 5 Added 10/2/17 Q25 - Q26 4 Added 9/20/17 Q24 3 Added 8/11/17 Q22 - Q23 2 Added 7/26/17 Q14 - Q21 1 Added 5/25/17 Q1 - Q13 Initial x 4/19/17

DARPA Spectrum Collaboration Challenge (SC2) Frequently Asked Questions Revision 14 Q52: My CIRN design relies upon open source software, such as the USRP UHD driver, as a key part of the design. If during a scored event (e.g., PE1, PE2 and SCE) an open source component of my design unexpectedly fails, crashes, or otherwise is disruptive to the desired functionality of the radio network, can the match be repeated? A52: No. Each team is wholly responsibility for the stability of their entire submitted design including any and all open source software used (including the USRP UHD driver). Q51: In order to qualify for PE1 per Q32, must the source_node and destination_node fields in the Performance message be populated? A51: No. Q50: Per Q32, option #1, for PE1, will any contiguous 1 minute period in which a flow achieves the target throughput be accepted as meeting the qualifying criteria? A50: Yes (throughputs are sliding-window averaged using a 1-minute window with the match interval, specified in the SC2 PE1 Scoring Procedures, as the step size). Q49: The USRP X310 can create an undesired transmission artifact which manifests as local oscillator leakage (i.e. a tone ) which may appear to be a transmission at a disallowed frequency. Will the unintentional local oscillator (LO) be counted as a transmission at a disallowed frequency? A49: If each USRP in a competitors CIRN is using the settings recommended below in Q29 specifically if the USRP is tuned -42 MHz with respect to the scenario center frequency given in colosseum.ini the competitor will not be penalized for the LO appearing at a disallowed frequency. Any other configuration may result in a competitor being assessed penalties for transmission at disallowed frequencies if the magnitude of the LO exceeds the allowed threshold. Q48: In PE1, will the scenario center frequency change during a match? A48: No.

Q47: In PE1, are competitors guaranteed to have a period of time at the beginning of a match without traffic? A47: No. Q46: To qualify for PE1, per Q32, Is there a minimum and maximum periodicity with which the Performance message must be sent? A46: The Performance message must be sent with a minimum periodicity of at least once every 10 seconds, and a maximum periodicity not to exceed 100 times per second. Q45: To qualify for PE1, per Q32, a CIRN must support at least the Hello and Performance Collaboration Protocol Messages. Is a CIRN required to provide the Performance message as a response to an InformationalQuery message, or is unsolicited and periodic transmission of these messages sufficient to qualify? A45: A CIRN is not required to send the Performance message in response to an InformationalQuery message. A CIRN must at least send the Performance message, in an unsolicited fashion, to all other CIRNs Q44: What range of center frequencies will be used for PE1 scenarios? A44: 900-1100 MHz. Q43: How many antennas will be available per SRN in PE1 scenarios? A43: 2 receive and 2 transmit antennas. Q42: Are packets from the IP Traffic Generator fragmented according to the maximum transmission unit (MTU) of the /dev/tr0 interface? A42: Yes.

Q41: During PE1, what is the minimum frequency bandwidth available for a CIRN to transmit over without incurring a penalty for transmitting on a disallowed frequency? A41: 5 MHz. Q40: How is average latency computed for a match scoring interval in which not all packets are delivered? A40: All packets with delay greater than the maximum possible delay, noted as LL CC MMMMMM in the Scoring Procedures, are assigned delay LL CC MMMMMM prior to computing the average. This includes dropped packets (i.e. those with infinite delay). Q39: In order to qualify for PE1, is there a latency requirement a CIRN must meet? A39: The throughput computation will only include traffic delivered up to 10 seconds after the 1- minute evaluation period (see Q32). There is no specified latency which must be achieved. Q38: How many submissions per team will DARPA evaluate for PE1? A38: A maximum of two submissions per team will be evaluated. One evaluation will be conducted following the initial submission deadline of 6:00AM US Eastern Time November 20 th 2017, and one evaluation will be conducted after the final submission deadline of 6:00 AM US Eastern Time November 29 th. Q37: If a team makes two PE1 submissions, which will be used for PE1? A37: The more recent of the two submissions which successfully qualifies per the criteria in Q32 will be used for PE1. Q36: For PE1, how will the power spectral density (PSD) in dbfs (decibels relative to Full-scale) be computed to assess transmission at disallowed frequencies? A36: The following shows the computation of the PSD for a complex-sampled signal xx, normalized by the maximum integer value that can be produced by the analog-to-digital converter such that the largest possible magnitude of xx is one, i.e. max abs(xx) = 1. PPPPPP(xx) = 20 log 10 (abs fft(xx) 20 log 10 (NN) 10 log 10 (ff ss )

Where N is the number of points in the Fourier transform (indicated as fft), and ff ss = 100 10 6 is the sampling rate. Q35: Will PE1 matches contain an unscored training period at the beginning of a match? A35: No. Q34: If my team s design successfully qualifies for PE1, but does not place among the top scoring teams, do I need to complete the Phase 2 Entrance Hurdle in order to remain in the competition? A34: No. All teams whose designs qualify for PE1 are automatically eligible to participate in Phase 2 regardless of the outcome of PE1. Q33: In order to qualify to participate in PE1, when and how must final Collaborative Intelligent Radio Network designs be submitted? A33: Initial CIRN designs are due by 6:00AM US Eastern Time November 20 th 2017. Finalized designs are due by 6:00 AM US Eastern Time November 29 th. Submissions must be compliant with the following criteria to qualify: A single radio container that works in batch mode with an MD5 hash of the container image The image is named pe1.tar.gz, and is located on your NAS shared at /share/nas/<team_name>/pe1/pe1.tar.gz You provide an md5 hash of the submitted image at /share/nas/<team_name>/pe1/pe1.md5 (This must be an **md5** hash) You provide a **SINGLE** modem configuration file for all radios submitted at /share/nas/<team_name>/pe1/config/radio.conf A complete and correct submission will consist of the following folders and files /share/nas/<teamname>/pe1/ /share/nas/<teamname>/pe1/pe1.tar.gz /share/nas/<teamname>/pe1/pe1.md5 /share/nas/<teamname>/pe1/config/ /share/nas/<teamname>/pe1/config/radio.conf /share/nas/<teamname>/pe1/config/batch.json Q32: Per Section 3.2.1 of the SC2 Rules Document, what performance must my CIRN achieve in order to qualify to participate in PE1? A32: In order for a submitted CIRN design to qualify to participate in PE1, the 5-node CIRN must achieve an average data rate of 1.0 Mbps (of application layer payload data, excluding overhead) under either of the following conditions:

1. Each radio link individually achieves an average 1.0 Mbps over any 1 minute period 2. All radio links achieve an average of 1.0Mbps over the same 1 minute period This criteria will be tested in the following configuration: 4 IP traffic flows between 5 radios nodes labeled A-E such that IP traffic flows as follows: o Flow 1: A->B, UDP Traffic of at least 1Mbps o Flow 2: B->C, UDP Traffic of at least 1Mbps o Flow 3: C->D, UDP Traffic of at least 1Mbps o Flow 4: D->E, UDP Traffic of at least 1Mbps At least 30 db SNR is present over 6 MHz of allowed bandwidth (submissions which transmit outside the 6MHz of bandwidth may be deemed ineligible) Additionally, the CIRN must properly register with the Collaboration Protocol Server, and support at least the following Collaboration Protocol Messages: Hello and Performance (with at least one of scalar_performance, bit_error_rate, or frame_loss_rate supported). The Performance message sub-field(s) must be filled with meaningful content; random or constantvalues which do not reflect the achieved throughput may be deemed ineligible. Q31: Is a CIRN allowed to deliver packets and packet fragments out of order to improve its score? A31: Yes. Q30: Do the training bots provided by DARPA maintain the same performance during tournament events (e.g. Scrimmages and scored events PE1, PE2 and SCE) if used as a competitor submission? A30: The training bots are designed in such a way that they will not operate successfully during tournament events. Q29: Are competitors wholly responsible for selecting all USRP settings, such as amplifier gain? A29: Yes, each team s CIRN software is responsible for setting all USRP settings. For convenience, a list of recommended USRP settings has been provided by the USRP hardware manufacturer (National Instruments) and is available at the following location: http://sc2colosseum.pbworks.com/w/page/120410898/optimizing%20srn%20usrp%20perfor mance

Q28: For PE1, will frequencies be disallowed at the onset or during the course of a match? A28: At the onset. Q27: For PE1, how many disallowed frequency bands must a CIRN obey? A27: Two. One disallowed frequency band starting at 0 Hz and extending to the lower most allowed frequency: [0 HHHH, ff mmmmmm HHHH), and one disallowed frequency band starting at the upper most allowed frequency and extending to infinity (ff mmmmmm HHHH, HHHH). Q26: To compute PE1 scores, will the latency of a fragmented IP packet be computed only after all fragments are successfully received? A26: Yes. Q25: During PE1, what is the maximum available frequency bandwidth a CIRN may transmit over without incurring a penalty for transmitting on a disallowed frequency? A25: 20 MHz. Q24: Are the emulated receive antennas guaranteed to be perfectly isolated from the transmit antennas of the same radio node? A24: No. Q23: During PE1, for non-degraded conditions, what nominal gateway-to-gateway bandwidth and latency can be expected for the collaboration channel? A23: A minimum average bandwidth of 1 Mbps and a maximum average round-trip latency of 50ms can be expected under nominal, non-degraded conditions. Q22: During PE1, how often will the collaboration channel experience degraded conditions which do not meet the definition of nominal?

A22: No more than 10% of the available points that can be scored will occur during periods of time when the collaboration channel operates under degraded conditions. Q21: Are incumbent network radios guaranteed to be fixed-position? A21: No. Q20: For PE1, what is the maximum speed of mobile nodes? A20: The maximum speed for mobile nodes is 50 mph. Q19: For PE1, How will scenario specific parameters such as bandwidth, Effective Isotropic Radiated Power (EIRP), and center frequency be communicated to our CIRNs? A19: All scenario specific parameters are contained in the RadioAPI Colosseum config file. Q18: What center frequency range will scenarios models? A18: 10 MHz - 6 GHz. Q17: During scrimmages and competition events (PE1, PE2, and SCE), will our container have access to the internet? A17: No. Q16: Will teams know which SRNs they will use in advance of the PE1 competition? How can we identify the specific radio node our container resides on at startup? A16: No. SRN nodes can be identified by inspecting the /etc/hostname file. Q15: For Phase 1 scrimmages and competition (PE1), how many containers can I deliver? A15: One. Q14: Will the full Colosseum bandwidth be available in every scenario?

A14: No. Q13: Do the eligibility requirements in the SC2 Rules Section 5 permit a non-us team comprised of non-us citizens to participate in the Spectrum Collaboration Challenge? A13: Yes. Q12: My approach includes techniques that require the use of hardware other than the Standard Radio Node. Can I provide my own hardware? A12: No. (see SC2 System Specification Document). Q11: Can I make use of International Traffic in Arms Regulated (ITAR) software in my Collaborative Intelligent Radio Network (CIRN)? A11: No. Q10: Is MIMO enabled (allowed) in the SRN nodes during competition? Is it a requirement? A10: MIMO is enabled and allowed during competition. Colosseum is designed to support 2 TX and 2 RX antennas (2x2 MIMO) per SRN. CIRN designs may choose to take advantage of the provided multiple antenna configuration or may choose to disregard the second antenna input and output. Q9: Will Collaborative Intelligent Radio Networks (CIRNs) be given their current score or other networks' scores by Colosseum infrastructure during PE1 so they can adapt online? A9: No. CIRNs may choose to share relevant performance metrics with each other through the Collaboration Protocol however. Q8: Will SRNs be given GPS-like localization information? A8: SRNs may receive localization information periodically throughout the match. The information's accuracy will reflect modeled conditions in the scenario. For example, localization data provided to indoor nodes may have increased uncertainty compared to that of outdoor nodes.

Q7: During the competition, can we (humans) provide control input to our nodes? A7: No. CIRNs must be fully autonomous. Q6: May teams add or change members as the competition evolves through the phases? A6: Yes, with the following restrictions: 1) A team may remove members at any point in the competition at the sole discretion of the team. 2) A team may add wholly new members at any point at the sole discretion of the team. Wholly new members are those that have not previously participated on any SC2 team. 3) A team may only add new team members who previously participated on another SC2 team under one of the following conditions: The new team member participated as a member of another team in a previous phase of SC2, and is not has not participated in the current phase. The new team member participated as a member of another team in a previous phase, and is changing teams during the allowed grace period, January 1 March 31 of the next phase. The new team member participated as a member of another team which has formally disbanded or otherwise withdrawn from the competition. 4) Teams may not add team members who currently or previously participated as part of the SC2 DARPA team. 5) All team membership changes must be registered with DARPA SC2 Team via email to sc2@darpa.mil. 6) The registered team lead is the only person authorized to make membership changes. Q5: Will our CIRN be told how long a match will last? A5: No. Information about match duration will not be provided to CIRNs. Q4: Can gateway nodes be assumed to be stationary since they have a wired backhaul? A4: No. The use of the term "wired network" in the SC2 Rules Document (Section 2.1) refers to the fact that no RF waveform is required for the gateway to use its second interface to collaborate with other CIRN gateways. The modeled internet connection statistics (such as throughput and latency) may model wired or wireless internet characteristics. Q3: Can teams choose which network node is the gateway? A3: No.

Q2: Can teams control the location of any of their nodes? A2: No. Q1: Will teams be allowed to practice with one another on Colosseum outside of scrimmages or match events? A1: No. Inter-team practices outside of scheduled scrimmage and match events are not allowed. Any attempt to organize such events may be considered collusion per the SC2 Rules Section 5, and may lead to disqualification.