A Cognitive PHY/MAC Proposal for IEEE WRAN Systems

A Cognitive PHY/MAC Proposal for IEEE 802.22 WRAN Systems
Month Year doc.: IEEE yy/xxxxr0 November 2005 A Cognitive PHY/MAC Proposal for IEEE WRAN Systems IEEE P Wireless RANs Date: Authors: Notice: This document has been prepared to assist IEEE It is offered as a basis for discussion and is not binding on the contributing individual(s) or organization(s). The material in this document is subject to change in form and content after further study. The contributor(s) reserve(s) the right to add, amend or withdraw material contained herein. Release: The contributor grants a free, irrevocable license to the IEEE to incorporate material contained in this contribution, and any modifications thereof, in the creation of an IEEE Standards publication; to copyright in the IEEE’s name any IEEE Standards publication even though it may include portions of this contribution; and at the IEEE’s sole discretion to permit others to reproduce in whole or in part the resulting IEEE Standards publication. The contributor also acknowledges and accepts that this contribution may be made public by IEEE Patent Policy and Procedures: The contributor is familiar with the IEEE 802 Patent Policy and Procedures including the statement "IEEE standards may include the known use of patent(s), including patent applications, provided the IEEE receives assurance from the patent holder or applicant with respect to patents essential for compliance with both mandatory and optional portions of the standard." Early disclosure to the Working Group of patent information that might be relevant to the standard is essential to reduce the possibility for delays in the development process and increase the likelihood that the draft publication will be approved for publication. Please notify the Chair Carl R. Stevenson as early as possible, in written or electronic form, if patented technology (or technology under patent application) might be incorporated into a draft standard being developed within the IEEE Working Group. If you have questions, contact the IEEE Patent Committee Administrator at Carlos Cordeiro, Philips John Doe, Some Company

Month Year doc.: IEEE yy/xxxxr0 November 2005 PHY Abstract Digital modulation systems presently make use of two basic modulation technologies: single carrier and multi-carrier. Their features are well-known since they have been deployed for several years around the world for broadcasting applications. Wireless access applications differ from broadcasting since they require : flexibility on the downstream link : variable number of user, variable throughput per user, variable level of protection, etc; multiple access on the upstream link. Single carrier modulation can tackle these objectives through time multiplexing techniques. Multi-carrier modulation is however more flexible since it enables to control the signal in both time and frequency domains. This gives the opportunity to define two dimensional (time and frequency) slots and to map the services to be transmitted in both directions onto a subset of these slots. Two types of multi-carrier modulation has been retained in IEEE (WiMAX) standard: OFDM in the fixed MAN version and OFDMA in the mobile version. In the continuity of IEEE , it is proposed here to consider OFDMA modulation for downstream and upstream links with two technological improvements: Spreading; OQAM waveforming. To meet the tight link budget requirements of WRAN, duo binary turbo code is proposed for service ranges up to 100 Km Carlos Cordeiro, Philips John Doe, Some Company

Month Year doc.: IEEE yy/xxxxr0 November 2005 MAC Abstract We propose the Cognitive MAC (CMAC) layer to be used as the basis for the future IEEE WRAN standard operating in the TV bands. The proposed CMAC is in some respects inspired by the IEEE standard, but it provides major extensions, improvements and also simplifications in order to meet the functional requirements. CMAC is based on a superframe architecture which is general enough to allow multiple wireless systems to coexist in addition to support the flexibility to group multiple vacant TV channels and hence achieve greater capacity. To coexist with incumbent services, CMAC is able to efficiently manage distributed incumbent measurement, control, and recovery procedures, while also providing the necessary spectrum management features. To coexist amongst systems, CMAC is the first of its kind to implement a novel coexistence beacon protocol (CBP) that allows BSs with overlapping coverage areas to coordinate and efficiently share the radio spectrum, hence minimizing interference. The efficiency of CBP is further improved by a new scheme that dynamically synchronizes overlapping BSs. Additional characteristics of CMAC include the support of various traffic types with different QoS requirements, flexible bandwidth management, and a combination of access mechanisms. Carlos Cordeiro, Philips John Doe, Some Company

Presentation Outline Introduction The Cognitive PHY Proposal
November 2005 Presentation Outline Introduction A Glimpse of IEEE The Cognitive PHY Proposal The Cognitive MAC Proposal Conclusions Carlos Cordeiro, Philips

November 2005 Carlos Cordeiro, Philips

The IEEE 802.22 From 18 Mbps to 24 Mbps
November 2005 The IEEE From 18 Mbps to 24 Mbps Propagation delays in excess of 300 µs Operates in TV bands 54 to 862 MHz 6 MHz, 7 MHz and 8 MHz channel bandwidth Carlos Cordeiro, Philips

Deployment Scenario Master/Slave relationship Entities
November 2005 Deployment Scenario Master/Slave relationship Entities Base Station (BS) Consumer Premise Equipment (CPE) 4W CPE transmit power Carlos Cordeiro, Philips

PHY Presentation Outline
November 2005 PHY Presentation Outline Background Top-level description of modulation/coding Channel bonding Modulation Parameters Spreading OFDMA Sensing techniques OQAM/OFDMA Duo-binary CTC Carlos Cordeiro, Philips

802.22 requirements consideration
November 2005 requirements consideration Regional Area Network (up to 30Km) Operate in vacant TV bands Detect vacant TV bands Large delay spread and roundtrip time Data rate: from 1.5 Mbps DS and ~300 Kbps US Should not cause harmful interference to other devices -70dB OOB emission detect and avoid Flexibility Bandwidth, bit rate, TX power, access mechanism, etc Spectral efficiency 4 watt transmission power Carlos Cordeiro, Philips

PHY Overview OFDMA both in uplink and downlink
November 2005 PHY Overview OFDMA both in uplink and downlink QPSK, 16-QAM, and 64-QAM, spreaded-QPSK More than 32 sub channels Contiguous channel bonding upto 3 TV channels ( and beyond in a stack manner) Data rate range from 5Mbps to 60Mbps Option of OQAM/OFDMA and turbo code Carlos Cordeiro, Philips

November 2005 Why OFDMA ? Single carrier and multi-carrier have been used for broadcasting, wireless access, etc Their behavior is well understood (capacity, filtering requirements, PAPR, equalization, flexibility, efficiency) Wireless access differ from broadcasting and most other system flexibility in downstream and upstream link variable # of users, variable throughput, large round-trip signal delay multiple access Multi-carrier system more suitable to meet these objectives It enables to control the signal in time and frequency Results in a two dimensional grid to assign resources to a user  OFDMA Resources can be allocated on a per user basis OFDM used in standards such as WiMedia UWB, WiMAX (Fixed MAN), DAB, DMB, DVB-T, DVB-H, ISDB-T OFDMA used in WiMax, DVB-RCT Carlos Cordeiro, Philips

OFDMA Based on OFDMA (sub-channels per user) November 2005 US/DS
Reduces overhead for short messages Flexibility in choosing modulation/coding for CPE Reduced PAPR for CPEs Carlos Cordeiro, Philips

November 2005 Coding Carlos Cordeiro, Philips

November 2005 Modulation Carlos Cordeiro, Philips

Channel Bonding More data rate Multi-path Diversity Interference
November 2005 Channel Bonding More data rate Multi-path Diversity Small BW signal can have deep fade or flat fade Wider-bandwidth signal provides more frequency/multipath diversity Interference Wider-band reduces the amount of interference Carlos Cordeiro, Philips

Channel Bonding/capacity
November 2005 Channel Bonding/capacity Aggregate TV channels to get more capacity Shannon: C = B.log2(1+S/N) Capacity proportional to BW, but logarithmic with SNR or signal power If S/N is fixed, then capacity increases linearly with bandwidth If signal power is fixed, but bandwidth is increased C = B.log2(1+S/(BNo)) Capacity still increases as bandwidth is increased Carlos Cordeiro, Philips

November 2005 Capacity of aggregated channels as a given signal power is spread over more channels Carlos Cordeiro, Philips

Channel bonding 6, 12, 18 MHz channels Depends on availability
November 2005 Channel bonding 6, 12, 18 MHz channels Depends on availability Several receiver techniques to deal with flexible BW Selectable analog filters Up sampling digital filters Carlos Cordeiro, Philips

Channel bonding structure
November 2005 Channel bonding structure 6K FFT over 3 TV channels 2K per TV channel Null out the outer carriers for 1 or 2 TV channels Fixed inter-carrier spacing Several implementation possibilities 6 MHz 12 MHz 18 MHz Carlos Cordeiro, Philips

Frame structure: Superframe
November 2005 Frame structure: Superframe Carlos Cordeiro, Philips

Spectrum of the signal (before further filtering)
November 2005 Spectrum of the signal (before further filtering) Produced using a 6K FFT for a single TV channel Carlos Cordeiro, Philips

802.22 proposed relative RF emission
November 2005 proposed relative RF emission Carlos Cordeiro, Philips

Inter-carrier spacing,
Table 2: Inter-carrier spacing and FFT/IFFT period values for different bandwidth options November 2005 Inter-carrier spacing and FFT/IFFT period values for different bandwidth options 6 MHz based channels (6, 12 and 18 MHz) 7 MHz based channels (7, 14 and 21 MHz) 8 MHz based channels (8, 16 and 24 MHz) Inter-carrier spacing, DF (Hz) FFT/IFFT period, TFFT (ms) Carlos Cordeiro, Philips

November 2005 OFDMA parameters Carlos Cordeiro, Philips

Modulation/coding modes and corresponding rates
November 2005 Modulation/coding modes and corresponding rates Carlos Cordeiro, Philips

Preamble Superframe preamble Frame preamble: 1-3 TV channels
November 2005 Preamble Superframe preamble Over 1512 sub-carriers (every fourth or second non-zero), 5 MHz BW Simply duplicate for additional TV channels 1 MHz gap between adjacent channels to relax filtering 2 symbol duration (1 more for data) Frame preamble: 1-3 TV channels 1728*N sub-carriers Short preamble is optional (short) (long) Carlos Cordeiro, Philips

Spreaded QPSK/OFDMA Spread data over some sub-carriers (Hadamard)
November 2005 Spreaded QPSK/OFDMA Spread data over some sub-carriers (Hadamard) Increases capturing of multipath diversity Increases resiliency to interferers Simple receiver structure (MMSE) Carlos Cordeiro, Philips

Simulation results for QPSK, rate 3/4
November 2005 Simulation results for QPSK, rate 3/4 Channels: ATSC Brazil D Profile A All with Doppler S-OFDMA gives 2-4dB gain! Carlos Cordeiro, Philips

Preliminary Link Budget(LOS)
November 2005 Preliminary Link Budget(LOS) Carlos Cordeiro, Philips

Other Features Ranging Transmitter Power Control (TPC)
November 2005 Other Features Ranging Transmitter Power Control (TPC) Consideration of multiple antenna Carlos Cordeiro, Philips

Channel Measurement Received signal strength Signal feature detection
November 2005 Channel Measurement Received signal strength Quality measurement of its own signal (TPC, modulation/coding) Fast channel ‘busy’ detection Signal feature detection Detection of the type of the signal ATSC, DVB-T, Part 74, .22, etc Should be robust to receiver imperfections Carlos Cordeiro, Philips

Received Signal Strength
November 2005 Received Signal Strength Several implementation techniques FFT, IOTA/FFT, simple low-pass filter etc Possibility to measure a part of the spectrum Various degrees of performance Integration time and threshold is very important BS sets essential parameters (constant) Either the BS makes the detection decision based on the collective measurement results or CPE’s can make the decision distributed measurement Carlos Cordeiro, Philips

Simulated performances of OFDM and OQAM: detecting ATSC pilot
November 2005 Simulated performances of OFDM and OQAM: detecting ATSC pilot 5ms integration time Carlos Cordeiro, Philips

DTV signal feature detection
November 2005 DTV signal feature detection Should not be sensitive to frequency selective fading, and receiver impairments (e.g., frequency error) Use field sync correlation detection for ATSC, similar correlation for other standards Compare correlation peak to the mean of the standard deviation of the correlation Characterized the theoretical performance Experimental tests Carlos Cordeiro, Philips

Experimental setup for DTV detection
November 2005 Experimental setup for DTV detection 8VSB_SOURCE MULTIPATH SIMULATOR RECEIVER ATTENUATOR Carlos Cordeiro, Philips

November 2005 Based on DTV Laboratory Test Plan (Group C.1)
Month Year doc.: IEEE yy/xxxxr0 November 2005 Based on DTV Laboratory Test Plan (Group C.1) “Static Echoes at various delays”. Carlos Cordeiro, Philips John Doe, Some Company

November 2005 Based on DTV Laboratory Test Plan (Group D.1)
Month Year doc.: IEEE yy/xxxxr0 November 2005 Based on DTV Laboratory Test Plan (Group D.1) “Static multipath with AWGN”. Carlos Cordeiro, Philips John Doe, Some Company

November 2005 Based on Doc.: IEEE802.22-05/0055r7. Profile A.
Month Year doc.: IEEE yy/xxxxr0 November 2005 Based on Doc.: IEEE /0055r7. Profile A. Carlos Cordeiro, Philips John Doe, Some Company

November 2005 Based on Doc.: IEEE802.22-05/0055r7. Profile B.
Month Year doc.: IEEE yy/xxxxr0 November 2005 Based on Doc.: IEEE /0055r7. Profile B. Carlos Cordeiro, Philips John Doe, Some Company

November 2005 Part 74 detection Part 74 devices occupy a small portion of the spectrum Thus, use spectral estimation and statistics of the estimated signal Spectral estimation using FFTs (windowing techniques can also be employed to better localize the spectrum) Perform FFT Average each freq bin Average across freq bin Compute mean and “variance” FFT avg W.F. V>k*avg Carlos Cordeiro, Philips

Part 74 detection (cont.) Detection Theoretical performance
November 2005 Part 74 detection (cont.) Detection Theoretical performance Carlos Cordeiro, Philips

Narrow-band detection (Part 74): Theoretical and simulated performance
November 2005 Narrow-band detection (Part 74): Theoretical and simulated performance Carlos Cordeiro, Philips

Probability of miss detection and false alarm
November 2005 Probability of miss detection and false alarm Carlos Cordeiro, Philips

November 2005 Detector Link Margin Carlos Cordeiro, Philips

OFDM/OQAM Outline Principles of OFDM/OQAM The IOTA Waveform
Month Year doc.: IEEE yy/xxxxr0 OFDM/OQAM Outline Principles of OFDM/OQAM The IOTA Waveform Advantages of OFDM/OQAM Simulation results John Doe, Some Company

OFDM/OQAM principles (1)
Month Year doc.: IEEE yy/xxxxr0 OFDM/OQAM principles (1) Aim: to increase OFDM spectral efficiency by : Removing the guard interval (cyclic prefix); Delivering a sharper spectral signal than OFDM. How: The waveform that modulates OFDM sub-carriers should be as much as possible localized in time and frequency domains to minimize inter-symbol and inter-carrier interferences. John Doe, Some Company

Month Year doc.: IEEE yy/xxxxr0 OFDM/OQAM principles (2) However, the waveform must guarantee orthogonality between sub-carriers and multi-carrier symbols. Appropriate waveform exist but guarantee orthogonality in the real domain  OffsetQAM modulation should be considered on each sub-carrier. Example: IOTA waveform (optimally localized in time and frequency). John Doe, Some Company

Month Year doc.: IEEE yy/xxxxr0 OFDM/OQAM principles (3) OFDM/QAM Transmitted signal: takes the complex value representing the transmitted encoded data sent on the mth sub-carrier at the nth symbol; and the basic functions are obtained by translation in time and frequency of a prototype function such as: With Rectangular function x used in OFDM has weak frequency localization. OFDM/OQAM Introduces a time offset between real and imaginary parts of symbols takes real values; And with As it can be shown that (2) is verified if g is even and if , where is the ambiguity function of g [3]. This is the case for the IOTA function presented hereafter. John Doe, Some Company

November 2005 OFDM/OQAM principles (3) Time-frequency lattice Carlos Cordeiro, Philips

IOTA = Isotropic Orthogonal Transform Algorithm
Month Year doc.: IEEE yy/xxxxr0 The IOTA function (1) IOTA = Isotropic Orthogonal Transform Algorithm IOTA is a prototype function obtained by the orthogonalization of the Gaussian function Its particularity: the IOTA waveform modulating each sub-carrier is: Quasi-optimally localized in both time and frequency Isotropic: it has the same shape as its Fourier transform => delay spread and Doppler are both equally managed John Doe, Some Company

The IOTA function (2) Month Year doc.: IEEE 802.22-yy/xxxxr0
John Doe, Some Company

The IOTA function (3) The IOTA function can be denoted by:
Month Year doc.: IEEE yy/xxxxr0 The IOTA function (3) The IOTA function can be denoted by: Where: The transmitted signal is: John Doe, Some Company

Advantages of OFDM/OQAM (1)
November 2005 Advantages of OFDM/OQAM (1) Spectrum is sharper : 70 dB instead of 30 dB This feature helps to protect the adjacent channels Carlos Cordeiro, Philips

Advantage of OFDM/OQAM (2)
November 2005 Advantage of OFDM/OQAM (2) Cyclic prefix not mandatory more useful bit-rate This extra bit-rate may be used to: Increase the global net bit-rate of the system; Increase the robustness and therefore the range, or decrease the power. Cyclic prefix 1/4 1/8 1/16 1/32 Bit-rate OFDM/QAM 1/2 11.57 Mbits/s 13.5 Mbits/s 14.47 Mbits/s 14.95 Mbits/s Bit-rate OFDM/OQAM 1/2 15.43 Mbits/s Carlos Cordeiro, Philips

Simulation results Simulation parameters: Constellation : 64 QAM
November 2005 Simulation results Simulation parameters: Constellation : 64 QAM Coding rate : ½ Bandwidth : 7 MHZ Channel: 641 MHz Channel model: Profile A of WRAN Carlos Cordeiro, Philips

OFDM/OQAM vs OFDM/QAM with Convolutional FEC
November 2005 OFDM/OQAM vs OFDM/QAM with Convolutional FEC Carlos Cordeiro, Philips

OFDM/OQAM vs OFDM/QAM with Duo-binary Turbo-codes
November 2005 OFDM/OQAM vs OFDM/QAM with Duo-binary Turbo-codes Carlos Cordeiro, Philips

November 2005 State of the art OQAM waveform has been standardized by TIA committee TR8 for Private Land Mobile applications Carlos Cordeiro, Philips

Duo-binary Turbo-codes Outline
November 2005 Duo-binary Turbo-codes Outline Duo-Binary Turbo Codes Internal interleaver Flexibility Performance Simulations Carlos Cordeiro, Philips

Duo-Binary Turbo-codes
November 2005 Duo-Binary Turbo-codes Information bits are encoded by couples Carlos Cordeiro, Philips

Duo-Binary Turbo-code
November 2005 Duo-Binary Turbo-code Duo-Binary input: two decoded bit output at a time Reduction of latency and complexity per decoded bit (compared to Binary TC) Better convergence Circular (tail-biting) encoding No trellis termination overhead Original interleaving scheme Larger minimum distances Improved asymptotic performances Carlos Cordeiro, Philips

Internal Interleaver Algorithmic permutation
November 2005 Internal Interleaver Algorithmic permutation One equation, 4 parameters (P0, P1, P2, P3) Parameters selected such that interleaver is contention-free Adjusting the TC to a blocksize only requires modification of the 4 parameters Quasi-regular permutation (easy connectivity) Inherent parallelism i = 0, …, N-1, j = 0, ...N-1 level 1: if j mod. 2 = 0, let (A,B) = (B,A) (invert the couple) level 2: - if j mod. 4 = 0, then P = 0; - if j mod. 4 = 1, then P = N/2 + P1; - if j mod. 4 = 2, then P = P2; - if j mod. 4 = 3, then P = N/2 + P3. i = P0*j + P +1 mod. N Carlos Cordeiro, Philips

Flexibility Can be easily adjusted to any blocksize
November 2005 Flexibility Can be easily adjusted to any blocksize Storage of the 4 parameters for all blocksizes considered Possibility of a generic approach (default parameters) All coding rates are possible Through puncturing patterns Natural coding rate is ½: increased robustness to puncturing Performance vs complexity: several adjustments are possible Number of iterations, Decoding algorithm, … Implementation: interleaver enables different degrees of parallelism Can be adjusted to meet complexity/throughput requirements Carlos Cordeiro, Philips

November 2005 Flexibility The number of iterations can be adjusted for a better performance-complexity trade-off Carlos Cordeiro, Philips

Performance Duo-Binary TC, 8 iterations, Max-Log-MAP decoding
November 2005 Performance Duo-Binary TC, 8 iterations, Max-Log-MAP decoding IEEE e structured LDPC, BP decoding, 50 iterations AWGN, R=1/2, QPSK N=576 and 2304 (coded blocksize) Carlos Cordeiro, Philips

Short blocksize performance
November 2005 Short blocksize performance Hardware measurements Low BER (down to 10-11) are achievable without error floor Carlos Cordeiro, Philips

Simulation results Simulation parameters Constellation : 64 QAM
November 2005 Simulation results Simulation parameters Constellation : 64 QAM Coding rate : ½ Bandwidth : 7 MHZ Channel: 641 MHz Channel model: Profile A of WRAN Carlos Cordeiro, Philips

Duo-binary Turbo-codes vs Convolutional with OFDM/QAM modulation
November 2005 Duo-binary Turbo-codes vs Convolutional with OFDM/QAM modulation Carlos Cordeiro, Philips

Duo-binary Turbo-codes vs Convolutional with OFDM/OQAM modulation
November 2005 Duo-binary Turbo-codes vs Convolutional with OFDM/OQAM modulation Carlos Cordeiro, Philips

Advantages of Duo-binary Turbo-codes
November 2005 Advantages of Duo-binary Turbo-codes Good performance for a very wide range of blocksizes Highly flexible scheme, enabling a very fine granularity Same encoder/decoder for all blocksizes/coding rates. Several trade-off in performance (number of iterations, decoding algorithm), implementation complexity (degrees of parallelism). Reasonable complexity Approximately 35% decrease in complexity per decoded bit compared to Binary TC. Carlos Cordeiro, Philips

State of the art Duo-binary Turbo-code is a mature technology
November 2005 State of the art Duo-binary Turbo-code is a mature technology This technology has already been selected by several standardization groups IEEE / WiMAX; DVB-RCS; DVB-RCT; ETSI HIPERMAN Carlos Cordeiro, Philips

Summary: Gains brought by OQAM and DTC
November 2005 Summary: Gains brought by OQAM and DTC OFDM/OQAM brings 10% more bit-rate When converted in error protection enables to go from ¾ rate to 2/3 Gain between 1 and 1,5 dB in C/N Duo-binary TC offers 3,5 to 4 dB When combined the gain is at least 4,5 dB that allows to increase the radius by 7,6 km (17%) with QPSK modulation in a Gaussian channel. Carlos Cordeiro, Philips

Final conclusion for WRAN PHY
November 2005 Final conclusion for WRAN PHY OFDMA/Channel bonding Good answer to flexibility requirements Spreading QPSK Captures multipath diversity and increases resiliency to interference (2-4 dB gain) OQAM waveform Increases efficiency and incumbent protection Duo-binary Turbo-codes Powerful error protection technique Carlos Cordeiro, Philips

CMAC Presentation Outline
November 2005 CMAC Presentation Outline Introduction The CMAC Protocol Architecture Data communication Superframe and Frame Structures Network entry and initialization Downstream and Upstream scheduling Coexistence Incumbents Self-Coexistence Synchronization of overlapping BSs Clustering Security Performance Evaluation Carlos Cordeiro, Philips

November 2005 Introduction A Cognitive MAC (CMAC) layer is proposed to be used as the future IEEE MAC for WRANs Some aspects of CMAC have been inspired by the IEEE MAC standard However, major enhancements have been made Support of multiple channel operation; Coexistence with both incumbents and itself (self-coexistence); Measurements (incumbents and itself) Spectrum management (time, frequency and power) The Coexistence Beacon Protocol (CBP) Synchronization of overlapping BSs The Incumbent Detection Recovery Protocol (IDRP) Embedded wireless microphone beacon mechanism Clustering support; etc. Carlos Cordeiro, Philips

Presentation Outline Introduction The CMAC Protocol
November 2005 Presentation Outline Introduction The CMAC Protocol Architecture Data communication Superframe and Frame Structures Network entry and initialization Downstream and Upstream scheduling Coexistence Incumbents Self-Coexistence Synchronization of overlapping BSs Clustering Security Performance Evaluation Carlos Cordeiro, Philips

November 2005 Overview Given the very long propagation delays in WRANs, the BS regulates the medium access Downstream: TDM (Time Division Multiplexing) Upstream: DAMA (Demand Assigned Multiple Access) TDMA Packet Size: 50 bytes Packet Size: 1500 bytes Carlos Cordeiro, Philips

November 2005 Overview (cont.) Combination of polling, contention and unsolicited bandwidth grants mechanisms Support of Unicast/Multicast/Broadcast for both management and data Connection-oriented MAC Connection identifier (CID) is a key component Defines a mapping between peer processes Defines a service flow (QoS provisioning) Carlos Cordeiro, Philips

Protocol Stack Architecture
November 2005 Protocol Stack Architecture Flexibility, scalability and efficiency are core elements Spectrum manager could be implemented in many ways Carlos Cordeiro, Philips

Protocol Stack Architecture (cont.)
November 2005 Protocol Stack Architecture (cont.) Flexible channel assignment Implementers decide on the algorithm Carlos Cordeiro, Philips

Basic Terms and Definitions
November 2005 Basic Terms and Definitions Superframe Defined and delimited by a preamble and the SCH (superframe control header). It is comprised of a number of Frames Frame Comprised of one DS and one US Subframe, where BS and CPEs use to communicate with each other Subframe Formed by a number of Bursts Burst Defined by a two dimensional segment of logical channel (frequency) and MAC slot (time). It may comprise of multiple MAC PDUs belonging to multiple CPEs MAC PDU The smallest unit of transmission/reception by the MAC. It is comprised of the MAC header, the payload, and CRC Carlos Cordeiro, Philips

Motivation The problem The fact The solution
November 2005 Motivation The problem How to offer enhanced capacity and higher data rates? The fact Spectrum occupancy measurements conducted by Shared Spectrum Company from January/2004 to August/2005 have shown that: “There is a significant amount of spectrum available in continuous blocks that are 1 MHz and wider ” “A dynamic spectrum sharing radio with a low agility, contiguous waveform will provide high utility” The solution Simultaneous use of multiple contiguous TV channels Carlos Cordeiro, Philips

Superframe Control Header (SCH)
November 2005 Superframe Structure Superframe Control Header (SCH) TV channels being bonded Coexistence and superframe information Number and size of frames Information on periodic quiet periods ID an transmit power of transmitter Location configuration information Carlos Cordeiro, Philips

Frame Structure CMAC is based on a TDD frame structure
November 2005 Frame Structure CMAC is based on a TDD frame structure Reduced complexity In general, less measurements overhead The flexible architecture (with the Spectrum Manager) already brings with it aspects of FDD The CMAC frame structure is comprised of two parts A predominantly downstream (DS) subframe An upstream (US) subframe Carlos Cordeiro, Philips

Frame Structure (cont.)
November 2005 Frame Structure (cont.) Carlos Cordeiro, Philips

Time/Frequency Structure of a MAC Frame
November 2005 Time/Frequency Structure of a MAC Frame Carlos Cordeiro, Philips

Network Entry and Initialization
November 2005 Network Entry and Initialization The key problem Carlos Cordeiro, Philips

Downstream (DS) Transmissions
November 2005 Downstream (DS) Transmissions DS = core messages + data (transmitted in bursts) Two core DS messages: DCD and DS-MAP Bursts identified by DIUC (Downstream Interval Usage Code) Each burst may contain data for several CPEs DCD (Downstream Channel Descriptor) Establishes association between DIUC and actual PHY parameters (e.g., modulation and coding) DS-MAP (Downstream map) Defines the usage (i.e., scheduling) of the downstream Critical, hence first message in each frame For self-coexistence purposes, the BS may choose to use part of DS subframe for CBP protocol operation Carlos Cordeiro, Philips

Upstream (US) Transmissions
November 2005 Upstream (US) Transmissions US = core messages + data (contention and bursts) Two core DS messages: UCD and US-MAP Bursts identified by UIUC (Upstream Interval Usage Code) The upstream can be segmented into several UIUC Contention-based Initialization, Bandwidth Request, Urgent Coexistence Situation (UCS), CBP slots (SCS) Data Bursts UCD (Upstream Channel Descriptor) Establishes association between UIUC and actual PHY parameters (e.g., modulation and coding) US-MAP (Upstream map) Defines the usage (i.e., scheduling) of the upstream Contains “grants”addressed to a particular CPE or a set of CPEs (e.g., for self-coexistence) Carlos Cordeiro, Philips

Bandwidth Management: Request/Grant Scheme
November 2005 Bandwidth Management: Request/Grant Scheme Self correcting No acknowledgement All errors are handled the same way (i.e., periodic aggregate requests) Bandwidth requests Are always per connection Can specify DS/US Traffic Constraint IEs for better self-coexistence Bandwidth grants Can be either per connection or per CPE Grants (given as durations) are carried in US-MAP messages CPE converts time into amount of data using information about the UIUC Carlos Cordeiro, Philips

Scheduling Services Unsolicited Grant Services (UGS)
November 2005 Scheduling Services Unsolicited Grant Services (UGS) For CBR and CBR-like flows (T1/E1) No specific bandwidth request issued by CPE Real-time Polling Service (rtPS) For rt-VBR-like service flows such as MPEG video CPEs are polled to meet delay requirements Non-real-time Polling Service (nrtPS) For non-real-time flows with better than best effort service such as bandwidth-intensive file transfer CPEs are polled and can use contention interval Best Effort (BE) E.g., Web surfing CPEs use contention interval only Carlos Cordeiro, Philips

Coexistence Two primary types Measurements can be classified as:
November 2005 Coexistence Two primary types With incumbents (TV service and Part 74 devices) With other overlapping cells Self-Coexistence Measurements can be classified as: In-band In case of incumbents, requires quiet periods (QP) Out-of-band No need for quiet periods Coexistence is achieved by a joint application of: Spectrum management (frequency and power) “Interference-free” traffic scheduling (time) Carlos Cordeiro, Philips

Measurements Measurements form a key component of CMAC
November 2005 Measurements Measurements form a key component of CMAC Protection of incumbents and self-coexistence The BS may request multiple measurements in a single management message E.g., ATSC, DVB, Wireless Microphone, Measurement messages may be transmitted through multicast Allows the implementation of advanced features such as clustering Bandwidth efficient Carlos Cordeiro, Philips

Measurements (cont.) Bulk Measurement Request (BLM-REQ)
November 2005 Measurements (cont.) Bulk Measurement Request (BLM-REQ) Transmitted by the BS to CPEs Includes information such as Channels to measure Multiple single measurement requests Bulk Measurement Response (BLM-RSP) Transmitted by CPE to BS If needed, acknowledges the receipt of the BLM-REQ message Bulk Measurement Report (BLM-REP) Returns multiple single measurement reports Bulk Measurement Acknowledgement (BLM-ACK) Transmitted by BS to CPE Acknowledges receipt of measurement report Carlos Cordeiro, Philips

Measurements (cont.) Single measurements can be of various types
November 2005 Measurements (cont.) Single measurements can be of various types Signal specific measurement request TV system and Wireless microphones Beacon measurement request CBP, BS, and Wireless microphone beacons CPE statistics measurement request Stop measurement request Location configuration measurement request A range of parameters can be specified Carlos Cordeiro, Philips

November 2005 Measurements (cont.) There is almost a one-to-one correspondence between measurement requests and reports Some of the individual reports are: Signal specific measurement report TV/Wireless Microphone system type, measured value, precision, etc. Beacon measurement report Information on any CBP, BS, or Wireless microphone beacons received CPE statistics measurement report E.g., Packet error rate Location configuration measurement report If known, location information (GPS, triangulization, and so on) Carlos Cordeiro, Philips

November 2005 Channel Management Channel management is key to effective network coordination, coexistence and sharing Included in two modes Embedded Non-embedded A set of messages are defined to allow flexible management of channels, including: Add/remove channel(s) to/from current set of channels Switch channel(s) of operation Quiet selected channel(s) – possibly to perform in-band measurement Carlos Cordeiro, Philips

Coexistence with Incumbents
November 2005 Coexistence with Incumbents Accomplished through the following steps: Measurements (discussed earlier) Detection TV: For more info, please see PHY proposal. Wireless Microphones PHY solution: For more info, please see PHY proposal. MAC solution: See next slide. Incumbent Notification Incumbent Detection Recovery Carlos Cordeiro, Philips

MAC Layer Detection of Wireless Microphones
November 2005 MAC Layer Detection of Wireless Microphones From the transmitter perspective, wireless microphone beacons (WMB) can be of two types Embedded device which has the additional capability of emitting WMBs Non-embedded Currently addressed by the Part 74 Study/Task Group Based on the proposed MAC layer, we have developed an embedded WMB approach that: Reliably detects multiple collocated networks Upon sending WMBs, this mechanism causes minimal, if any, harmful interference to collocated networks Once either BSs or CPEs detect the WMB, a dissemination is made and all present networks vacate the channel Carlos Cordeiro, Philips

Incumbent Notification
November 2005 Incumbent Notification The problem How to notify the BS about the presence of incumbents in a timely fashion? Two solutions are possible CPEs with upstream bandwidth allocation Send report provided bandwidth and time are available; and/or Set dedicated bits in MAC header CPEs without upstream bandwidth allocation Urgent Coexistence Situation (UCS) Notification slots reserved specifically for incumbent notification purposes Can use either contention-based or contention-based CDMA access The size of a slot fits the smallest MAC frame necessary to perform the incumbent notification Carlos Cordeiro, Philips

Incumbent Notification (cont.)
November 2005 Incumbent Notification (cont.) Carlos Cordeiro, Philips

Incumbent Notification (cont.)
November 2005 Incumbent Notification (cont.) The BS can use various strategies depending upon how reliable it wants the notification to be Trade-off between overhead and data efficiency Scalability Carlos Cordeiro, Philips

Incumbent Detection Recovery
November 2005 Incumbent Detection Recovery The problem How does the cell recover from an UCS with incumbents in a timely fashion? Carlos Cordeiro, Philips

Incumbent Detection Recovery (cont.)
November 2005 Incumbent Detection Recovery (cont.) The Incumbent Detection Recovery Protocol (IDRP) Introduces the concept of Backup Channel The network not only performs in-band measurements, but also out-of-band measurements Out-of-band measurements will determine a suitable Backup Channel The network falls back to the Backup Channel in case communication is preempted by an incumbent The algorithms at both the BS and CPEs are provided These algorithms also account for the case when no Backup Channel is available Carlos Cordeiro, Philips

Self-Coexistence The general problem November 2005 TDMA Schedule
Carlos Cordeiro, Philips

Self-Coexistence (cont.)
November 2005 Self-Coexistence (cont.) Indeed a major issue E.g., h Becomes even more critical in given The large coverage range Its unlicensed nature Directional antennas at CPEs do not address the problem Carlos Cordeiro, Philips

November 2005 Self-Coexistence (cont.) Some approaches to better self-coexistence Over the backhaul Pros can wash its hands (throw the “hot potato” to somebody else) Cons Will there be really a “common backhaul” between competing WISPs? Can rely on that? What if this “common backhaul” is down? Can rely on the “upper layers” to take care of self-coexistence? Coordination is an active process (e.g., quiet periods), and not a “once-in-a-month thing” Over-the-air Built-in and self-healing system More complex MAC layer (but just a little more) Carlos Cordeiro, Philips

November 2005 Self-Coexistence (cont.) Two solutions are proposed BS beacon based The Coexistence Beacon Protocol (CBP) Both solutions: Can be implemented either over-the-air or via a backbone Here, we focus on the over-the-air implementation Allow either one-way or two-way (i.e., negotiation) communication The BS and its CPEs shall participate in the self-coexistence task Carlos Cordeiro, Philips

November 2005 Self-Coexistence (cont.) BS beacon based Implemented through overheard BS beacons BS beacons carry various information: Channels used Quiet periods Frame information Transmit power level If needed, can use sensing antenna for this purpose Allows better TPC and sharing in frequency only Case 1: Case 2: Carlos Cordeiro, Philips

November 2005 Self-Coexistence (cont.) Coexistence Beacon Protocol (CBP) CBP is executed by CPEs but under BS control CPEs transmit coexistence packets carrying two types of information About the cell About a CPE’s reservations with the BS Allows better TPC and sharing in both frequency and time CBP beacon CBP beacon Carlos Cordeiro, Philips

November 2005 Self-Coexistence (cont.) Carlos Cordeiro, Philips

But what does the CPEs do with this information?
November 2005 But what does the CPEs do with this information? If previously requested by the BS, report it Future upstream bandwidth reservation requests can contain time allocation constraints For example, a CPE can specify: “Give me 100Kb of airtime, but not between T1 and T2” Note on the BS Traffic Constraint (TRC-REQ/RSP) management messages are also available to the BS For example, can be used before the BS allocates any time for the CPE Allow the BS to inquire CPE about possible time allocation constraints Carlos Cordeiro, Philips

Then, what does the BS do about all this?
November 2005 Then, what does the BS do about all this? If possible and desirable, avoid each other by switching channels Better TPC Implement “interference-free” scheduling Sharing in time and frequency Carlos Cordeiro, Philips

Synchronization of Overlapping BSs
November 2005 Synchronization of Overlapping BSs The problem Frames of co-channel overlapping BSs are asynchronous, which makes coexistence even harder Numerous benefits to synchronization Incumbent protection Quiet period synchronization of overlapping BSs Improved detection Self-Coexistence Logical channel amongst overlapping BSs Efficient sharing of resources BS1: BS2: Time Carlos Cordeiro, Philips

Synchronization of Overlapping BSs (cont.)
November 2005 Synchronization of Overlapping BSs (cont.) Synchronization is proposed amongst multiple collocated networks Carlos Cordeiro, Philips

November 2005 Clustering Alleviate much of the redundancy involved in the execution of the coexistence mechanisms So, very suitable for Can be employed in all coexistence mechanisms, except for the protection of Wireless Microphone services Based on key observations Sensing outcome of close-by CPEs are likely to be “similar” CPEs are stationary It is a two-step process conducted by the BS Formation of Physical Cluster Formation of Logical Cluster Carlos Cordeiro, Philips

Clustering: Physical Cluster (cont.)
November 2005 Clustering: Physical Cluster (cont.) Creation of Physical Clusters is totally localized at the BS No direct involvement from CPEs The BS groups together CPEs sensing “similar” characteristics of the incumbent signal Could also be based on location relative to the incumbent transmitter Carlos Cordeiro, Philips

Clustering: Physical Cluster (cont.)
November 2005 Clustering: Physical Cluster (cont.) Based on the well-known k-means clustering algorithm The algorithm Initially, no clustering CPEs report measurements to the BS (BLM-REP) which constructs incumbent profiles Then, the BS runs the clustering algorithm Carlos Cordeiro, Philips

Clustering: Logical Cluster (cont.)
November 2005 Clustering: Logical Cluster (cont.) Formed by CPEs belonging to different Physical Clusters Allows the BS to group those CPEs that are less likely to contend for the same airtime CPEs within a Logical Cluster perform the same “coexistence task” Carlos Cordeiro, Philips

Security Sublayer Based on IEEE 802.16e/D12 security sublayer
November 2005 Security Sublayer Based on IEEE e/D12 security sublayer Generic security framework made specifically for BWA networks Meets all the security requirements identified for the WRAN Standard Deeply studied and improved by various security experts (including IEEE and IETF ones) Composed of two sublayers A Privacy Key Management protocol (PKM) which provides authentication, authorization and secure key distribution between the BS and the CPE An encapsulation protocol which provides data packets protection Carlos Cordeiro, Philips

Security Sublayer (cont.)
November 2005 Security Sublayer (cont.) Mutual Authentication of the devices Either using RSA and digital certificates Or using EAP and EAP-method specific credentials Authentication of the subscribers (optional) Using EAP and EAP-method specific credentials Authorization based on authenticated CPE and/or subscriber identity Give access to dedicated service flows Carlos Cordeiro, Philips

Security Sublayer (cont.)
November 2005 Security Sublayer (cont.) Data packets encryption Using strong cryptographic algorithms (AES) Management frames integrity protection Using keyed message authentication codes Protection against Deny of Service and other attacks Protection of management frames against forgery and replay attacks Protection of data frames against replay attacks Protection of EAP packets during subscribers authentication Protection of every key negotiation phase, using digital signatures and random numbers Carlos Cordeiro, Philips

Performance Evaluation
November 2005 Performance Evaluation All aspects of CMAC are being implemented in OPNET OPNET is considered the most well-reputated and reliable network simulation tool available today In all simulations: In case of quiet periods (QP), every CPE performs detection in all in-band channels (e.g., N-1, N, and N+1 in case of a single TV channel) DFS model is implemented as per the requirements document No fragmentation or packing Some common simulation parameters Superframe size = 12 frames, where Frame size = 40 ms Packet size = 1 Kbyte Detection time per TV channel = 13 ms 64-QAM rate 2/3 and Symbol time = 310 µs Carlos Cordeiro, Philips

Throughput at the MAC SAP
November 2005 Throughput at the MAC SAP Evaluate the throughput of CMAC under varying number of bonded TV channels 1 BS and 127 CPEs Carlos Cordeiro, Philips

Throughput at the MAC SAP (cont.)
November 2005 Throughput at the MAC SAP (cont.) Impact of QP on throughput is more confined to high load scenarios The scheduler can properly handle this Channel bonding provides significant performance improvement Carlos Cordeiro, Philips

Channel Efficiency Evaluate the channel utilization
November 2005 Channel Efficiency Evaluate the channel utilization The overall impact of QPs is only noticeable in high loads Fragmentation and packing can improve these figures even more Carlos Cordeiro, Philips

November 2005 Network Joining Time Evaluate, for the worst case scenario, the average network joining time by a CPE CPEs first scan channel for a time equivalent to a frame size CPE stays in a channel for a superframe duration after that This is followed by network entry and initialization More efficient algorithms can be easily implemented Carlos Cordeiro, Philips

Network Joining Time (cont.)
November 2005 Network Joining Time (cont.) 1 BS and 127 CPEs BS is powered up at simulation startup CPEs power up at random times FRD requires joining time under 10 sec Carlos Cordeiro, Philips

November 2005 Impact on QoS Evaluate the impact of quiet periods and incumbents on QoS Traffic pattern A total of 3 Mbps constant aggregate US traffic DS traffic varies between 3 Mbps and 15 Mbps All 127 CPEs establish connections with BS Out of these, 4 real time (QoS) connections at 32 Kbps each Other connections are BE or non-real time Carlos Cordeiro, Philips

November 2005 Impact on QoS (cont.) The overall impact on average downstream delay is very small QoS can be satisfied to a large extent Effect of Queuing (sec) Carlos Cordeiro, Philips

November 2005 Impact on QoS (cont.) The overall impact on average upstream delay is not so small as in the downstream case Despite of that, QoS can still be satisfied to a significant extent (sec) Carlos Cordeiro, Philips

Handling of Incumbents
November 2005 Handling of Incumbents Evaluate the detection, notification and recovery capability of CMAC 1 BS and 9 CPEs TV station starts in-band operation at a random time Incumbent is detected during quiet period Carlos Cordeiro, Philips

Handling of Incumbents (cont.)
November 2005 Handling of Incumbents (cont.) Network operation is quickly restored BS and unaffected CPEs switch to Backup Channel CPEs who do not receive switch message go to Backup Channel after timeout (2 frames) Channel A Channel B Carlos Cordeiro, Philips

Handling of Incumbents (cont.)
November 2005 Handling of Incumbents (cont.) Evaluate the dynamics of channel bonding Together with handling of incumbents Network can switch to one or more Backup Channel Channel A Channel B Carlos Cordeiro, Philips

November 2005 Conclusions Proposed a PHY and MAC that addresses the requirements set forth by the WG PHY Based on OFDMA Spreaded OFDMA O-QAM Flexible channel configurations (6, 12, and 18 MHz) TV and Part 74 detection MAC Coexistence is a key feature Incumbent protection Self-coexistence CBP and IDRP protocols, superframes, support of channel bonding, etc. Carlos Cordeiro, Philips

A Cognitive PHY/MAC Proposal for IEEE WRAN Systems

Similar presentations

Presentation on theme: "A Cognitive PHY/MAC Proposal for IEEE WRAN Systems"— Presentation transcript:

Similar presentations

About project

Feedback

Log in

Auth with social network:

A Cognitive PHY/MAC Proposal for IEEE WRAN Systems

Similar presentations

Presentation on theme: "A Cognitive PHY/MAC Proposal for IEEE WRAN Systems"— Presentation transcript:

Similar presentations

About project

Feedback