Cognitive Radio Networks Advanced Mobile Communication Networks Integrated Communication Systems Group Ilmenau University of Technology
Outline Introduction Cognitive Radio Technology Spectrum Sensing Spectrum Management Spectrum Mobility Spectrum Sharing Cognitive MAC Protocols Conclusions Example From ICS Research References 2
Introduction 3
Introduction Limited spectrum Inefficiency in spectrum usage Utilization of the assigned licensed spectrum varies between 15% and 85% A new technology is required to enable better utilization of unused licensed spectrum Cognitive radio technology 4
Cognitive Radio Technology 5
What is a Cognitive Radio Definitions Federal Communications Commission Definition: a cognitive radio is a radio that can change its transmitter parameters based on interaction with the environment in which it operates Mitola Definition: the cognitive radio identifies the point at which wireless PDAs and the related networks are sufficiently computationally intelligent on the subject of radio resources and related computer-to-computer communications to detect user communication needs as a function of use context, and to provide radio resources and wireless services most appropriate to those needs 6
What is a Cognitive Radio Characteristics Cognitive capability: the capability of the radio technology of capturing or sensing information from its radio environment (monitoring the power in some frequency band) Reconfigurability: the ability of the radio to be dynamically programmable according to the radio environment (transmission and receipt of data on a variety of frequencies and using various radio access technologies) 7
What Should Cognitive Radio Enable to Do? Determination of unused portion of spectrum (spectrum holes or white space) Selection of the best available channel Coordination of the access with other users Detection of the appearance of licensed users vacate the channel 8
Cognitive Radio Network Architecture 9
Cognitive Radio Network Architecture Primary networks Networks with access right to certain spectrum bands, e.g. common cellular systems and TV broadcast networks Users of these networks are referred to as primary users. They have the right to operate in licensed spectrum Users of certain primary network do not care of other primary or secondary network users Secondary networks Do not have license to operate in the spectrum band they currently use or aim at using Opportunistic spectrum access Users of these networks are referred to as secondary users. They have no right to access licensed bands currently used Additional functionalities are required to share licensed spectrum bands with other secondary or primary networks 10
Cognitive Radio Cycle Radio environment Transmitted signal RF stimuli Spectrum Sharing Spectrum Mobility Primary user detection Spectrum Sensing Channel capacity Decision request Spectrum hole Spectrum characterizations Spectrum Decision 11
Spectrum Sensing 12
Spectrum Sensing Spectrum sensing is determined as the capability of detection of spectrum holes Spectrum Sensing Transmitter Detection Cooperative Detection Interference - Based Detection Matched Filter Detection Energy Detection Cyclostationary Feature Detection 13
Transmitter Detection Location of primary receiver is not known (no signaling between primary and secondary users) detection of the weak signal from a primary transmitter through local observations of secondary users Mechanisms Matched filter detection Energy detection Cyclostationary feature detection 14
Transmitter Detection Matched filter detection Optimal detector when all information of the primary user signal is known (it maximizes the received signal-to-noise ratio) Produces poor performance if primary users information not known Energy detection Used if primary users information not known Cyclostationary feature detection Mean and autocorrelation of modulated signals exhibit periodicity (modulated signals are coupled normally with sine wave carriers, pulse trains, repeating spreading, etc.) Differentiates noise from modulated signal (noise is a wide-sense stationary signal with no correlation) Computationally complex and requires long observation time 15
Transmitter Detection Primary and secondary networks are physically separated transmitter detection can not avoid interference due to the lack of the primary receiver s information Even if the secondary user has a line of sight to the transmitter, the secondary user may not be able to detect the transmitter, e.g. due to shadowing, etc. Can not prevent hidden terminal problem Primary transmitter range Interference Interference Secondary transmitter range Secondary user Primary user Transmitter Can not detect transmitter Can not detect transmitter 16
Cooperative Detection Incorporating information from multiple secondary users to detect primary users Clasified into Centralized cooperative detection Controlled mostly by cognitive BSs Cognitive BS collects sensing information from all secondary users it serves and detects spectrum holes Distributed cooperative detection No centralized infrastructure Observations are exchanged among secondary users More accurate detection than that based on single secondary user observations, e.g. transmitter detection. Moreover, multi-path fading and shadowing effects are mitigated 17
Interference-Based Detection Interference takes place at the receiver side (i.e. the receiver is disrupted). However, it is controlled at the transmitter side, e.g. using power control, etc. Interference temperature is well-known example (receiverdetection model) Power at receiver Licensed signal New opportunities for spectrum access Minimum service range with interference cap Service range at orginal noise floor Secondary users are allowed to communicate as long as they do not produce more interference than this limit Orginal noise floor 18
Challenges Interference temperature measurement Secondary users are aware of their locations and transmission power. They are not aware of primary users locations Currently, no practical way for a cognitive radio to measure or estimate the interference temperature at neighbor primary users Spectrum sensing in multi-user networks Multi-user environment makes it more difficult to sense primary users (secondary networks coexist with each other as well as with primary networks) Cooperative detection can be considered as possible solution for such environments Detection capability Detection of primary users in very short time is essential OFDM-based secondary users are best adequate (multi-carrier sensing can be exploited) 19
Spectrum Decision 20
Managing Available Spectrum Spectrum bands are spread over wide frequency range including licensed and unlicensed bands Radio environment characteristics show fast and mostly not predictable variation over time Secondary users have to select the best spectrum band meeting their QoS requirements spectrum management functions are required Spectrum management includes following steps 1. Getting data from spectrum sensing 2. Performing spectrum analysis 3. Finally making a spectrum decision 21
Spectrum Analysis Characterizes sensed spectrum holes to obtain the band appropriate for user s requirements Characteristics of spectrum holes Interference Some spectrum bands are more crowded than others Based on the interference at primary receivers, the allowed sending power of secondary user can be derived channel capacity is estimated Path loss Path loss increases as frequency increases. To retain the capacity when switching to higher frequency, sending power should be increased more interference produced Wireless link errors Modulation scheme and interference affect strongly the error rate Link layer delay Affected by the interference, path loss, etc. Holding time Expected time duration the secondary user can occupy the channel 22
Spectrum Decision Entity or CR Engine Once spectrum bands are characterized, the band best meeting QoS requirements should be selected spectrum decision function should be aware of QoS requirements of current ongoing applications Spectrum decision rules are required QoS requirements for secondary user Data rate Acceptable error rate Delay... 23
Challenges Decision Model Development of suitable decision rules that consider spectrum bands characters is until now an open issue Multiple spectrum band decision In case secondary users are capable of using multiple channels for transmission simultaneously, it is important to determine the number of spectrum bands available and select the bands appropriately Spectrum decision over heterogeneous spectrum bands Support spectrum decision operations on both licensed and unlicensed bands is challenging 24
Spectrum Mobility 25
Spectrum Mobility The process when a secondary user changes its frequency of operation, also called spectrum handoff Reasons - Operating channel becomes worse - Primary user wants to communicate on the channel - User movements (available spectrum bands change) Requirements - Low latency - Transparence to upper layers protocols if possible - No impairments on ongoing applications (ideal case) Multi-layer mobility management is required with which protocols of many layers cooperate to support mobility, 26
Challenges Smooth spectrum mobility schemes Synchronization between protocols of many layers and possibly with applications to support smooth spectrum handoffs (e.g. applications or protocols switch from operation mode to another upon prediction of a spectrum handoff, etc.) Support of horizontal (changing channels while staying in the same secondary network) and vertical handoffs (between secondary networks) Performing spectrum handoffs to maintain QoS requirements satisfied 27
Spectrum Sharing 28
Spectrum Sharing Considered similar to Medium Access Control (MAC) issue in existing systems. However, different challenges arise due to Coexistence with licensed users Wide range of available spectrum Spectrum sharing steps Spectrum sensing: detect unused spectrum holes Spectrum allocation: allocation of possible target channels based on spectrum sensing results and allocation policies Spectrum access: coordination of access to the allocated channel to avoid collisions Transmitter-receiver handshake: negotiation of communication channel between sender and receiver Spectrum mobility: enable continuous communication between sender and receiver in spite of primary user appearance on the used channel 29
Spectrum Sharing Techniques 1/3 Spectrum sharing techniques are classified according to Architecture Centralized - Centralized entity controls the spectrum allocation and access - Secondary users do observations and report to the centralized entity, which creates spectrum allocation map Distributed - Applied when construction of infrastructure is not possible or not preferable - Each node is responsible for the spectrum allocation 30
Spectrum Sharing Techniques 2/3 Spectrum allocation behavior Cooperative - Observation results of each node are shared with other nodes spectrum allocation is done based on these measurements - These techniques result in better spectrum utilization at the cost of considerable signaling between nodes Non-cooperative (selfish) - Each node does its observations and allocates its spectrum band - These techniques result in reduced spectrum utilization. However, they may be practical for certain applications or situations 31
Spectrum Sharing Techniques 3/3 Spectrum access technology Interweave spectrum sharing (sometimes called overlay in the noninformation-theoretic literature) - Secondary nodes access spectrum holes not used by primary networks Interference to primary users is minimized Underlay spectrum sharing - Based on spread spectrum techniques developed for cellular networks - After acquiring spectrum allocation map, secondary users begin sending, so that their transmission power is regarded as noise by licensed users - Overlay spectrum sharing - Cooperation with the primary transmission - SUs devote part of its transmit power to enhance the primary signal - Interference level may be increased 32
Intra/Inter-Network Spectrum Sharing Inter-network spectrum sharing Secondary user (operator1) Secondary user (operator2) Intra-network spectrum sharing Inter-network spectrum sharing Centralized inter-network spectrum sharing: secondary networks organize cooperatively the spectrum allowed to be accessed by users of each secondary network, e.g. by means of central spectrum policy server, etc. Distributed inter-network spectrum sharing: BSs of secondary networks compete to allocate spectrum holes 33
Challenges Common control channel (CCC) Tasks Transmitter-receiver handshake Communication with a central entity organizing the spectrum allocation Sensing information exchange Problems Fixed CCC is infeasible (CCC must be vacated when a primary user appears on it) CCC for all users seems to be topology-dependent, thus CCC varies over time If no CCC is allocated, transmitter-receiver handshake becomes a challenge Dynamic radio range Radio range and characteristics change with operating frequency CCC must be selected carefully (better to select CCC in lower spectrum bands and data channels in higher ones) Spectrum unit Existing techniques consider channel as the basic spectrum unit. As known, channel may be time slot, frequency, code, etc. 34
Cognitive Radio MAC Protocols 35
Classification Cognitive Radio (CR) MAC Protocols Random Access MAC Protocols Time Slotted MAC Protocols Hybrid MAC Protocols CR Centralized MAC (CSMA/CA-like random access for control packets and data) Single Radio (Time synchronized control and data slots) (Partially time slotted and partially random access) CSMA-MAC IEEE 802.22 DSA Driven MAC CR Ad Hoc MAC Single Radio Multi Radi o SCA-MAC HC-MAC DOSS DSA-MAC C-MAC OS-MAC SYN-MAC 36
Classification Classified according to the access method into Random access MAC protocols No need for time synchronization Based on Carrier Sense Multiple Access (CSMA) Time slotted MAC protocols Need for network-wide synchronization Time is divided into slots for both control and data channels Hybrid MAC protocols Partially slotted transmission, in which Signaling generally occurs over synchronized time slots Data transmission may have random channel access schemes without time synchronization Classified according to the architecture into CR centralized MAC Central entity manages, synchronizes and coordinates operations among secondary users CR Ad Hoc MAC No central entity, neighbors cooperate to gain access to available channels 37
CSMA-MAC Protocol PS Busy RTS RTS CTS Data CRN Busy Carrier sense(t s ) RTS CTS Data (a) PS Busy RTS RTS CTS Data CRN Busy Carrier sense(t s ) RTS (b) Collision Still no successfull transmisssion PS Busy RTS RTS CRN Busy Carrier sense(t s ) RTS CTS Data (c) PS CRN Channel is free Busy Busy Carrier sense(t s ) RTS CTS Data (d) Primary System (PS) Cognitive Radio Network (CRN) 38
CSMA-MAC Protocol Case (a): CRN out of the range of PS Primary user gains the access to the channel and sends its data Secondary user senses the channel for a period t s The secondary user finds the channel vacant (assuming the primary and secondary users are separated by a large distance) The secondary user gains access to the channel and sends its data Case (b): CRN in the range of PS but not of the PS transmitter RTS packet sent by the secondary user experiences collision The secondary user waits for the next transmission opportunity after repeating the previous sensing process Case (c): PS experiences internal collision Primary user sends repeated RTS packets. However, a collision incurs each time Secondary user can start transmission Case (d): Channel is not used by PS Primary user has no packets to send, thus the channel is free Secondary user can start transmission 39
SYN-MAC Protocol Available channels 2 2 4 4 1 2 5 A B C D E F 1 3 5 1 1 3 3 4 4 Slot for Ch1 Slot for Ch2 Slot for Ch3 Slot for Ch4 Slot for Ch5 Ch1 Ch2 Ch3 Ch4 RTS1 CTS1 Data IE IE Primary user active Ch5 RTS CTS DATA Information Event (IE) Backoff 40
SYN-MAC Protocol Time is divided into fixed-time intervals (slots) Each time slot is dedicated to one channel Each node has two radios, one for listening to control messages and one for sending data C wants to send data to D Each node knows the available channel sets of their neighbors Channels 1 and 5 are common C chooses Ch1 for transmission and starts negotiation over it using RTS/CTS Once negotiation is successful, data transmission takes place on Ch1 B observes that primary user of Ch 4 has returned B knows that it can reach its neighbors (A and C) through Ch2 B waits for the time slot which represents Ch2 B sends Information Events (IE) control message with its new channel list Nodes A and C, on receiving this information learn that node B will not be available on Ch 4 E applies the same procedure as B to notify D and F that Ch 4 is removed from its channels list 41
CR Standards IEEE 802.11af Super Wi-Fi (Feb. 2014) IEEE 802.22 (July 2011) TV white space spectrum in the VHF and UHF, 54 790 MHz, OFDM Cognitive sensing, flexible adjustment of bandwidth, modulation, power and coding Database driven approach: Access granted by Geolocation DataBase (GDB) Local sensing is used in certain special cases Up to 1 km, 35.6 Mbit/s for 8 MHz channels Access points (AP) and stations Wireless local area network (WLAN) 30-100 km, 19 Mbit/s for 6 MHz channels Base stations (BS) and customerpremises equipment (CPE) Wireless regional area network (WRAN) 42
Example from ICS Research Flexible MAC as Spectrum Mobility Solution by André Puschmann 43
Motivation TDMA CSMA/CA Power consumption Low High Bandwidth utilization Maximum Low Preferred traffic level High Low Dynamics (network change) Poor Good Effect of packet failure High (latency) Low Synchronization Crucial - Reference: TDMA Protocol Requirements for Wireless Sensor Networks, IEEE 2008 44
Proposed Improvement Flexible Link Layer Architecture [1] Features: Multiple configurations Component reuse Complexity reduction Reuse of existing MACs Separated control logic Frame Error Rate (FER) / Channel Busy Fraction (CBF) switching metric [1] Puschmann et al., A Component-based Approach for Constructing Flexible Link-Layer Protocols, CROWNCOM 2013, Washington DC, USA. 45
Results Gain: 46
Conclusions Cognitive radio technology enables better utilization of unused licensed spectrum Basic idea Allocation of an unused spectrum band Utilization of the allocated band for communication as long as no primary user appears on this band or no interference produced for primary users working around Requirements Fast and accurate spectrum sensing equipment as well as mechanisms Adequate spectrum decision approaches Seamless spectrum mobility techniques Efficient spectrum sharing methods However Lots of new open issues and challenges to solve IEEE 802.22, IEEE 802.11af are standards for cognitive radio networks 47
References Cognitive Radio Technology, Spectrum Sensing, Management, Mobility and Sharing I.F. Akyildiz, W.Y. Lee, M.C. Vuran, S. Mohanty, NeXt Generation/Dynamic Spectrum Access/Cognitive Radio Wireless Networks: A Survey, Computer Networks Journal, 2006 S. Haykin, Cognitive radio: brain-empowered wireless communications, IEEE Journal on Selected Areas in Communications, 2005 H. Arslan Cognitive radio, software defined radio, and adaptive wireless systems, Springer, 2007 J.J. Mitola, Cognitive Radio - An Integrated Agent Architecture for Software Defined Radio, Doctoral thesis, Royal Institute of Technology (KTH), Teleinformatics, ISSN 1403 ISSN 1403 5286. 5286, Stockholm, 2000 Cognitive MAC Protocols A. Chia-Chun Hsu, D. S. L. Weit, C. C. J. Kuo, A cognitive mac protocol using statistical channel allocation for wireless ad-hoc networks, in Proceeding of IEEE Wireless Communications and Networking Conference (WCNC 2007), March 2007 J. Jia, Q. Zhang, X. Shen, Hc-mac: A hardware-constrained cognitive mac for efficient spectrum management, IEEE Journal on Selected Areas In Communications, vol. 26, no. 1, January 2008 L. Ma, X. Han, C. C. Shen, Dynamic open spectrum sharing mac protocol for wireless ad hoc networks, in Proceeding of the first IEEE International Symposium on New Frontiers in Dynamic Spectrum Access Networks (DySPAN 2005), November 2005 C. Cordeiro, K. Challapali, C-mac: A cognitive mac protocol for multi-channel wireless networks, in Proceeding of the 2nd IEEE International Symposium on New Frontiers in Dynamic Spectrum Access Networks (DySPAN 2007), April 2007 L. Le, E. Hossain, Osa-mac: A mac protocol for opportunistic spectrum access in cognitive radio networks, in Proceeding of the IEEE Wireless Communications and Networking Conference (WCNC 2008), April 2008 Y. R. Kondareddy, P. Agrawal, Synchronized mac protocol for multi-hop cognitive radio networks, in Proceeding of IEEE International Conference on Communications (ICC 2008), May 2008 48
References Claudia Cormio, Kaushik R. Chowdhury, A survey on MAC protocols for cognitive radio networks, Ad Hoc Networks journal, 2009 IEEE Standards Activities C. Stevenson, G. Chouinard, L. Zhongding, H. Wendong, S. Shellhammer, W. Caldwell, IEEE 802.22: The first cognitive radio wireless regional area network standard, IEEE Communications Magazine, 2009 IEEE P802.22 / DRAFTv1.0 Draft Standard for Wireless Regional Area Networks Part 22: Cognitive Wireless RAN Medium Access Control (MAC) and Physical Layer (PHY) specifications: Policies and procedures for operation in the TV Bands, April 2008 IEEE 802.11af -2013 Standard for Information Technology - Telecommunications and Information Exchange Between Systems - Local and Metropolitan Area Networks - Specific Requirements - Part 11: Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY) Specifications Amendment 5: TV White Spaces Operation, February 2014 49
Contact Integrated Communication Systems Group Ilmenau University of Technology Prof. Dr.-Ing. habil. Andreas Mitschele-Thiel Dr.-Ing. Oleksandr Artemenko fon: +49 (0)3677 69 2819/2788 fax: +49 (0)3677 69 1226 e-mail: mitsch, oleksandr.artemenko@tu-ilmenau.de Visitors address: Technische Universität Ilmenau Helmholtzplatz 5 Zuse Building, room 1032/1071 D-98693 Ilmenau www.tu-ilmenau.de/ics Integrated Communication Systems Group Ilmenau University of Technology