MIMO-aware Cooperative Cognitive Radio Networks Hang Liu
Outline Motivation and Industrial Relevance Project Objectives Approach and Previous Results Future Work Outcome and Impact [2]
Motivation & Relevance Cognitive radio to alleviate spectrum scarcity enables secondary users (SUs) to dynamically access (DSA) and share the spectrum licensed to primary users (PUs). MIMO offers another dimension for sharing Increase throughput by sending multiple independent data streams (spatial multiplexing) Suppress interference from neighboring links (beam-forming) Company and government labs are investigating use of CR and MIMO to improve spectrum utilization [3]
MIMO-Aware Cooperative Cognitive Radio Networks Developing a general MIMO-aware Cooperative Dynamic Spectrum Access framework enables multiple primary and multiple secondary users to cooperate in dynamic spectrum sharing and to exploit the cognitive radio and MIMO techniques to achieve significant gains on spectrum efficiency and physical layer security. Novel centralized and distributed relay selection and resource management algorithms based on game theory for joint optimization of multiple PUs and multiple SUs. Design and investigate cooperative jamming mechanisms to enhance the secrecy capacity of the primary network. Develop cooperative protocols between PUs and SUs Testbed implementation that supports a number of experimental tasks for validating the effectiveness of approaches and methodologies developed in the project. Other outside funding to leverage the research: NSF MIMO-Aware Cooperative Dynamic Spectrum Access. 300K Industry and university collaboration (Technicolor, NYU, GWU) [4]
Dynamic Spectrum Access DSA Operation Modes Interweave DSA: an SU can transmit only on a spectrum band where the PU is not active. Underlay DSA: an SU can transmit on a spectrum band under condition its interference to PUs is limited. Cooperative DSA or Cooperative Cognitive Radio Networks: PUs and SUs cooperate in dynamic spectrum access. PUs improve their performance with the help of SUs, while the SUs can access the PUs spectrum for secondary data transmission. a win-win situation MIMO allows cooperative DSA in temporal, frequency, and spatial dimensions. Current research on MIMO DSA mainly focuses on physical layer. Conventional cooperative cognitive radio schemes are based on Temporal-Only sharing Low efficiency to PUs and SUs Higher layer protocols are needed to fully exploit MIMO and CR capabilities in cooperative DSA. [5]
MIMO in CCRN A Motivating Example of Using MIMO in CCRN PUs select some SUs as the cooperative relay, in return, grant the SUs some opportunities to access the licensed channel for secondary data transmission In phase one, PT broadcasts to SU2 and SU3, and SU1 can transmit to SU2 simultaneously; In phase two, SU2 and SU3 cooperatively relay the primary traffic to PR, and SU3 can transmit to SU4 simultaneously. [6]
System Model One scenario we have studied (with some previous results) A primary link coexists with several secondary MIMO links. PUs are assumed to be legacy devices with a single antenna. How to select the relays? How to schedule transmissions considering MIMO? How relays access the channel fairly? [7]
MIMO-CCRN Framework In Phase One, PT continuously broadcasts its data to the chosen relays. The pairs with SR selected as the relay access the channel in a TDMA fashion. In Phase Two, the selected relays cooperatively forward the primary data to the PR. The pairs with ST selected as the relay access the channel in a TDMA fashion. How to select relays? [8]
Game Theory Analysis Stackelberg Game: Primary link is the leader. By using backward-induction, selecting the best relay set based on the knowledge of the secondary users game. Secondary users are the followers, they conduct a power control game to determine the relaying power. Secondary Power Control Game: Strategy Space: The power each secondary relay used to relay the primary traffic Utility Function: Fairness: we let T k (i), the time allocated to relay k s transmission proportional to k s consumed power for relaying. The utility then is: throughput gain consumed power [9]
Primary Utiltity Optimization The primary link aims at solving the following optimization problem Time allocation Relaying power allocation Relay set selection where Link rate region, Parameters: T i k : time fraction allocated to user i in Phase k P r : the power allocated for relaying by relay r S i : the relay set in Phase i [10]
Algorithm Backward Induction Algorithm at the PU 1. Rp_best = Rdir; 2. For each possible relay set S1, S2 3. Find the Nash Equilibrium of the secondary power control game; 4. Solve the primary utility optimization problem, achieve Rp(S1, S2); 5. if Rp(S1, S2) > Rp_best 6. Rp_best = Rp(S1, S2) 7. end if 8. end for [11]
Average utility of the primary link Performance Evaluation Significant improvement compared to non-mimo CCRN (single antenna) More SUs (larger K) result in higher utility to the primary link [12]
Performance Evaluation Average Utility of the Secondary Links Both primary link and secondary links achieve larger utility than the conventional CCRN scheme; MIMO-CCRN creates better win-win situation [13]
Future Work PI and 2 graduate students for three years For the first year: Develop a general MIMO-aware Cooperative Dynamic Spectrum Access framework, that enables multiple primary and multiple secondary users to cooperate. PT2 PT1 SUr1 SUr2 PR2 PT3 PT1 Spe ctrum PT2 Spe ctrum PT3 Spe ctrum SUr4 SUr3 PR3 PR1 In the real world, there usually exist multiple concurrent PU transmissions and SUs. The PUs will compete with each other in selecting a good SU as partner. Similarly, multiple SUs also compete for the spectrum resource. This many-to-many cooperation/competition with MIMO techniques has not been systemically investigated before. the existing schemes are no longer applicable [14]
Future Work PI and 2 graduate students for three years For the first year: Develop a general MIMO-aware Cooperative Dynamic Spectrum Access framework, that enables multiple primary and multiple secondary users to cooperate. PT2 PT1 SUr1 SUr2 PR2 PT3 PT1 Spe ctrum PT2 Spe ctrum PT3 Spe ctrum SUr4 SUr3 PR3 PR1 Due to the use of MIMO and multiple frequency channels, the primary and secondary transmissions should be intelligently scheduled in temporal, frequency, and spatial domains to exploit channel and user diversities for optimal resource use. [15]
Future Work PI and 2 graduate students for three years For the first year: Develop a general MIMO-aware Cooperative Dynamic Spectrum Access framework, that enables multiple primary and multiple secondary users to cooperate. PT2 PT1 SUr1 SUr2 PR2 PT3 PT1 Spe ctrum PT2 Spe ctrum PT3 Spe ctrum SUr4 SUr3 PR3 PR1 Each entity wants to maximize its own interest, while the overall framework has to ensure system optimization and fairness. To this end, distributed methods are desirable when global knowledge is not available. The transmitter might not always have knowledge of the channel state. The cooperation scheme needs to properly handle this case to enhance transmission capacity. [16]
Future Work PI and 2 graduate students for three years For the first year: Develop a general MIMO-aware Cooperative Dynamic Spectrum Access framework, that enables multiple primary and multiple secondary users to cooperate. PT2 PT1 SUr1 SUr2 PR2 PT3 PT1 Spe ctrum PT2 Spe ctrum PT3 Spe ctrum SUr4 SUr3 PR3 PR1 Each entity wants to maximize its own interest, while the overall framework has to ensure system optimization and fairness. To this end, distributed methods are desirable when global knowledge is not available. The transmitter might not always have knowledge of the channel state information (CSI). The cooperation scheme needs to properly handle this case to enhance transmission capacity. [17]
Future Work Develop novel centralized and distributed algorithms based on bargaining game and matching theory (Stackelberg game not applicable) for many-tomany cooperative relay selection channel sharing PU and SU data transmission resource allocation and scheduling Design practical protocols to enable information dissemination, cooperation, and radio resource management negotiation among multiple MIMOempowered PUs and SUs for joint optimization. Investigated adaptive directional antennas for relay capacity enhancement. [18]
Cooperative Jamming Multiple MIMO helpers are employed to transmit noise Cooperatively jam the eavesdropper s channel No interference to the legitimate receivers (jamming signal is nullified) Enhance secrecy capacity of primary networks MIMO helpers can simultaneously transmit their own data. Primary Transmitter SU2 SU1 Primary Transmission Jamming Eavesdropper Secondary Transmission SU3 Primary Receiver SU4 Secondary jammers enhance PU security capacity Optimization strategy for multiple PUs and multiple helpers. Helper selection Novel beamforming techniques and signal design for cooperative jamming and transmitter artificial interference [19]
Testbed Validation build a cognitive radio network testbed based on software-defined network (SDN) technology provides a platform to validate the proposed algorithms and protocols gain deep insights and valuable feedbacks that can facilitate the further development of our approaches and methodologies. SDN Controller Pica8 Open Switch Pica8 Open Switch USRP/GNURadio [20]
Outcome and Impact A general framework for MIMO-aware cooperative dynamic spectrum access with multiple PUs and SUs. New centralized and distributed algorithms for relay selection and resource sharing New algorithms for cooperative jamming and artificial interference Practical protocols to enable multiple MIMO-empowered PUs and SUs to cooperate for joint optimization. Software implementation and testbed results We expect Significant spectrum efficiency gains Physical layer security enhancement [21]
Thanks Q & A [22]