Energy-Efficient and Trust-aware Cooperation in Cognitive Radio Networks Networks Ning Zhang Department of Electrical and Computer Engineering U i University of Waterloo i f W l 1
Outline Outline • Introduction Introduction • Related works • Motivation and objectives Motivation and objectives • System model • Proposed scheme • Proposed scheme – Trust values – Stackelberg game Stackelberg game • Numerical results • Conclusion and future work • Conclusion and future work 2
Introduction Introduction Cognitive radio network (CRN) • – Spectrum scarcity S i – Cognitive radio [1] primary users (PU) and secondary users (SU) – Spectrum holes Shortcomings of Spectrum Sensing • – Inaccurate (multipath fading, shadowing) ( l i h f di h d i ) – Energy consumption – Dynamic Cooperation • – Cooperative sensing (Cooperation among SUs) p g ( p g ) – Cooperation between PU and SU [1] S. Haykin, Cognitive radio: brain ‐ empowered wireless communications, IEEE Journal on Selected Areas in 3 Communications 23 (2) (2005) 201–220.
Related works Related works In [2], the PU gives its licensed bandwidth for a fraction of the transmission In [2], the PU gives its licensed bandwidth for a fraction of the transmission • time to a subset of secondary transmitters in exchange for cooperation to increase its transmission rate. Zhang et al. in [3] propose a novel cooperative cognitive radio framework, • where the PU selects some appropriate SUs as cooperative relays to increase its transmission rate and revenue. In return, an interval of time is given to the selected secondary relays as a reward. gi en to the selected secondar rela s as a re ard Han et al. in [4] propose a two-phase cooperation scheme in CRN. The PU • transmits its signal in the first phase which is overheard by the SU Then transmits its signal in the first phase, which is overheard by the SU. Then the SU decodes and superimposes the PU’s signal with its own signal using a different power level to broadcast. [2] O. Simeone. et al, “Spectrum leasing to cooperating secondary ad hoc networks,” Selected Areas in Communications, IEEE Journal on , 2008 [3] J. Zhang and Q. Zhang, “Stackelberg game for utility-based cooperative cognitive radio networks,” in Proceedings of the tenth ACM international symposium on Mobile ad hoc networking and computing. ACM, 2009. [4] Y. Han et al, “Cooperative decode-and-forward relaying for secondary spectrum access,” Wireless Communications,IEEE Transactions on, 4 2009.
Motivation and objectives j Security concerns: • All previous works assume that SUs are trustworthy and well behaved which All previous works assume that SUs are trustworthy and well-behaved, which may not be always true in reality. There may exist some dishonest users, even malicious ones in the system, corrupting or disrupting the normal operation of Cooperative CRN. p Energy efficiency concerns: • Since the PU can transmit its data at any time, it may have an incentive to Since the PU can transmit its data at any time, it may have an incentive to cooperate for energy saving due to power limitation. Motivation: Motivation: • We make an effort to guarantee and improve the performance of cooperation and SU selection, considering energy efficiency and security issues. Objectives: • We try to answer the following questions: when to cooperate with whom to cooperate and how to cooperate ? when to cooperate, with whom to cooperate and how to cooperate ? 5
System model I y Primary network: • PU communicates with the base station (BS) in TDMA mode. – The time slot duration for each PU’s transmission is T . – Secondary network: • Ad hoc network Ad hoc network – Cooperation • Three-phase cooperation and Amplify-and-Forward (AF) – (a) Primary transmission (b) Cooperative transmission (c) Secondary transmission 6
System model II y Assumptions: • – Channels are modeled as independent complex Gaussian random Ch l d l d i d d l G i d variables, constant within each slot, but generally varying over the slots. – A common control channel is assumed for exchanging information g g among PU, SU and BS. – There exist a trusted center keeping the trust values. Security threats: • – When a malicious SU is selected to cooperate with the PU, it may When a malicious SU is selected to cooperate with the PU it may misbehave, e.g., it may drop PU’s data ,etc. – A dishonest SU may not obey the cooperation rule during the cooperative transmission to pursue more self-benefits. – Moreover, considering the mobility of SUs, the malicious or dishonest SUs may misbehave at one place then move to other places SUs may misbehave at one place then move to other places. 7
System model III y Notations • the channel gain between the PU and the BS; h pb the channel gain between the PU and SUi ; i h ps the channel gain between SUi and the BS; h h l i b SUi d h BS i h sb the channel gain between SUi and its corresponding receiver. i h s transmit power of PU without cooperation transmit power of PU without cooperation. P P d transmit power of PU with cooperation. P c transmit power of SU in cooperative and secondary transmission 1 i P s the one-sided power spectral density of the independent additive white N 0 Gaussian noise at the both base station and secondary receivers . 1 This rule is enforced in order to ensure that the secondary users spend for cooperation at least the same power that they are willing to spend for their own transmission. 8
Trust values Bayesian framework to evaluate the SU’s behaviours • – Each SU will behave well with probability p, and misbehave with probability (1-p) ˆ – Posteriori probability estimate of the above binary events can be represented as p beta distributions. Trust computation model based on beta function • v – let be the observed number of good behaviours. – let be the observed number of misbehaviours. ( 2 ) v ( k 1 ) ( l 1 ) ˆ ˆ ˆ ( | , ) ( 1 ) f p k l p p ( ( 1 1 ) ) ( ( 1 1 ) ) v v Tr - The trust value of SUi: i 1 Beta function of event after 7 observations of ˆ ( ) E p 2 good behaviors and 1 observation of misbehaviors. v 9
Stackelberg game between PU and SU g g As the PU is a licensed user and has a higher priority on using the spectrum • b band, it has the right to decide the parameters ( i, and Pc) so as to d it h th i ht t d id th t ( i d P ) t maximize its own utility in terms of energy efficiency. The SU decides its transmission power, under the given (i , and Pc ) to • maximize the transmission rate without spending too much energy. Stackelberg game : • Leader: Leader: PU i , and Pc PU i and Pc Follower: SU transmission power 10
Utility function of PU y • Primary user: y Without cooperation, the primary user’s direct transmission rate 2 P h d pb log log ( ( 1 1 ) ) R d R W W 2 N (1) 0 Through cooperation with SUi the achievable cooperative rate [5] Through cooperation with SUi, the achievable cooperative rate 2 i P h a W 2 2 c pb i i i i i i log [ 1 ( , )] R f P h P h c 2 c ps s sb 2 N (2) (2) 0 0 2 2 i i i i P h P h 1 2 2 c ps s sb i i i i ( , ) f P h P h c ps s sb 2 2 N i i i i P h P h N 0 c ps s sb 0 R In order to maintain PU’s transmission rate: R c d [5] J. Laneman, D. Tse, and G. Wornell, “Cooperative diversity in wireless networks: Efficient protocols and outage behavior,” Information Theory, IEEE Transactions on, 2004 11
Utility function of PU y Non-cooperative: • P The PU uses to transmit for a period of . Energy consumption is T P T d d Cooperative: • a i T The PU cooperates with SUi , it uses to transmit for a period of . i P 2 c a a T T c Th The energy consumption is i i i i i P 2 The energy saving for the PU is the ratio between the amount of energy • reduction and the total power dissipation without cooperation: p p p a T i i P T P i a P a d c 2 i 1 i c 1 i i P P i c i d 2 2 P T P d d i denotes the ratio between and (power allocation coefficient ) P P • i c d , PU chooses cooperation with SUi 1 i 1 1 , PU doesn t choose cooperation with SUi PU doesn’t choose cooperation with SUi i Taking the trust value into consideration, • a Expected energy saving Expected energy saving i ( ( 1 1 i i ) ) U U Tr Tr p i 2 12
Recommend
More recommend
