Research On Energy-Efficiency And Security For Cooperative Wireless Sensor Networks

Wireless Sensor Networks are composed of many tiny sensor nodes,which can provide data acquisition,processing and transmission,to perform tasks such as environment monitoring,battlefield reconnaissance and topographic mapping.Due to the limited data acquisition range,processing ability and transmission distance of a single sensor node,it is impossible to use one sensor to accomplish these missions.Therefore,cooperative communication,which exploits the broadcast nature of wireless medium and adopts idle nodes as relays to forward signals,is widely used in wireless sensor networks.This technology is acknowledged as an effective way to enlarge signal coverage and improve communication rate and reliability.However,compared with conventional cooperative communication networks,cooperative wireless sensor networks have their own features.Firstly,in conventional cooperative communication networks,transmission distances are usually large,so transmit power is high enough to make circuit power consumption negligible.While in cooperative wireless sensor networks,transmission distances are much smaller due to the dense distribution of sensors,so transmit power is much lower and circuit power consumption might be comparable.Moreover,circuit power and transmit power are both supplied by energy-limited batteries in sensor nodes.Consequently,it is of great importance to take circuit power consumption into consideration.Secondly,wireless sensor networks provide efficient sensing and computing solutions for various applications,many of which have demands for high information security,for example,battlefield reconnaissance and outer space exploration.In these applications,sensor nodes are usually unprotected which makes them easily spotted or captured.Hence,it is necessary to consider the existence of eavesdroppers and improve system confidentiality through physical-layer security.Aiming at the aforementioned two features,this dissertation studies the energy-efficient and secure transmission schemes in cooperative wireless sensor networks.The main contributions of this dissertation are listed as follows.(1)Optimal relay selection and power allocation schemes are proposed under a total power constraint for cooperative wireless sensor networks assisted by amplify-and-forward and decode-and-forward relay nodes,respectively.The total power constraint contains not only transmit power but also the power dissipated by relays to drive the circuit blocks.Relay ordering is introduced into the proposed schemes so that they can obtain the same optimal solutions as exhaustive search scheme does but with less computation.Moreover,a suboptimal relay selection and power allocation scheme with further computation reduce is also provided for amplify-and-forward cooperative networks,which enjoys almost the same performance as the optimal scheme does.Simulation results verify that when the circuit power consumption is taken into consideration,adopting more relays may degrade the system performance.It is also shown that the proposed partial relay selection and power allocation schemes provide higher received signal-to-noise ratio.(2)The security performance of a decode-and-forward cooperative wireless sensor network is investigated in terms of secrecy outage probability,based on which,transmit power is allocated between both hops of cooperation.The decoding ability of relay nodes is taken into account.Only the relays that can successfully decode the source message form a decoding set.Two transmission schemes are studied,namely,the best relay selection scheme and the all-relay based beamforming scheme.It is revealed that in both schemes,there is a tradeoff between the two ways of improving system security – increasing the number of relays in decoding set and increasing the transmit power of relays.Simulation results show that the theoretical expressions of secrecy outage probability are accurate and the proposed power allocation schemes outperform the traditional equal power allocation scheme.(3)Beamforming schemes are studied for both decode-and-forward and amplify-and-forward cooperative wireless sensor networks to enhance system security.In a decode-and-forward system,the lower bound of ergodic secrecy capacity over the distribution of relay-to-eavesdropper channels is obtained as a Rayleigh quotient and the optimal beamforming solution is readily derived.However,in an amplify-and-forward system,the lower bound is obtained as a product of two correlated Rayleigh quotients which is non-convex.Therefore,Taylor series expansion is used to approximate the product of two Rayleigh quotients to a single Rayleigh quotient,and a suboptimal beamforming solution is then derived.Simulation results show that the practical secrecy capacity of the proposed scheme for decode-and-forward is higher than that of the conventional scheme,while the proposed scheme for amplify-and-forward outperforms both the conventional scheme and the bound maximization scheme.(4)A relay-selection and beamforming framework is proposed for a cooperative wireless sensor network with multiple decode-and-forward relays and an eavesdropper.In the proposed framework,partial relays are selected to forward the source signal in a beamforming manner.The circuit power consumption is considered along with the transmit power.Two relay selection strategies are presented,namely the exponential-complexity exhaustive search and linear-complexity relay ordering.For comparison purposes,the conventional all-relay based beamforming scheme and the best relay selection scheme are also improved by incorporating circuit power.It is verified that adopting more relays may degrade the system security and is shown that the proposed schemes combine the advantage of the all-relay based beamforming scheme in high power range and that of the best-relay selection scheme in low power range,and outperform them both in the medium power range.Additionally,the relay ordering based beamforming scheme achieves almost the same secrecy capacity as the exhaustive-search scheme does,but with a reduced computational complexity.