CN103780317B - Double-threshold cooperative spectrum sensing method based on degree of belief - Google Patents

Abstract

The invention discloses a trust-based dual-threshold cooperative spectrum sensing method, which takes into account the reliability and low load of the cognitive radio system. The method includes: the system preferentially utilizes cognitive nodes that meet the dual-threshold requirements to participate in cooperative sensing, When the number of cognitive nodes meeting the double-threshold requirements is insufficient, increase the cognitive nodes that meet the requirements of the trust parameter to participate in collaborative sensing. The fusion center stores the detection records of the cognitive nodes, and uses them to set fusion weights for local detection results. The sensing parameters required to be transmitted by the method are reduced, and the occupied channel bandwidth is reduced. At the same time, the detection performance of this method is further improved due to the reduction of unreliable cognitive nodes. In addition, compared with most methods that only improve a certain performance index of the system, this system can adapt to different types of wireless services by adjusting the value of n _t , which has certain flexibility.

Description

A trust-based dual-threshold cooperative spectrum sensing method

technical field

The invention relates to the technical field of computer wireless communication, in particular to a trust-based double-threshold cooperative spectrum sensing method.

Background technique

Computer wireless communication and personal communication have developed rapidly in the past few decades, and various system structures, protocol standards and network architectures have emerged in an endless stream. With the rapid growth of wireless communication services, the contradiction between the increasing spectrum demand and limited spectrum resources has become one of the main bottlenecks restricting the development of wireless communication. In addition to the allocation of very few ISM open frequency bands, most of the current frequency usage policies in the world adopt a license system, that is, authorized users use the specified frequency bands, and non-authorized users are not allowed to use them. And studies have shown that the actual utilization of licensed spectrum is low. Therefore, no matter how the spectrum utilization rate is improved technically, as long as the authorized user temporarily does not use the specified frequency band, the frequency resources allocated to the user will be wasted. In order to make full use of the allocated frequency resources in such a dynamic environment, cognitive radio technology emerges as the times require. It can effectively utilize spectrum holes under the premise of causing only limited interference to existing licensed users. The purpose of spectrum sensing is to quickly discover spectrum holes that appear dynamically in the frequency domain, time domain, and region, so that cognitive users can use the spectrum opportunistically. At the same time, in order not to cause interference to the primary user, the cognitive user needs to be able to quickly perceive the reappearance of the primary user during the communication process using the spectrum hole, and perform spectrum switching in time to free up the channel for the primary user or continue to use it. The original frequency band, but it is necessary to adjust the transmission power or change the modulation method to avoid interference. Spectrum sensing is mainly a technology of the physical layer, which is the premise of spectrum management, spectrum sharing and spectrum mobility management.

The existing spectrum sensing methods mainly include: matched filter detection, energy detection and cyclostationary feature detection. Among them, the matched filter detection needs to know the relevant information (such as modulation mode, pulse waveform, etc.) Extract the characteristic information of the signal; in contrast, energy detection has the advantages of simple implementation and no need for additional information of the system. However, in the actual wireless propagation environment, due to unfavorable factors such as fading and noise of a single detection user in non-cooperative sensing, it is easy to cause deviations between the detection results and the actual situation. In order to further improve the detection accuracy without excessively increasing the complexity of the equipment, people have proposed the research of cooperative spectrum sensing technology. Using the mutual cooperation between cognitive users can effectively eliminate the influence of shadow fading and reduce the detection requirements of a single cognitive user. Collaborative sensing collects perception data from multiple distributed cognitive users, and through the mutual cooperation among multiple cognitive users in different geographical locations, it weakens the negative impact of external adverse factors on a single cognitive user, and ultimately improves cognition Detection performance of radio systems.

Existing cooperative spectrum sensing methods generally focus on a certain performance index of the system, and cannot take into account various performance requirements. For example, although a system that requires all cognitive nodes to participate in cooperation can improve the perceived reliability of the system, the load of the system often causes channel congestion, and the perceived delay is often large, which cannot be performed well in limited channel bandwidth and real-time communication. get applied. Another example is that although a system that requires some cognitive nodes to participate in cooperation can reduce the perception overhead, the detection performance is often not ideal. Many existing methods select some cognitive nodes that participate in cooperation based on double thresholds. If the double threshold is not set properly, it will There are too many or too few cognitive nodes participating in the collaboration. When there are too many cognitive nodes, the load of the system will not be greatly improved; when there are too few cognitive nodes, the system performance will be poor or even cause perception failure. If the weighted fusion method is used to further improve the perceptual reliability, the transmission fusion weight parameters will further affect the channel congestion and perceptual delay. Although some methods do not use double thresholds to select cognitive nodes (such as using trust parameters), verifying whether each cognitive node is reliable through trust parameters and then deciding whether to participate in cooperation will increase the complexity of the system and lead to perception time. delay increase. And the present invention can well solve the above problems.

Contents of the invention

The purpose of the present invention is to propose a cooperative spectrum detection method that takes into account the reliability and low load of the cognitive radio system in view of the existing deficiency of only improving a certain performance index of the system. The method is based on the following content: 1. Verification of local detection results based on adaptive double thresholds and trust parameters; 2. The fusion center sets fusion weights for the cognitive nodes based on the detection records of the cognitive nodes, and the weight parameters are stored in the fusion center to avoid The system overhead caused by the transmission is eliminated.

The technical solution adopted by the present invention to solve its technical problems is: the system of the present invention preferentially utilizes the cognitive nodes that meet the double-threshold requirements to participate in cooperative sensing, and when the number of cognitive nodes that meet the dual-threshold requirements is insufficient, increase the number of cognitive nodes that meet the requirements of the trust degree parameter Cognitive nodes participate in collaborative perception. The fusion center stores the detection records of the cognitive nodes, and sets fusion weights for the local detection results, and weights the fusion to obtain the global detection results.

On the basis of the traditional dual-threshold cooperative spectrum sensing method, the present invention makes the following improvements: 1. For the selection of reliable cognitive nodes, the present invention not only relies on self-adaptive double-thresholds, but also introduces Confidence parameter. Because the traditional double-threshold is greatly affected by noise in the threshold setting, once the setting is inappropriate, it will lead to too many or too few cognitive nodes participating in the cooperation. In the present invention, by increasing the trust degree parameter, when the number of cognitive nodes participating in the collaboration does not reach a preset value, the cognitive nodes located between the double thresholds verify the reliability according to the trust degree parameter. If the trust parameter requirements are met, you can participate in the collaboration. This makes up for the lack of a small number of cognitive nodes.

Method flow:

Step 1: Set the dual threshold of the cognitive node; by constructing the effective utilization function U _i (λ _i ) of CR _i and solving its maximum value, that is, the optimal single threshold λ _opt,i , and setting the adaptive dual threshold on this basis Thresholds λ _1,i and λ _2,i ;

Step 2: Cognitive nodes independently perform local energy detection, if T _i <λ _1,i , decide d _i =0; if T _i >λ _2,i , decide d _i =1, and pass d _i to The fusion center participates in collaborative perception. If λ _1,i ≤T _i ≤λ _2,i , no judgment will be made temporarily;

Step 3: If n _c ≥ [N/n _t ], discard all nodes in λ _1,i ≤ T _i ≤ λ _2,i in the above step 2; where n _c means sending local detection in the above step 2 The result is the number of cognitive nodes to the fusion center, N represents the total number of cognitive nodes, n _t is an adjustable parameter (n _t ≥ 1), [] represents the largest integer; if n _c <[N/n _t ], fusion The center broadcasts uncertain information. At this time, the cognitive nodes satisfying λ _1,i ≤T _i ≤λ _2,i adopt a single threshold decision, that is, according to T _i ≤λ _opt,i or T _i >λ _opt,i , the decision d _i =0 or d _i =1, and conduct further verification according to the trust degree parameter; if the verification d _i is credible, transfer d _i to the fusion center; otherwise, discard d _i ;

Step 4: Cognitive node weight setting; in each round of sensing process, the fusion center will store the uploaded local detection results d _i and system global results D, and set the weight of cognitive nodes accordingly;

Step 5: The fusion center performs weighted fusion, and broadcasts the global detection result 0 or 1 to all cognitive nodes; if the fusion center does not receive any local detection results, it broadcasts information missing signals, and the system immediately starts the next round of spectrum sensing ;

Step 6: In order to take into account the two indicators of system accuracy and efficiency, n _t can be adjusted appropriately; when n _t increases, the probability of participating in the collaboration of cognitive nodes satisfying λ _1,i ≤T _i ≤λ _2,i decreases is small, the amount of data transmitted by the system will decrease, but the detection accuracy will decrease; when n _t decreases, the probability of cognitive nodes that satisfy λ _1,i ≤ T _i ≤ λ _2,i to participate in cooperation increases, and the data transmitted by the system The amount will increase, and the detection accuracy will be improved; the system can properly adjust _nt according to different communication services.

Beneficial effect:

1. The reason why the present invention does not select cognitive nodes based solely on the trust parameter is that the verification of the cognitive node by the trust parameter is more complicated than double thresholds, and requires more time, thereby reducing the perception efficiency of the system.

2. For the selection of the fusion weight, the present invention does not use the signal-to-noise ratio setting in most methods, but uses the cognitive node decision records stored in the fusion center. In this way, the channel congestion caused by the transmission of weight coefficients of each cognitive node is avoided. The fusion center only sets fusion weights based on the latest rounds of detection records, which not only ensures the accuracy and timeliness of weight setting, but also reduces the storage cost of the fusion center.

3. The present invention can adapt to different types of wireless services by adjusting the parameter n _t , and has certain flexibility.

Description of drawings

Fig. 1 is a block diagram of an energy detector of the present invention.

Fig. 2 is a diagram of a cooperative spectrum sensing model of the present invention.

Fig. 3 is a flow chart of the method of the present invention.

detailed description

The present invention will be further described in detail below in conjunction with the accompanying drawings.

As shown in Figure 2, the method of the present invention comprises the following steps:

Step 1: Set the dual threshold of the cognitive node; by constructing the effective utilization function U _i (λ _i ) of CR _i and solving its maximum value, that is, the optimal single threshold λ _opt,i , and setting the adaptive dual threshold on this basis Thresholds λ _1,i and λ _2,i .

Step 2: Cognitive nodes perform local energy detection independently; if T _i <λ _1,i , decide d _i =0; if T _i >λ _2,i , decide d _i =1, and pass d _i to The fusion center participates in collaborative sensing; if λ _1,i ≤T _i ≤λ _2,i , no judgment will be made temporarily.

Step 3: If n _c ≥ [N/n _t ], discard all nodes in λ _1,i ≤ T _i ≤ λ _2,i in step 2; where n _c means sending local detection results to The number of cognitive nodes in the fusion center, N represents the total number of cognitive nodes, n _t is an adjustable parameter (n _t ≥ 1), [] represents the largest integer; if n _c <[N/n _t ], the fusion center broadcasts Uncertain information, at this time, the cognitive nodes satisfying λ _1,i ≤T _i ≤λ _2,i adopt a single-threshold decision, that is, according to T _i ≤λ _opt,i or T _i >λ _opt,i , the decision d _i = 0 or d _i =1, and conduct further verification according to the trust parameter; if it is verified that d _i is credible, transfer d _i to the fusion center; otherwise, discard d _i .

Step 4: Cognitive node weight setting; in each round of sensing, the fusion center will store the uploaded local detection results d _i and system global results D, and set the weights of cognitive nodes.

Step 5: The fusion center performs weighted fusion, and broadcasts the global detection result 0 or 1 to all cognitive nodes; if the fusion center does not receive any local detection results, it broadcasts information missing signals, and the system immediately starts the next round of spectrum sensing .

Step 6: In order to take into account the two indicators of system accuracy and efficiency, n _t can be adjusted appropriately; when n _t increases, the probability of participating in the collaboration of cognitive nodes satisfying λ _1,i ≤T _i ≤λ _2,i decreases is small, the amount of data transmitted by the system will decrease, but the detection accuracy will decrease; when n _t decreases, the probability of cognitive nodes that satisfy λ _1,i ≤ T _i ≤ λ _2,i to participate in cooperation increases, and the data transmitted by the system The amount will increase and the detection accuracy will increase. Therefore, the system can properly adjust _nt according to different communication services.

To sum up, the above is only a preferred embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Within the technical scope disclosed in the invention, any easily conceivable changes or substitutions shall be covered in the technical scope disclosed in the present invention. Therefore, the protection scope of the present invention should be determined by the protection scope of the claims.

Claims (1)

1. a double-threshold cooperative spectrum sensing method based on degree of belief, it is characterised in that Described method comprises the steps:

Step 1: the double threshold of cognitive nodes is set；By structure CR_iEffectively utilize function U_i(λ_i) and solve its maximum, i.e. optimum simple gate limit λ_opt,i, and arrange on this basis certainly Adapt to double threshold λ_1,iAnd λ_2,i；

Step 2: cognitive nodes carries out local energy detection independently, if T_i<λ_1,i, then adjudicate d_i=0；If T_i>λ_2,i, then d is adjudicated_i=1, by d_iReach fusion center and participate in collaborative sensing, if λ_1,i≤T_i≤λ_2,i, wouldn't adjudicate；

Step 3: if n_c≥[N/n_t], then give up and all in above-mentioned steps 2 be in λ_1,i≤T_i≤λ_2,i In node；Wherein, n_cRepresent and above-mentioned steps 2 sends local testing result to fusion center Cognitive nodes number, N represent cognitive nodes sum, n_tIt is adjustable parameter, n_t>=1, [] table Show and take its maximum integer；If n_c<[N/n_t], fusion center broadcast uncertain information, now Meet λ_1,i≤T_i≤λ_2,iCognitive nodes use single threshold judgement, i.e. according to T_i<λ_opt,iOr T_i>λ_opt,i, adjudicate d_i=0 or d_i=1, and make checking further by degree of belief parameter；If checking d_iIt is believable, by d_iReach fusion center；Otherwise, d is given up_i；

Step 4: cognitive nodes weight is arranged；Take turns in perception each, fusion center meeting Local testing result d that storage is uploaded_iWith system global outcome D, and cognitive nodes is set with this Weight；

Step 5: fusion center is weighted merging, and to the broadcast overall situation inspection of all cognitive nodes Survey result 0 or 1；If fusion center does not receives any local testing result, then broadcast letter Breath deleted signal, now system starts next round frequency spectrum perception immediately；

Step 6: in order to take into account accuracy and the efficiency two indices of system, adjusts n_t；Work as n_t During increase, meet λ_1,i≤T_i≤λ_2,iCognitive nodes participate in cooperation probability reduce, system transfers Data volume will reduce, but detection degree of accuracy reduce；Work as n_tDuring reduction, meet λ_1,i≤T_i≤λ_2,i Cognitive nodes participate in the probability of cooperation and increase, the data volume of system transfers will increase, and improve inspection Survey degree of accuracy；System can adjust n according to the difference of communication service_t, adapt to different types of Wireless traffic；

Wherein CR_iRepresent i-th cognitive nodes；U_i(λ_i) represent i-th cognitive nodes CR_i? Simple gate limit λ_iEffectively utilize function；T_iRepresent the energy statistics amount of i-th cognitive nodes；

Described method, when cognitive nodes number is less, introduces degree of belief parameter；

Described method is by degree of enhancing trust parameter, when the cognitive nodes number participating in cooperation does not reaches During to preset value, the cognitive nodes between double threshold is reliable according to degree of belief Verification Property；If meeting degree of belief parameter request, participate in cooperation；

System Priority in described method utilizes and meets the cognitive nodes that double threshold requires and participate in association Make perception, when meeting the cognitive nodes number deficiency of double threshold requirement, increase and meet degree of belief The cognitive nodes of parameter request participates in collaborative sensing；Fusion center stores the detection of cognitive nodes Record, and fusion weight is set as local detection result, and Weighted Fusion obtains overall situation inspection Survey result；Described fusion center only arranges fusion power according to up-to-date some detection records taken turns Weight；

Described method is a kind of cooperation frequency taking into account cognitive radio system reliability and low-load Spectrum detection method；Described method is based on following content: 1) based on self adaptation double threshold and letter Appoint the checking of the local testing result of degree parameter；2) fusion center is according to the detection of cognitive nodes Being recorded as it and arrange fusion weight, weight parameter is stored in fusion center.