CN114726402A - Anonymous frequency hopping sequence design method suitable for multi-antenna cognitive wireless network - Google Patents

Abstract

The invention belongs to the technical field of communication, and particularly relates to an asynchronous frequency hopping sequence design method suitable for a multi-antenna cognitive wireless network. The invention aims to design an anonymous frequency hopping sequence generation method suitable for control information interaction for a distributed control cognitive radio network with asynchronous clock and heterogeneous antennas, ensure that two cognitive nodes with different local available channel sets and different antenna numbers can realize frequency hopping convergence on all public available channels of the two cognitive nodes under any frequency hopping starting time difference, provide frequency hopping convergence capacity backward compatible with a single-antenna cognitive node for a multi-antenna cognitive node, and realize the effect of being superior to a theoretical upper limit O (N) under the condition that the number of the local available channels is far less than the total number N of accessible channels of a cognitive radio network²) The MTTR value of the wireless sensor network, so that the anonymous control information interaction performance of any two antennas and the available channel heterogeneous cognitive node is effectively improved.

Description

Anonymous frequency hopping sequence design method suitable for multi-antenna cognitive wireless network

Technical Field

The invention belongs to the technical field of communication, and particularly relates to a control information interaction mechanism design based on frequency hopping, which is suitable for a wireless communication network.

Background

Cognitive wireless networks are typically made up of a set of network communication nodes that do not have legally licensed spectrum resources. On the premise of ensuring that authorized users do not interfere with legal communication, the cognitive radio network nodes need to flexibly access to a proper authorized communication frequency band to complete the communication function among the nodes. In a cognitive wireless network based on distributed control, different physical locations of different cognitive nodes and different spectrum sensing capabilities of the different cognitive nodes cause different sets of available communication bands (or channels) sensed locally by the different cognitive nodes, thereby causing difficulty in selecting the communication bands among the different cognitive nodes. Therefore, each cognitive node needs to quickly find channels available for neighboring nodes among all locally available channels based on a proper frequency hopping mechanism, so that convergence is achieved, and an efficient and reliable frequency band basis is provided for interaction of various types of control information including frequency spectrum sensing results, clock information, node numbering, handshake negotiation and the like.

Specifically, in the arrangement of R_ACognitive node A of root antenna and configured with R_BIn the process of realizing frequency hopping convergence among cognitive node Bs of a root antenna, at least one antenna of a node A and at least one antenna configured by the node B can establish communication and interaction control information only when the at least one antenna and the at least one antenna are switched to the same channel in the same time slot, wherein R_ANot less than 1 and R_BNot less than 1. Therefore, a frequency hopping sequence set composed of R frequency hopping sequences needs to be generated for each cognitive node with R antennas to respectively guide the frequency hopping process of the R antennas, so as to improve the control information interaction performance between adjacent nodes as much as possible. Generally, the performance parameters for measuring the goodness of a frequency hopping sequence set generation method include:

aggregation Degree (DoR), which is the total number of channels that can be aggregated by any two cognitive nodes according to the frequency hopping sequence set generated by the method. Generally, the maximization of the convergence can effectively improve the anti-interference performance of the control information interaction of the cognitive radio network.

A Maximum aggregation time interval (MTTR for short), which is a Maximum time interval from when any two cognitive nodes start frequency hopping to when they realize aggregation for the first time when the frequency hopping sequence set generated by the method is adopted by any two cognitive nodes. When the MTTR value is smaller, the longest time delay for realizing aggregation of the two cognitive nodes is smaller, and the control information interaction performance between the two cognitive nodes is better.

Since different wireless communication devices are usually configured with different numbers of antennas at present, it is necessary to design a frequency hopping sequence generation algorithm for two cognitive nodes with different numbers of antennas, which can ensure convergence. Particularly, when the frequency hopping sequence generation method adopted by the cognitive node has the backward compatibility (that is, the cognitive node supports a multi-antenna cognitive node and a single-antenna cognitive node to realize frequency hopping convergence on a common available channel), the flexibility and the application range of the cognitive node with any two antennas in a heterogeneous structure to realize frequency hopping convergence are effectively guaranteed.

In addition, the existing frequency hopping sequence aggregation method generally adopts two modes, namely anonymity and non-anonymity. The non-anonymous generation method needs unique addressing (such as 48-bit MAC address) of each cognitive node and proper expansion of each symbol in the addressing to generate the frequency hopping sequence set of the node, and the anonymous generation method does not need to use addressing information of the cognitive node to generate the frequency hopping sequence set of the node at all. Because the MTTR of the frequency hopping sequence set generated based on the non-anonymous mode is correspondingly increased along with the increase of the scale of network nodes and the increase of the addressing length required by the nodes, and the frequency hopping sequence set is easy to be subjected to security defects such as malicious node addressing analysis, monitoring and attack, the frequency hopping sequence set generated based on the non-anonymous mode is generally safer, and better frequency hopping convergence performance can be obtained under the condition that the scale of the cognitive wireless network is increased.

However, although the existing frequency hopping convergence algorithm can support any two cognitive nodes to realize anonymous convergence on any N different channels, the MTTR of the algorithm is O (N) when the number N of convergence channels is increased²) Is increased, resulting in poor frequency hopping convergence performance thereof. To compensate for this deficiency, each cognitive node a may pass N, which is only locally perceived_aThe frequency hopping mode on the available channels avoids the waste of time resources caused by accessing the local unavailable channels, thereby the number N of the local available channels_aFar less than cognitive wireless network canThe theoretical upper limit value of the realization of the total number N of the input channels is better than that of MTTR (MTTR) and is O (N)²) Anonymous frequency hopping convergence performance.

Disclosure of Invention

The invention aims to design an anonymous frequency hopping sequence generation method suitable for control information interaction for a distributed control cognitive radio network with clock asynchronism and antenna isomerism, ensure that two cognitive nodes with different local available channel sets and different antenna number can realize frequency hopping convergence on all public available channels under the condition of any frequency hopping starting time difference, provide frequency hopping convergence capability backwards compatible with a single-antenna cognitive node for a multi-antenna cognitive node, and realize the effect of being superior to a theoretical upper limit value O (N is far less than the total number N of accessible channels of a cognitive radio network) under the condition that the number of the local available channels is far less than the total number N of the accessible channels of the cognitive radio network²) The MTTR value of the wireless sensor network, so that the anonymous control information interaction performance of any two antennas and the available channel heterogeneous cognitive node is effectively improved.

For the purpose of illustrating and understanding the technical solutions of the present invention, the basic concepts and principles involved in the present invention will be briefly described as follows:

definitions 1 if set Z_nA subset of k elements a ═ {0,1, …, n-1}, a ═ a { (a)₀,a₁,…,a_k-1Satisfies the condition that for any non-zero integer d ∈ Z_nEach having at least one ordered pair of elements (a)_i,a_j) Satisfies a_i∈A，a_je.A and d ═ a_i-a_jmodulo n, then the set A is called an (n, k) -relaxation cycle difference set or simply (n, k) -DS, where Z is_nRepresenting the set of all integers modulo n.

Definition 2. for a set of k elements

The execution distance is r epsilon [0, n-1]Can result in a set of k elements, i.e.

For any n ≧ 2, (n, k) -DS always exists, and two corollaries hold as follows:

inference 1 if a set of k elements

Is (n, k) -DS, then its rotating set

Is also an (n, k) -DS.

Inference 2 for (n, k) -DS A

r_j∈[0,n-1]This is true.

Definition 3. if a Mk element set

Can be divided into M mutually disjoint (n, k) -DSs, the set U is then referred to as an M-dimensional disjoint (n, k) -difference set combination or simply as an (M, n, k) -UDDS.

Inference 3. if a set of Mk elements

Is an (M, n, k) -UDDS, then the set ROT (U, i)

Also constitute a (M, n, k) -UDDS.

Definitions 4 if a (3,15,5) -UDDSU^(I)Can be divided into 3 mutually disjoint (15,5) -DSs

And

and satisfy

r∈[1,14]Then the UDDS is referred to as a first class (3,15,5) -UDDS.

Definition 5. for a given (15,5) -DS that can be partitioned into 3 disjoint groups

And

class I (3,15,5) -UDDSU^(I)If there is a (15,5) -DS that can be partitioned into 3 mutually disjoint groups

And

and satisfies the conditions

(3,15,5) -UDDS

Then the UDDS is called an AND U^(I)Associated second class (3,15,5) -UDDSs. Further, for any two different ones of U and U^(I)Associated second class (3,15,5) -UDDS

And

to say, they need to satisfy

For example, because

So that the system is composed of 3 mutually exclusive (15,5) -DSUs₀＝{0,8,9,10,11}, U ₁1,2,4,7,12 and U₂(3,15,5) -UDDSU consisting of {3,5,6,13,14} is not a first class of (3,15,5) -UDDS. On the other hand, the data transmission system is composed of 3 mutually exclusive (15,5) -DSs

And

constructed (3,15,5) -UDDS

Satisfy the requirement of

r∈[1,14]Thus constituting a first class (3,15,5) -UDDS. Furthermore, with the first class (3,15,5) -UDDS

The associated 12 second class (3,15,5) -UDDS can be represented as

Wherein

Can be divided into 3 mutually disjoint (15,5) -DSs

And

can be divided into 3 mutually disjoint (15,5) -DSs

And

can be divided into 3 mutually disjoint (15,5) -DSs

And

can be divided into 3 mutually disjoint (15,5) -DSs

And

can be divided into 3 mutually disjoint (15,5) -DSs

And

can be divided into 3 mutually disjoint (15,5) -DSs

And

can be divided into 3 mutually disjoint (15,5) -DSs

And

can be divided into 3 mutually disjoint (15,5) -DSs

And

can be divided into 3 mutually disjoint (15,5) -DSs

And

can be scribedDivided into 3 mutually disjoint (15,5) -DSs

And

can be divided into 3 mutually disjoint (15,5) -DSs

And

can be divided into 3 mutually disjoint (15,5) -DSs

And

the technical scheme of the invention is as follows:

for a cognitive wireless network with the total number of the accessible channels being N, each cognitive wireless network is provided with any N_aAn available channel

And R_aCognitive node CU of root antenna_aWherein

R_aNot less than 1 and N_a≥3R_aCan be based on the following steps for R thereof_aThe root antenna generates different periodic frequency hopping sequences:

s1, cognizing node CU_aR of (A) to (B)_aThe root antennas are numbered 0,1,2, …, R in sequence_a-1。

S2, cognitive node CU_aN of (A)_aThe available channels are divided into R satisfying the following conditions_aOne channel group:

each channel group i e 0, p-1]For a set V comprising G channels_a,i＝{v_a,iG,v_a,iG+1,…,v_a,(i+1)G-1Where p is N_amodulo R_aAnd

and each channel group j ∈ [ p, R_a-1]For a set V comprising H channels_a,j＝{v_a,p(G-H)+jH,v_{a,p(G-H)+jH+1},…,v_{a,p(G-H)+(j+1)H-1}Therein of

S3, initializing cognitive node CU_aThe antenna number of (1) is r ═ 0.

S4, initializing the periodic frequency hopping sequence of the antenna r as

S5, initializing frequency hopping sequence RS_a,rThe frame number of (1) is i-0.

S6, initializing set W_r,i＝V_a,r\{V_a,r[i]In which V is_a,r[i]Representative set V_a,rThe i +1 th channel number in (1).

S7, initializing frequency hopping sequence RS_a,rThe subframe number j of frame i in (b) is 0.

S8, if (| V)_a,r1) cannot be divisible by 2 and

then set V is initialized_r,i,j＝{V_a,r[i],W_r,i[0],W_r,i[|V_a,r|-2]}; otherwise, set V is initialized_r,i,j＝{V_a,r[i],W_r,i[|V_a,r|-3-2j],W_r,i[|V_a,r|-2-2j]}. Here W_r,i[j]Representative set W_r,iThe j +1 th channel number in (1).

S9, collecting the V_r,i,jIs rearranged in order from small to large so that V_r,i,j[0]<V_r,i,j[1]<V_r,i,j[2]And is provided with u_a,0＝V_r,i,j[0]，u_a,1＝V_r,i,j[1]And u and_a,2＝V_r,i,j[2]。

s10, adopting the following 8 steps based on 3 available channels u_a,0,u_a,1And u_a,2Generate a length of

Frequency hopping sequence T of one time slot_a,r,i,j：

S10.1, numbering each available channel u_a,hRepresented as a binary sequence u_a,h[0]u_a,h[1]…u_a,h[l₁-1]Wherein

h∈[0,2]And u and_a,h[0]and u_a,h[l₁-1]Representing the highest and lowest weighted bits in the binary sequence, respectively.

S10.2, find u_a,0[L_a]＝u_a,1[L_a]≠u_a,2[L_a]Or u_a,0[L_a]≠u_a,1[L_a]＝u_a,2[L_a]Smallest integer of L_aAnd mixing L_aExpressed as an eleven-ary sequence L_a[0]L_a[1]…L_a[l₂-1]Wherein L is_a∈[0,l₁-1]And

s10.3, if u_a,0[L_a]＝u_a,1[L_a]≠u_a,2[L_a]Then further find u_a,0[M_a]≠u_a,1[M_a]Minimum integer of true M_aAnd combining M_aExpressed as an eleven-ary sequence M_a[0]M_a[1]…M_a[l₂-1]Then generate a 2l₂+1 element sequence D_a＝{11,L_a[0],L_a[1],…,L_a[l₂-1],M_a[0],M_a[1],…,M_a[l₂-1]}; if u is_a,0[L_a]≠u_a,1[L_a]＝u_a,2[L_a]Then further find u_a,1[M_a]≠u_a,2[M_a]Minimum integer of true M_aAnd combining M_aExpressed as an eleven-ary sequence M_a[0]M_a[1]…M_a[l₂-1]Then generate a 2l₂+1 element sequence D_a＝{11,M_a[0],M_a[1],…,M_a[l₂-1],L_a[0],L_a[1],…,L_a[l₂-1]In which M is_a∈[0,l₁-1]。

S10.4 based on the first class (3,15,5) -UDDS

An available channel u is generated as follows_a,0,u_a,1And u_a,2Up-hopped 15-slot periodic hopping sequence S_a,*：

In each 15-slot period, the hopping sequence S_a,*Need to be in each time slot

Internally switching to an available channel u_a,hWhere h is [0,2 ]]。

S10.5Based on and

associated each second class (3,15,5) -UDDS

Wherein g ∈ [0,11 ]]Generating a channel u in 3 available channels by using the following method_a,0,u_a,1And u_a,2Up-hopped 15-slot periodic frequency hopping sequence S_a,g：

In each 15-slot period, the hopping sequence S_a,gNeed to be in each time slot

Internally handing over to channel u_a,hWhere g ∈ [0,11 ]]And h is [0,2 ]]。

S10.6, initialization

And k is 0.

S10.7, update

Wherein D_a[k]Represents sequence D_aThe (k + 1) th element and the symbol | | | in (b) represent the concatenation of two hopping sequences.

S10.8, if k<2l₂Then k +1 is updated and S10.7 is returned; otherwise, go to S11.

S11, update RS_a,r＝RS_a,r||T_a,r,i,j。

S12, if

Then update j to j +1 and return to S8; otherwise, go to S13.

S13, if i<|V_a,rI-1, then i +1 is updated and S6 is returned; otherwise, go to S14.

S14, if r<R_a-1, then update r ═ r +1 and return to S4; otherwise, ending the algorithm execution and outputting the hopping sequenceColumn(s) of

The invention has the beneficial effects that:

according to definition 4 and definition 5, when two are each provided with N_aAn available channel

And N_bAn available channel

Cognitive node CU_aAnd CU_bIn case of starting frequency hopping at the same time, they are based on the first class (3,15,5) -UDDS

Respectively generated 30-slot periodic frequency hopping sequence S_a,*||S_a,*And S_b,*||S_b,*Can realize (u)_a,0,u_b,0),(u_a,1,u_b,1) And (u)_a,2,u_b,2) Convergence between equal 3 channel pairs, and CU_aAnd CU_bBased on the second class (3,15,5) -UDDS

And

wherein g is₁≠g₂,g₁∈[0,10]And g₂∈[0,10]Respectively generated 15-slot periodic frequency hopping sequences

And

can realize (u)_a,0,u_b,1),(u_a,0,u_b,2),(u_a,1,u_b,0),(u_a,1,u_b,2),(u_a,2,u_b,0) And (u)_a,2,u_b,1) And 6 channel pairs. On the other hand, when two cognitive nodes CU_aAnd CU_bIn the case of starting the frequency hopping at any different time, they are based on the first class (3,15,5) -UDDS

Respectively generated 30-slot periodic frequency hopping sequence S_a,*||S_a,*And S_b,*||S_b,*All 9 channel pairs described above can be implemented (i.e., (u)_a,i,u_b,j)

) Convergence between them. Therefore, CU_aAnd CU_bRespectively generated 45 time slot periodic frequency hopping sequence

And

the frequency hopping convergence among all the 9 channel pairs can be realized under any frequency hopping starting time difference, wherein g₁≠g₂And g₁,g₂∈[0,11]. Based on this fact, when R_aAntenna cognitive node CU_aAnd R_bAntenna cognitive node CU_bHaving at least one common available channel v_a,x＝v_b,yIn time, no matter how big the difference of the frequency hopping starting time of the two cognitive nodes is, the former is accessed to the channel v_a,xOf (a) an antenna r_a,xWith the latter accessing the channel v_b,yR of_b,yCan always be at

Frequency hopping convergence on the common available channel is achieved within a time slot.

Therefore, in a cognitive wireless network with the total number of accessible channels being N, if one has N_aR of locally available channel_aAntenna cognitive node CU_aAnd one has N_bR of a locally available channel_bAntenna cognitive node CU_bThere are at least 1 common available channel in between, then the invention can support them to realize frequency hopping convergence on all C common available channels under any frequency hopping starting time difference, wherein C is [1, min { N ]_a,N_b}]And ensure that their maximum time interval from the start of frequency hopping to the first time hopping convergence (i.e., MTTR) is no greater than

And a time slot. Since the theoretical upper limit of MTTR is

Therefore, the invention is particularly suitable for cognitive nodes CU_aNumber of local available channels N_aAnd cognitive node CU_bNumber of local available channels N_bThe theoretical upper limit value of the method is O (N) better than that of MTTR (maximum transmission rate) under the condition that the number of the accessible channels is far less than the total number N of the cognitive radio network²) The convergence performance of the existing anonymous frequency hopping algorithm.

On the other hand, there are anonymous frequency hopping convergence algorithms of the same type, i.e. documents [1 ]]MTP and document [2 ]]EE and document [3 ]]The theoretical upper limit values of MTTR for algorithms 3-5 are

And

thus, at R_aAnd R_bUnder given conditions, the method can always ensure that the total number N of the accessible channels of the cognitive wireless network is large enough or two cognitive nodes CU_aAnd CU_bNumber of locally available channels N_aAnd N_bLess than that of the document [1 ] is obtained without much difference]“Z.Gu,H.Pu,Q.-S.Hua,and F.C.MTP of M.lau, "Improved reconstructed surfaces of audios for terrestrial cognitive radio networks," in Proc. IEEE INFOCOM,2015, pp.154-162 ", document [2]EE and document [3 ] of "Y. -C.Chang, C. -S.Chang, and J. -P.Sheu," An enhanced fast multi-radio reproducing in a heterologous audio in heterologous radio network, "IEEE trans. Cogn. Commun. Net., vol.4, No.4, pp.847-859, December 2018"]' Dunmejun, research on heterogeneous cognitive radio network frequency hopping blind convergence technology, university of electronic technology Master ' graduate thesis, 6 months in 2021 ' MTTR theoretical upper limit of algorithm 3-5.

In addition, the invention can also fully support two cognitive nodes with any number of antennas to realize frequency hopping convergence, and has the capability that the existing multi-antenna anonymous frequency hopping algorithm (for example, documents [4] "L.Yu, H.Liu, Y.Leung, X.Chu, and Z.Lin," Multiple radios for fast rendezvous in coherent radio networks, "IEEE Trans. Mobile Comp., vol.14, No.9, pp.1917-1931, September 2015" and [5] "J. -P.Sheuu and J." J.Lin, "A multi-radio rendezvous in coherent radio networks," IEEE Trans. Mobile. Comp., 17, vol.9, No. 1980-1990, and September 8 ") can not provide a multi-antenna anonymous frequency hopping algorithm (i.e., the capability of the September 201compatible with the common single-antenna) and can be realized on the single-antenna convergent node, therefore, the flexibility and the application range of the cognitive wireless network with the antenna heterogeneous characteristic for realizing frequency hopping convergence can be effectively guaranteed.

Finally, the frequency hopping sequence generated for each cognitive node is completely irrelevant to the addressing of the node, and the frequency hopping convergence capability based on an anonymous mode can be provided for any two cognitive nodes with at least 1 public available channel, so that the safety of control information interaction is effectively improved.

Drawings

Fig. 1 shows a cognitive node CU in which the total number N of accessible channels in a cognitive radio network is 512_aAnd CU _b3 and 1 antennas are configured, their available channel number sets are 58,232,138,313,421,25,97,70,145,113,37,79,180,326,245 and 25,77,138,325,421,490, respectively, and a cognitive node CU_aPrior to cognitive node CU_bUnder the condition that the frequency hopping starts at exactly 5 time slots, the cognitive node CU_a3-antenna frequency hopping sequence set and cognitive node CU generated based on method_bThe convergence diagram of the single-antenna frequency hopping sequence generated based on the invention. Each double arrow in the figure represents a cognitive node CU_aAnd CU_bOne frequency hopping aggregation on a common available channel.

FIG. 2 shows that when N accessible channels of the cognitive radio network are numbered 0,1, …, N-1, a single-antenna cognitive node CU_aIs numbered 1,2, …,0.5N, and a single antenna cognitive node CU_bAre numbered 0.1N,0.1N +1, …,0.6N, they are according to the invention, document [1 ]]And document [2 ]]And document [3 ]]And (3) a simulation comparison graph of the maximum convergence time interval (namely MTTR) obtained by the four anonymous single-antenna frequency hopping convergence algorithms and the change of the total number N of the accessible channels of the cognitive radio network.

FIG. 3 shows a cognitive node CU configured with 2 antennas when N accessible channels of the cognitive wireless network are numbered 0,1, …, N-1_aIs 0.3N,0.3N +1, …,0.7N, and a cognitive node CU is configured with 5 antennas_bAre numbered 0.4N,0.4N +1, …,0.6N, they are in accordance with the invention and the document [2 ]]And (3) a simulation comparison graph of the maximum convergence time interval (namely MTTR) obtained by the two anonymous antenna heterogeneous frequency hopping convergence algorithms along with the total number N of the accessible channels of the cognitive radio network.

Detailed Description

The technical scheme of the invention is described in detail in the following with reference to the accompanying drawings and embodiments:

examples

Given that the total number of accessible channels in the cognitive wireless network is N-512, the cognitive node CU_aConfiguring 3 antennas and having available channel number set {58,232,138,313,421,25,97,70,145,113,37,79,180,326,245}, and adopting the method to realize the node CU by the following steps_aGenerate its hop sequence set containing 3 hop sequences:

s1, cognizing node CU _a3 antennasNumbered 0,1,2 in sequence.

S2, cognitive node CU_aIs divided into 3 channel groups V_a,0＝{58,232,138,313,421}， V _a,125,97,70,145,113, and V _a,237,79,180,326,245, such that antennas 0,1, and 2 are in channel group V, respectively_a,0,V_a,1And V_a,2And (4) up frequency hopping.

S3, initializing cognitive node CU_aThe antenna number of (1) is r ═ 0.

S4, initializing the periodic frequency hopping sequence with the antenna r being 0

S5, initializing frequency hopping sequence RS_a,0The frame number of (1) is i-0.

S6, initializing set W_r,i＝W_0,0＝V_a,0\{V_a,0[0]Where V is 232,138,313,421_a,0[0]Representative set V_a,0The 1 st channel number in (1), i.e., 58.

S7, initializing frequency hopping sequence RS_a,0The subframe number j of frame 0 is 0.

S8, if (| V)_a,r1) cannot be divisible by 2 and

then set V is initialized_r,i,j＝{V_a,r[i],W_r,i[0],W_r,i[|V_a,r|-2]}; otherwise, set V is initialized_r,i,j＝{V_a,r[i],W_r,i[|V_a,r|-3-2j],W_r,i[|V_a,r|-2-2j]}。

At this time, | V _a,04 can be divided by 2 and

thus initializing set V_0,0,0＝{V_a,0[0],W_0,0[|V_a,0|-3],W_0,0[|V_a,0|-2]}＝{58,313,421}。

S9, mixingSet V_0,0,0Is rearranged in order from small to large so that V_0,0,0[0]<V_0,0,0[1]<V_0,0,0[2]Thereby obtaining an updated V_0,0,0Is {58,313,421} and set u_a,0＝V_0,0,0[0]＝58，u_a,1＝V_0,0,0[1]＝313,u_a,2＝V_0,0,0[2]＝421。

S10, adopting the following steps based on 3 available channels u_a,0,u_a,1And u_a,2Generate a length of

Frequency hopping sequence T of one time slot_a,0,0,0：

S10.1, numbering each available channel u_a,hExpressed as a length of

Binary sequence u of_a,h[0]u_a,h[1]…u_a,h[8]Where h is [0,2 ]]And u is_a,h[0]And u_a,h[8]Representing the highest and lowest weighted bits in the binary sequence, respectively.

At this time, u is due to_a,0＝58，u _a,1313, and u _a,2421 so that there is u_a,0＝000111010，u_a,1100111001, and u_a,2＝110100101。

S10.2, find u_a,0[L_a]＝u_a,1[L_a]≠u_a,2[L_a]Or u_a,0[L_a]≠u_a,1[L_a]＝u_a,2[L_a]Smallest integer of L_aAnd mixing L_aExpressed as an eleven-ary sequence L_a[0]L_a[1]…L_a[l₂-1]Wherein L is_a∈[0,l₁-1]And

at this time, u is 0_a,0[0]≠u_a,1[0]＝u_a,2[0]1, therefore, L is set_aIs equal to 0, and_aexpressed as a length of

Eleven system sequence L_a[0]＝0。

S10.3, if u_a,0[L_a]＝u_a,1[L_a]≠u_a,2[L_a]Then further find u_a,0[M_a]≠u_a,1[M_a]Minimum integer of true M_aAnd M is_aExpressed as an eleven-ary sequence M_a[0]M_a[1]…M_a[l₂-1]Then generate a 2l₂+1 element sequence D_a＝{11,L_a[0],L_a[1],…,L_a[l₂-1],M_a[0],M_a[1],…,M_a[l₂-1]}; if u is_a,0[L_a]≠u_a,1[L_a]＝u_a,2[L_a]Then further find u_a,1[M_a]≠u_a,2[M_a]Minimum integer of true M_aWill M_aExpressed as an eleven-ary sequence M_a[0]M_a[1]…M_a[l₂-1]Then generate a 2l₂+1 element sequence D_a＝{11,M_a[0],M_a[1],…,M_a[l₂-1],L_a[0],L_a[1],…,L_a[l₂-1]In which M is_a∈[0,l₁-1]。

At this time, u is due to_a,0[L_a]≠u_a,1[L_a]＝u_a,2[L_a]And l ₂1, therefore further find u_a,0[M_a]≠u_a,1[M_a]Minimum integer of true M _a1, mixing M_aExpressed as a length l ═ 1₂Eleven sequence M of 1_a[0]1 and generates one 2l₂+ 1-3 element sequence D_a＝{11,1,0}。

S10.4 based on the first class (3,15,5) -UDDS

An available channel u is generated as follows_a,0,u_a,1And u_a,2Up-hopped 15-slot periodic hopping sequence S_a,*：

In each 15-slot period, the hopping sequence S_a,*Need to be in each time slot

Internal switching to available channel u_a,hWhere h is [0,2 ]]。

At this time, since the first class (3,15,5) -UDDS

Can be divided into 3 mutually disjoint (15,5) -DSs

And

thus having S_a,*＝{313,421,313,58,313,58,58,421,421,58,58,421,313,313,421}。

S10.5, based on

Associated each second class (3,15,5) -UDDS

Wherein g ∈ [0,11 ]]Generating a channel u in 3 available channels by using the following method_a,0,u_a,1And u_a,2Up-hopped 15-slot periodic frequency hopping sequence S_a,g：

In each 15-slot period, the hopping sequence S_a,gNeed to be in each time slot

Internally handing over to channel u_a,hWhere g ∈ [0,11 ]]And h is [0,2 ]]。

At this time, according to the first type (3,15,5) -UDDS

Associated 3 second class (3,15,5) -UDDS

And

can be defined as S_a,0＝{58,313,421,421,58,421,313,313,58,421,421,313,58,58,313},S_a,1(58, 313,421, 58,313,58, 421) and S_a,11＝{58,313,58,421,421,313,421,58,58,421,421,313,58,313,313}。

S10.6, initialization

And k is 0.

S10.7, updating T_a,0,0,0＝T_a,0,0,0||S_a,*||S_a,*||S_a,Da[0]＝S_a,*||S_a,*||S_a,11The symbol | | | represents the concatenation of two hopping sequences.

At this time, there is T_a,0,0,0＝{313,421,313,58,313,58,58,421,421,58,58,421,313,313,421,313,421,313,58,313,58,58,421,421,58,58,421,313,313,421,58,313,58,421,421,313,421,58,58,421,421,313,58,313,313}。

S10.8, if k<2l₂Then k +1 is updated and S10.7 is returned; otherwise, go to S11.

At this time, since k is 0<2l₂Thus, update k +1 k 2 and return to S10.7.

S10.7, updating T_a,0,0,0＝T_a,0,0,0||S_a,*||S_a,*||S_a,Da[1]＝S_a,*||S_a,*||S_a,11||S_a,*||S_a,*||S_a,1。

At this time, there is T_a,0,0,0＝{313,421,313,58,313,58,58,421,421,58,58,421,313,313,421,313,421,313,58,313,58,58,421,421,58,58,421,313,313,421,58,313,58,421,421,313,421,58,58,421,421,313,58,313,313,313,421,313,58,313,58,58,421,421,58,58,421,313,313,421,313,421,313,58,313,58,58,421,421,58,58,421,313,313,421,58,313,421,421,58,58,313,58,421,313,421,313,313,58,421}。

S10.8, if k<2l₂Then k +1 is updated and S10.7 is returned; otherwise, go to S11.

In this case, k is 1<2l₂Thus, update k + 12 and return to S10.7.

S10.7, update

At this time, there is T_a,0,0,0＝{313,421,313,58,313,58,58,421,421,58,58,421,313,313,421,313,421,313,58,313,58,58,421,421,58,58,421,313,313,421,58,313,58,421,421,313,421,58,58,421,421,313,58,313,313,313,421,313,58,313,58,58,421,421,58,58,421,313,313,421,313,421,313,58,313,58,58,421,421,58,58,421,313,313,421,58,313,421,421,58,58,313,58,421,313,421,313,313,58,421,313,421,313,58,313,58,58,421,421,58,58,421,313,313,421,313,421,313,58,313,58,58,421,421,58,58,421,313,313,421,58,313,421,421,58,421,313,313,58,421,421,313,58,58,313}。

S10.8, if k<2l₂Then k +1 is updated and S10.7 is returned; otherwise, go to S11.

In this case, k is 2l or 2l₂Thus, go to S11, 2.

S11, update RS_a,0＝RS_a,0||T_a,0,0,0＝T_a,0,0,0。

S12, if

Then j +1 is updated and S8 is returned; otherwise, go to S13.

At this time, since

Thus updating j +1 to 1 and returning to S8.

S8, if (| V)_a,r1) cannot be divisible by 2 and

then set V is initialized_r,i,j＝{V_a,r[i],W_r,i[0],W_r,i[|V_a,r|-2]}; otherwise, set V is initialized_r,i,j＝{V_a,r[i],W_r,i[|V_a,r|-3-2j],W_r,i[|V_a,r|-2-2j]}。

At this time, | V _a,04 can be divided by 2, thus initializing the set V_0,0,1＝{V_a,0[0],W_0,0[|V_a,0|-3-2],W_0,0[|V_a,0|-2-2]}＝{58,232,138}。

S9, collecting the V_0,0,1Is rearranged in order from small to large so that V_0,0,1[0]<V_0,0,1[1]<V_0,0,1[2]Thereby obtaining an updated V_0,0,1Is {58,138,232} and set u_a,0＝V_0,0,1[0]＝58，u_a,1＝V_0,0,1[1]＝138,u_a,2＝V_0,0,1[2]＝232。

S10, adopting the following steps based on 3 available channels u_a,0,u_a,1And u_a,2Generate a length of

Frequency hopping sequence T of one time slot_0,0,1：

S10.1, numbering each available channel u_a,hExpressed as a length of

Binary sequence u of_a,h[0]u_a,h[1]…u_a,h[8]Where h is [0,2 ]]And u and_a,h[0]and u_a,h[8]Representing the highest and lowest weighted bits in the binary sequence, respectively.

At this time, u is due to_a,0＝58，u _a,1138, and u _a,2232, thus has u_a,0＝000111010，u_a,1010001010, and u_a,2＝011101000。

S10.2, find u_a,0[L_a]＝u_a,1[L_a]≠u_a,2[L_a]Or u_a,0[L_a]≠u_a,1[L_a]＝u_a,2[L_a]Smallest integer of L_aAnd mixing L_aExpressed as an eleven-ary sequence L_a[0]L_a[1]…L_a[l₂-1]Wherein L is_a∈[0,l₁-1]And

at this time, since 0 ═ u_a,0[1]≠u_a,1[1]＝u_a,2[1]1, therefore, L is set_a1, and mixing L_aIs expressed as a length of

Eleven system sequence L_a[0]＝1。

S10.3, if u_a,0[L_a]＝u_a,1[L_a]≠u_a,2[L_a]Then further find u_a,0[M_a]≠u_a,1[M_a]Smallest integer M of true_aAnd combining M_aExpressed as an eleven-ary sequence M_a[0]M_a[1]…M_a[l₂-1]Then generate a 2l₂+1 element sequence D_a＝{11,L_a[0],L_a[1],…,L_a[l₂-1],M_a[0],M_a[1],…,M_a[l₂-1]}; if u is_a,0[L_a]≠u_a,1[L_a]＝u_a,2[L_a]Then further find u_a,1[M_a]≠u_a,2[M_a]Minimum integer of true M_aA 1, M_aExpressed as an eleven-ary sequence M_a[0]M_a[1]…M_a[l₂-1]Then generate a 2l₂+1 element sequence D_a＝{11,M_a[0],M_a[1],…,M_a[l₂-1],L_a[0],L_a[1],…,L_a[l₂-1]In which M is_a∈[0,l₁-1]。

At this time, u is due to_a,0[L_a]≠u_a,1[L_a]＝u_a,2[L_a]And l ₂1, thus further found so that u_a,0[M_a]≠u_a,1[M_a]Minimum integer of true M_aWhen M is equal to 2, mixing_a2 is expressed as a length l₂Eleven sequence M of 1_a[0]2 and generates one 2l₂+ 1-3 element sequence D_a＝{11,2,1}。

S10.4 based on the first class (3,15,5) -UDDS

An available channel u is generated as follows_a,0,u_a,1And u_a,2Up-hopped 15-slot periodic hopping sequence S_a,*：

In each 15-slot period, the hopping sequence S_a,*Need to be in each time slot

Internally switching to an available channel u_a,hWhere h is [0,2 ]]。

At this time, since the first type (3,15,5) -UDDS

Can be divided into 3 mutually disjoint (15,5) -DSs

And

thus having S_a,*＝{138,232,138,58,138,58,58,232,232,58,58,232,138,138,232}。

S10.5, based on

Associated each second class (3,15,5) -UDDS

Wherein g ∈ [0,11 ]]Generating a channel u in 3 available channels by using the following method_a,0,u_a,1And u_a,2Up-hopped 15-slot periodic frequency hopping sequence S_a,g：

In each 15-slot period, the hopping sequence S_a,gNeed to be in each time slot

Internally handing over to channel u_a,hWhere g ∈ [0,11 ]]And h is [0,2 ]]。

At this time, according to the first type (3,15,5) -UDDS

Associated 3 second class (3,15,5) -UDDS

And

can be defined as S_a,1＝{58,138,232,232,58,58,138,58,232,138,232,138,138,58,232}，S_a,258,138, 58,138,232,138,232,58,232, 138,58,232,58 and S_a,11＝{58,138,58,232,232,138,232,58,58,232,232,138,58,138,138}。

S10.6, initialization

And k is 0.

S10.7, update

At this time, there is T_a,0,0,1＝{138,232,138,58,138,58,58,232,232,58,58,232,138,138,232,138,232,138,58,138,58,58,232,232,58,58,232,138,138,232,58,138,58,232,232,138,232,58,58,232,232,138,58,138,138}。

S10.8, if k<2l₂Then k +1 is updated and S10.7 is returned; otherwise, go to S11.

At this time, since k is 0<2l₂Thus, update k +1 k 2 and return to S10.7.

S10.7, update

At this time, there is T_a,0,0,1＝{138,232,138,58,138,58,58,232,232,58,58,232,138,138,232,138,232,138,58,138,58,58,232,232,58,58,232,138,138,232,58,138,58,232,232,138,232,58,58,232,232,138,58,138,138,138,232,138,58,138,58,58,232,232,58,58,232,138,138,232,138,232,138,58,138,58,58,232,232,58,58,232,138,138,232,58,138,138,58,138,232,138,232,58,232,232,138,58,232,58}。

S10.8, if k<2l₂Then k +1 is updated and S10.7 is returned; otherwise, go to S11.

At this time, since k is 1<2l₂Thus, update k + 12 and return to S10.7.

S10.7, update

At this time, there is T_a,0,0,1＝{138,232,138,58,138,58,58,232,232,58,58,232,138,138,232,138,232,138,58,138,58,58,232,232,58,58,232,138,138,232,58,138,58,232,232,138,232,58,58,232,232,138,58,138,138,138,232,138,58,138,58,58,232,232,58,58,232,138,138,232,138,232,138,58,138,58,58,232,232,58,58,232,138,138,232,58,138,138,58,138,232,138,232,58,232,232,138,58,232,58,138,232,138,58,138,58,58,232,232,58,58,232,138,138,232,138,232,138,58,138,58,58,232,232,58,58,232,138,138,232,58,138,232,232,58,58,138,58,232,138,232,138,138,58,232}。

S10.8, if k<2l₂Then k +1 is updated and S10.7 is returned; otherwise, go to S11.

In this case, k is 2l or 2l₂Thus, go to S11, 2.

S11, update RS_a,0＝RS_a,0||T_a,0,0,1＝T_a,0,0,0||T_a,0,0,1。

S12, if

Then j +1 is updated and S8 is returned; otherwise, go to S13.

At this time, since

Thus jumping to S13.

S13, if i<|V_a,rI-1, then i +1 is updated and S6 is returned; otherwise, go to S14.

At this time, since i is 0<|V_a,rI-1-4, so i +1 is updated and S6 is returned.

……

By analogy, the cognitive node CU can be finally generated_aOf the hopping sequence RS_a,0,RS_a,1And RS_a,2. FIG. 1 shows that the cognitive node CU_aFrequency hopping sequence RS for antenna 0_a,0First 270 time slots of (1), the frequency hopping sequence RS of antenna_a,1First 270 time slots of, and the hopping sequence RS of antenna 2_a,2The first 270 slots. The cycle lengths of the hopping sequences are all

And a time slot.

Similarly, when the total number N of accessible channels in the cognitive radio network is 512, the cognitive user CU_bFIG. 1 also shows a single-antenna cognitive node CU when 1 antenna is configured and the available channel number set {25,77,138,325,421,490} is available_bFrequency hopping sequence RS generated by adopting the invention_b,0The first 265 time slots of {421,490,421,25,421, … }. The frequency hopping sequence has a period length of

And a time slot.

When cognitive node CU as shown in FIG. 1_aPrior to cognitive node CU_bWhen the frequency hopping starts at exactly 5 time slots, the cognitive node CU_aUsing 3 hopping sequences RS_a,0,RS_a,1And RS_a,2And cognitive node CU_bUsing a single hopping sequence RS_b,0At cognitive node CU_aThe frequency hopping convergence can be achieved on all the commonly available channels, i.e., channels 25,138 and 421, within time slot 0-152, and their time interval from the beginning of the frequency hopping to the first time that the frequency hopping convergence is achieved on channel 421 is 3 time slots. This time interval is much less than

The MTTR per slot is a theoretical upper bound.

The following combined simulation shows that the invention achieves the effects:

as shown in fig. 2, when N accessible channels of the cognitive wireless network are numbered 0,1, …, N-1, the cognitive node CU with single antenna_aIs numbered 1,2, …,0.5N, and a single antenna cognitive node CU_bIs numbered 0.1N,0.1N +1, …,0.6N, it is determined that the local available channel is not availableAccording to the present invention, reference [1 ]]And document [2 ]]And document [3 ]]And (3) a simulation comparison graph of the maximum convergence time interval (namely MTTR) obtained by the four anonymous single-antenna frequency hopping convergence algorithms and the change of the total number N of the accessible channels of the cognitive radio network. It can be seen that the present invention always achieves a lower maximum convergence time interval than the other three anonymous single-antenna frequency hopping convergence algorithms.

As shown in fig. 3, when N accessible channels of the cognitive wireless network are numbered 0,1, …, N-1, a cognitive node CU with 2 antennas is configured_aIs 0.3N,0.3N +1, …,0.7N, and a cognitive node CU is configured with 5 antennas_bAre numbered 0.4N,0.4N +1, …,0.6N, they are in accordance with the invention and the document [2 ]]And (3) a simulation comparison graph of the maximum convergence time interval (namely MTTR) obtained by the two anonymous antenna heterogeneous frequency hopping convergence algorithms along with the total number N of the accessible channels of the cognitive radio network. It can be seen that the invention can always obtain the comparison document [2 ] when the total number N of accessible channels of the cognitive wireless network is more than 600]The anonymous antenna heterogeneous frequency hopping convergence algorithm has a lower maximum convergence time interval.

Claims (1)

1. An anonymous frequency hopping sequence design method suitable for a multi-antenna cognitive wireless network defines the total number of accessible channels of the multi-antenna cognitive wireless network as N, and each cognitive node CU_aIs provided with N_aAn available channel

And R_aRoot antenna, where 0 ≦ v_a,0<v_a,1<v_a,2≤N-1，R_a≥1，N_a≥3R_aCharacterized by being R_aThe method for generating different periodic frequency hopping sequences by the root antenna respectively comprises the following steps:

s1, cognitive node CU_aR of (A) to (B)_aThe root antennas are numbered 0,1,2, …, R in sequence_a-1；

S2, cognizing node CU_aN of (A)_aThe available channels are divided into R satisfying the following conditions_aOne channel group: each timeOne channel group i belongs to [0, p-1 ]]For a set V comprising G channels_a,i＝{v_a,iG,v_a,iG+1,…,v_a,(i+1)G-1Where p is N_amodulo R_aAnd

and each channel group j ∈ [ p, R_a-1]For a set V comprising H channels_a,j＝{v_a,p(G-H)+jH,v_{a,p(G-H)+jH+1},…,v_{a,p(G-H)+(j+1)H-1}Therein of

S3, initializing cognitive node CU_aThe antenna number of (a) is r ═ 0;

s4, initializing the periodic frequency hopping sequence of the antenna r as

S5, initializing frequency hopping sequence RS_a,rThe frame number of (1) is i-0;

s6, initializing set W_r,i＝V_a,r\{V_a,r[i]In which V is_a,r[i]Representative set V_a,rThe i +1 th channel number in (a);

s7, initializing frequency hopping sequence RS_a,rThe subframe number j of the frame i is 0;

s8, if (| V)_a,r1) cannot be divisible by 2 and

then set V is initialized_r,i,j＝{V_a,r[i],W_r,i[0],W_r,i[|V_a,r|-2]}; otherwise, set V is initialized_r,i,j＝{V_a,r[i],W_r,i[|V_a,r|-3-2j],W_r,i[|V_a,r|-2-2j]}，W_r,i[j]Representative set W_r,iThe j +1 th channel number in (a);

s9, collecting the V_r,i,j3 elements ofRearranging from small to large in order so that V_r,i,j[0]<V_r,i,j[1]<V_r,i,j[2]And is provided with u_a,0＝V_r,i,j[0]，u_a,1＝V_r,i,j[1]And u and_a,2＝V_r,i,j[2]；

s10, adopting the following steps based on 3 available channels u_a,0,u_a,1And u_a,2Generate a length of

Frequency hopping sequence T of one time slot_a,r,i,j：

S101, numbering each available channel u_a,hRepresented as a binary sequence u_a,h[0]u_a,h[1]…u_a,h[l₁-1]Wherein

h∈[0,2]And u and_a,h[0]and u_a,h[l₁-1]Respectively representing the bits with the highest weight and the lowest weight in the binary sequence;

s102, finding u_a,0[L_a]＝u_a,1[L_a]≠u_a,2[L_a]Or u_a,0[L_a]≠u_a,1[L_a]＝u_a,2[L_a]Smallest integer of L_aAnd mixing L_aExpressed as an eleven-ary sequence L_a[0]L_a[1]…L_a[l₂-1]Wherein L is_a∈[0,l₁-1]And

s103, if u_a,0[L_a]＝u_a,1[L_a]≠u_a,2[L_a]Then further find u_a,0[M_a]≠u_a,1[M_a]Smallest integer M of true_aAnd M is_aExpressed as an eleven-ary sequence M_a[0]M_a[1]…M_a[l₂-1]Then generate a 2l₂+1 element sequence D_a＝{11,L_a[0],L_a[1],…,L_a[l₂-1],M_a[0],M_a[1],…,M_a[l₂-1]}; if u is_a,0[L_a]≠u_a,1[L_a]＝u_a,2[L_a]Then further find u_a,1[M_a]≠u_a,2[M_a]Minimum integer of true M_aAnd M is_aExpressed as an eleven-ary sequence M_a[0]M_a[1]…M_a[l₂-1]Then generate a 2l₂+1 element sequence D_a＝{11,M_a[0],M_a[1],…,M_a[l₂-1],L_a[0],L_a[1],…,L_a[l₂-1]In which M is_a∈[0,l₁-1]；

S104, UDDS based on first class (3,15,5)

An available channel u is generated as follows_a,0,u_a,1And u_a,2Up-hopped 15-slot periodic hopping sequence S_a,*Of the first class (3,15,5) -UDDS

Is defined as if a (3,15,5) -UDDSU^(I)Can be divided into 3 mutually disjoint (15,5) -DSs

And

and satisfy

Then the UDDS is referred to as a first class (3,15,5) -UDDS:

in each 15-slot period, frequency hoppingSequence S_a,*Need to be in each time slot

Internally switching to an available channel u_a,hWhere h is [0,2 ]]；

S105, based on

Associated each second class (3,15,5) -UDDS

Wherein g ∈ [0,11 ]]Generating a channel u in 3 available channels by using the following method_a,0,u_a,1And u_a,2Up-hopped 15-slot periodic frequency hopping sequence S_a,gClass II (3,15,5) -UDDS

Is defined as if there is one (15,5) -DS which can be divided into 3 mutually disjoint groups

And

and satisfies the conditions

(3,15,5) -UDDS

Then the UDDS is called an AND U^(I)Associated second class (3,15,5) -UDDSs:

in each 15-slot period, the hopping sequence S_a,gNeed to be in each time slot

Internally handing over to channel u_a,hWhere g ∈ [0,11 ]]And h is [0,2 ]]。

S106, initialization

And k is 0;

s107, update

Wherein D_a[k]Represents sequence D_aThe (k + 1) th element and the symbol | | | in the sequence represent the series connection of two frequency hopping sequences;

s108, if k<2l₂Then k +1 is updated and returns to S107; otherwise, jumping to S11;

s11, update RS_a,r＝RS_a,r||T_a,r,i,j；

S12, if

Then j +1 is updated and S8 is returned; otherwise, jumping to S13;

s13, if i<|V_a,rI-1, then i +1 is updated and S6 is returned; otherwise, jumping to S14;

s14, if r<R_a-1, then update r ═ r +1 and return to S4; otherwise, ending the algorithm execution and outputting the hopping sequence

Images