CN104168592A - Recognition parameter adjustment method based on multi-target artificial physics optimization - Google Patents
Recognition parameter adjustment method based on multi-target artificial physics optimization Download PDFInfo
- Publication number
- CN104168592A CN104168592A CN201410323449.1A CN201410323449A CN104168592A CN 104168592 A CN104168592 A CN 104168592A CN 201410323449 A CN201410323449 A CN 201410323449A CN 104168592 A CN104168592 A CN 104168592A
- Authority
- CN
- China
- Prior art keywords
- particulate
- domination
- order value
- particle
- cognitive
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Landscapes
- Mobile Radio Communication Systems (AREA)
Abstract
The invention relates to a recognition parameter adjustment method based on multi-target artificial physics optimization, which mainly solves the problem that effect of the data adjustment performed by the prior art is poor. The recognition parameter adjustment method comprises steps of (1) adjusting recognition parameters if a wireless channel and recognition user needs change, (2) mapping the recognition parameter codes to artificial physic particles and determining an evaluation function of fitness, (3) initializing population, (4) calculating fitness values, (5) arranging the particles according to the ascending order of the particle order values, (6) rearranging the particles having same orders according to Hamming distances of the particles, distributing an only order value to each particle, (8) calculating the particle mass, (9) calculating the applying force of the particles, (10) updating the particle motion, (11) updating non-dominated particle group, (12) stopping if the maximum time of evolution is achieved and turning to the process (4) if the maximum time of evolution is not achieved. The invention can perform optimization adjustment on the engine parameter of the recognition wireless network.
Description
Technical field
The invention belongs to the cognitive radio networks communications field, relate to a kind of cognitive parameter regulation means of optimizing based on multiple target artificial physics, can be used for the engine parameters of cognitive radio networks to be optimized adjustment.
Background technology
Intelligent being mainly reflected in of cognitive radio networks can be according to the demand of the variation of external wireless environment and cognitive user, adaptive adjustment transformation parameter, optimization system performance.How cognitive parameter being carried out to adaptive optimization adjustment, is the hot issue in cognitive radio networks research.From in essence, cognitive parameter optimization is a multi-objective optimization question, and intelligent optimization algorithm is the efficient algorithm that solves this problem.At present, emerge corresponding achievement in research, as particle cluster algorithm, ant group algorithm, immune clone algorithm etc.But mostly existing intelligent optimization algorithm is that the multi-objective optimization question of cognitive parameter is converted to single goal problem by weighting method to be solved.Determine because weights are more difficult, and algorithm can only obtain the optimal solution in a kind of weights situation at every turn and easily miss some optimal solutions, therefore, arithmetic result need to continue to improve.
Summary of the invention
Problem to be solved by this invention is, overcome the deficiencies in the prior art, a kind of cognitive parameter optimization method based on multiple target artificial physics is provided, improve the effect of optimizing and revising of cognitive parameter, solved prior art the multi-objective optimization question of cognitive parameter is converted to single goal problem solves and the problem of easily missing some optimal solutions by weighting method.
The present invention solves its technical problem and takes following technical scheme to realize:
According to a kind of cognitive parameter regulation means of optimizing based on multiple target artificial physics provided by the invention, it comprises the following steps:
(1) if wireless channel, cognitive user demand change, adjust cognitive parameter;
(2) cognitive parameter coding is mapped as to the particulate of artificial physics, determines fitness function;
(3) initialization of population, sets relevant parameter: establishing evolutionary generation is 0, the population that random initializtion scale is S
, the particulate that is K to each length
be expressed as
; Random initializtion each
be 0 or 1; If the initial velocity of particulate
, gravitational constant G is set, maximum evolutionary generation is
, the set of depositing non-domination disaggregation is
;
(4) calculate fitness value, determine particulate order value: calculate the adaptive value of each particulate under each sub-goal
, retain the arbitrary particulate of domination
fine-grained quantity, be designated as
; Find out non-domination particulate, be stored in non-domination set
in, and be 1 to each non-domination particulate distribution order value, be that all the other particulates distribute order values
;
(5) press particulate order value ascending order to particle alignment;
(6) order is worth to identical particulate, by the descending of hamming distance between particulate, particulate is rearranged, give corresponding order value; If hamming distance is still identical, compose order value to it at random;
(7) each particulate is composed to a unique order value: according to above-mentioned ranking results, each particulate is composed to a unique serial number, be each particulate and distribute one 1 to S(population scale) between natural number conduct
;
(8) calculate particle mass according to following formula:
, wherein:
represent the quality of particulate i in g generation;
(9) calculate the suffered active force of particulate: calculate the active force of suffered other particulate of particulate according to following formula,
Wherein:
for the active force of particulate to particulate i,
Calculate particulate according to following formula suffered with joint efforts:
(10) calculate particle movement and upgrade:
;
Wherein: for inertia weight (
)
it is the stochastic variable of an obedience (0,1) normal distribution;
The position renewal equation of particulate is:
Wherein:
it is the random number of any [0,1] producing;
(11) upgrade the set of non-domination solution: upgrade
collection, recalculates the functional value of each particulate under each sub-goal, retains the particle number of the each particulate of domination
, distributing order value to each non-domination particulate is 1, for all the other particulates distribute order value
; By each non-domination particulate be stored in
concentrated particulate comparison, if cannot mutually arrange, deposits this particulate in
concentrate; If this particulate domination
concentrated particulate, will by domination particulate from
concentrate and delete; If desired the particulate of storage exceedes
the scale of collection, deletes particulate minimum other particle numbers of domination;
(12) if algorithm reaches maximum evolution number of times
, algorithm stops, will
optimal solution mapping output, be cognitive engine parameter optimization result; Otherwise, evolutionary generation
, go to step (4).
It is to take following technical scheme further to realize that the present invention solves its technical problem:
The cognitive parameter coding of aforesaid step (2) is mapped as the particulate of artificial physics, is to adopt modulation system and the transmitting power of binary system to each subcarrier to encode.
Aforesaid fitness function is:
,
Wherein:
for the through-put power of subcarrier
,
for the maximum transmission power of all subcarriers,
for the number of subcarrier;
Wherein:
for the average error rate of sub-channels;
Maximum data throughput
Wherein:
for the number of subcarrier,
be l the corresponding modulation system of subcarrier number,
for maximum modulation system number,
for lowest modulation system number;
.
Aforesaid step (4) particulate order value is defined as follows: be located at
the population that generation generates is
,
in expression, arrange the fine-grained quantity of particulate i, particulate i is defined as in the order value in g generation:
.
Aforesaid step (6) hamming distance is: to each particulate
,
,
be called the gene position of coding; To any two particulates
with
, its hamming distance definition is:
。
The present invention compared with prior art has significant advantage and beneficial effect:
1. the artificial physics optimized algorithm that the present invention adopts, has parameter less, and the advantage of converges faster is applicable to cognitive engine parameter to adjust;
2. the optimal solution set of finding algorithm of the present invention, has avoided the part optimal solution that may miss when multi-objective problem is converted to single goal and solves.
3. the particulate method for measuring similarity based on hamming distance is that multiple target particulate is composed order value, has ensured that evolutionary process searches for and keep the diversity of non-domination disaggregation towards Pareto optimal solution set.
The specific embodiment of the present invention is provided in detail by following examples and accompanying drawing thereof.
Brief description of the drawings
Fig. 1 is flow chart of the present invention;
Fig. 2 is that 1 time parameter optimization of the pattern of the present invention under awgn channel is adjusted result figure;
Fig. 3 is that 2 times parameter optimizations of the pattern of the present invention under awgn channel are adjusted result figure;
Fig. 4 is that the present invention's parameter optimization under the mode 3 under awgn channel is adjusted result figure;
Fig. 5 is that 4 times parameter optimizations of the pattern of the present invention under awgn channel are adjusted result figure.
Embodiment
Below in conjunction with accompanying drawing and preferred embodiment, to according to embodiment provided by the invention, structure, feature and effect thereof, be described in detail as follows.
A kind of cognitive parameter regulation means of optimizing based on multiple target artificial physics as shown in Fig. 1, it comprises the following steps:
(1) if wireless channel, cognitive user demand change, adjust cognitive parameter;
In embodiments of the present invention, by under IEEE802.11a model, manual switchover channel carrys out the dynamic translation of analog wireless channel between the different modes such as No Fading, Flat Fading, AWGN carrys out the variation of analog wireless channel and cognitive user demand.Cognitive user type is divided into four kinds of patterns, and the low transmitting power of pattern 1 preference, as file transfer; The pattern 2 preference error rates are lower, need the application of high reliability, as secure communication; Mode 3 prefers to high data rate, as wideband video communication; Pattern 4 without special preferences, is sought a kind of balance to each target.
(2) by cognitive parameter coding, be mapped as the particulate that artificial physics is optimized, determine fitness function;
In embodiments of the present invention, adopt binary coding to realize modulation system to each subcarrier and the coding of transmitting power.Suppose, with the coding representing subcarrier modulation modes, to use
represent the coding to transmitting power that subcarrier adopts, particulate length
by
with
coding be in series,
.Dynamic channel fading factor is by distributing the random number between [0 1] to realize to each subcarrier; Noise power is initially 0.01mw; Transmitting power has 64 kinds of possibility values, and scope is set to 0-25.2dBm, is spaced apart 0.4dBm(
), therefore need 6 binary codings to realize; Tetra-kinds of the optional BPSK of subcarrier modulation modes, QPSK, 16QAM and 64QAM (
), therefore need 2 binary codings.The number of subcarrier
therefore, coding total length
.
In embodiments of the present invention, fitness function is:
(3) initialization of population, sets relevant parameter: establishing evolutionary generation is 0, and random initializtion scale is
population
, to each length be
=256 particulate
be expressed as
; Random initializtion each
be 0 or 1; If the initial velocity of particulate
, gravitational constant G=2 are set, maximum evolutionary generation is
, the set of depositing non-domination disaggregation is
.
(4) calculate fitness value, determine particulate order value: according to fitness function formula, calculate the adaptive value of each particulate under 3 sub-goals
, retain the arbitrary particulate of domination
fine-grained quantity, be designated as
; Find out non-domination particulate, be stored in non-domination set
in, and be 1 to each non-domination particulate distribution order value, be that all the other particulates distribute order values
;
Wherein: particulate order value is defined as follows: being located at g for the population generating is,
middle domination particulate
fine-grained quantity, particulate is defined as in the order value in generation:
;
(5) press particulate order value ascending order to particle alignment;
(6) order is worth to identical particulate, by the descending of hamming distance between particulate, particulate is rearranged, give corresponding order value; If hamming distance is still identical, compose order value to it at random; Wherein hamming distance is: to each particulate,
,
be called the gene position of coding; To any two particulates
, its hamming distance definition is:
;
(7) each particulate is composed to a unique order value: according to above-mentioned ranking results, each particulate is composed to a unique serial number, be the natural number conduct that each particulate distributes 1 to 100
; The particulate that order value is less is relatively more excellent;
(8) calculate particle mass according to following formula:
, represent the quality of particulate in generation;
(9) calculate the suffered active force of particulate: calculate the active force of suffered other particulate of particulate according to following formula,
Wherein:
for the active force of particulate j to particulate i,
Calculate particulate according to following formula suffered with joint efforts:
,?
；
(10) calculate particle movement and upgrade:
;
Wherein:
it is the stochastic variable of an obedience (0,1) normal distribution;
The position renewal equation of particulate is:
Wherein: the random number that is any [0,1] producing;
(11) upgrade the set of non-domination solution: upgrade
collection, recalculates the functional value of each particulate under each sub-goal, retains the particle number of the each particulate of domination
, distribute order value to be to each non-domination particulate
+ 1, for all the other particulates distribute order value; By each non-domination particulate be stored in
concentrated particulate comparison, if cannot mutually arrange, deposits this particulate in
concentrate; If the particulate in this particulate dominant set, will be arranged
particulate is deleted from concentrating; If desired the particulate of storage exceedes
the scale of collection, deletes particulate minimum other particle numbers of domination;
(12) if algorithm reaches maximum evolution number of times
, algorithm stops, will
optimal solution mapping output, be cognitive engine parameter optimization result; Otherwise, evolutionary generation
, go to step (4), carry out the g time operation, until reach maximum number of run 200.
Effect of the present invention can further illustrate by following experiment:
1. experiment condition:
In WINDOWS XP system, use Matlab2009a to carry out emulation.
2. experiment content and result
Emulation experiment has been verified the adjustment result of cognitive parameter under different radio channel condition.By under IEEE802.11a model, manual switchover channel carrys out the dynamic translation of analog wireless channel between the different modes such as No Fading, Flat Fading, AWGN.Table 1 has been listed under 3 kinds of different channels conditions, and under 4 kinds of different user models, the representational satisfactory solution of part.
Satisfactory solution under table 1 different channels condition
Can find out from the experimental result of table 1, the present invention can be according to the variation optimized transmission parameter adaptively of wireless channel conditions variation and cognitive user service mode.
Fig. 2～Fig. 5 has provided the present invention under awgn channel, and under four kinds of different modes, parameter optimization is adjusted result, can find out, the present invention can be according to cognitive user demand and wireless environment, and self adaptation is adjusted parameter.
Claims (5)
1. a cognitive parameter regulation means of optimizing based on multiple target artificial physics, is characterized in that: it comprises the following steps:
(1) if wireless channel, cognitive user demand change, adjust cognitive parameter;
(2) cognitive parameter coding is mapped as to the particulate of artificial physics, determines fitness function;
(3) initialization of population, sets relevant parameter: establishing evolutionary generation is 0, the population that random initializtion scale is S
, the particulate that is K to each length
be expressed as
; Random initializtion each
be 0 or 1; If the initial velocity of particulate
, gravitational constant G is set, maximum evolutionary generation is
, the set of depositing non-domination disaggregation is
;
(4) calculate fitness value, determine particulate order value: calculate the adaptive value of each particulate under each sub-goal
, retain the arbitrary particulate of domination
fine-grained quantity, be designated as
; Find out non-domination particulate, be stored in non-domination set
in, and be 1 to each non-domination particulate distribution order value, be that all the other particulates distribute order values
;
(5) press particulate order value ascending order to particle alignment;
(6) order is worth to identical particulate, by the descending of hamming distance between particulate, particulate is rearranged, give corresponding order value; If hamming distance is still identical, compose order value to it at random;
(7) each particulate is composed to a unique order value: according to above-mentioned ranking results, each particulate is composed to a unique serial number, be each particulate and distribute one 1 to S(population scale) between natural number conduct
;
(8) calculate particle mass according to following formula:
, wherein:
represent the quality of particulate i in g generation;
(9) calculate the suffered active force of particulate: calculate the active force of suffered other particulate of particulate according to following formula,
Wherein:
for the active force of particulate to particulate i,
Calculate particulate according to following formula suffered with joint efforts:
(10) calculate particle movement and upgrade:
;
Wherein: for inertia weight (
),
it is the stochastic variable of an obedience (0,1) normal distribution;
The position renewal equation of particulate is:
Wherein:
it is the random number of any [0,1] producing;
(11) upgrade the set of non-domination solution: upgrade
collection, recalculates the functional value of each particulate under each sub-goal, retains the particle number of the each particulate of domination
, distributing order value to each non-domination particulate is 1, for all the other particulates distribute order value
; By each non-domination particulate be stored in
concentrated particulate comparison, if cannot mutually arrange, deposits this particulate in
concentrate; If this particulate domination
concentrated particulate, will by domination particulate from
concentrate and delete; If desired the particulate of storage exceedes
the scale of collection, deletes particulate minimum other particle numbers of domination;
(12) if algorithm reaches maximum evolution number of times
, algorithm stops, will
optimal solution mapping output, be cognitive engine parameter optimization result; Otherwise, evolutionary generation
, go to step (4).
2. according to the cognitive parameter regulation means of optimizing based on multiple target artificial physics described in claim 1, it is characterized in that: described step (2) is mapped as cognitive parameter coding the particulate of artificial physics, is to adopt modulation system and the transmitting power of binary system to each subcarrier to encode.
3. according to the cognitive parameter regulation means of optimizing based on multiple target artificial physics described in claim 1, it is characterized in that: described fitness function is:
,
，
Wherein:
for the through-put power of subcarrier (
),
for the maximum transmission power of all subcarriers,
for the number of subcarrier;
Wherein:
for the average error rate of L sub-channels;
Maximum data throughput
Wherein: the number that L is subcarrier,
be
the corresponding modulation system of individual subcarrier number,
for maximum modulation system number,
for lowest modulation system number;
.
4. according to the cognition network parameter optimization method of adjustment described in claim 1, it is characterized in that: described step (4) particulate order value is defined as follows: being located at g for the population generating is
the quantity of all particulate i of middle domination particulate i, particulate is defined as in the order value in g generation:
.
5. according to the cognitive parameter regulation means of optimizing based on multiple target artificial physics described in claim 1, it is characterized in that: described step (6) hamming distance is: to each particulate
, be called the gene position of coding; To any two particulates
, its hamming distance definition is:
.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410323449.1A CN104168592A (en) | 2014-07-09 | 2014-07-09 | Recognition parameter adjustment method based on multi-target artificial physics optimization |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410323449.1A CN104168592A (en) | 2014-07-09 | 2014-07-09 | Recognition parameter adjustment method based on multi-target artificial physics optimization |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104168592A true CN104168592A (en) | 2014-11-26 |
Family
ID=51912184
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410323449.1A Pending CN104168592A (en) | 2014-07-09 | 2014-07-09 | Recognition parameter adjustment method based on multi-target artificial physics optimization |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104168592A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106254008A (en) * | 2016-10-09 | 2016-12-21 | 哈尔滨工程大学 | A kind of frequency spectrum sensing method of energy acquisition green cognitive radio |
| CN106257849A (en) * | 2016-09-13 | 2016-12-28 | 哈尔滨工程大学 | Frequency spectrum sensing method based on multi-target quantum Lampyridea search mechanisms |
| CN107634811A (en) * | 2017-09-27 | 2018-01-26 | 天津工业大学 | A kind of cognition Internet of Things frequency spectrum detecting method based on mimicry physics multiple-objection optimization |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080010044A1 (en) * | 2006-07-05 | 2008-01-10 | Ruetsch Gregory R | Using interval techniques to solve a parametric multi-objective optimization problem |
| CN103249179A (en) * | 2013-04-25 | 2013-08-14 | 中山大学 | Multi-objective mother foraging algorithm based optimization method for relay node deployment in wireless sensor network |
-
2014
- 2014-07-09 CN CN201410323449.1A patent/CN104168592A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080010044A1 (en) * | 2006-07-05 | 2008-01-10 | Ruetsch Gregory R | Using interval techniques to solve a parametric multi-objective optimization problem |
| CN103249179A (en) * | 2013-04-25 | 2013-08-14 | 中山大学 | Multi-objective mother foraging algorithm based optimization method for relay node deployment in wireless sensor network |
Non-Patent Citations (4)
| Title |
|---|
| YAN WANG: "A multi-objective artificial physics optimization algorithm based on ranks of individuals", 《SOFT COMPUT》 * |
| 于洋: "基于二进制混沌粒子群算法的认知决策引擎", 《哈尔滨工业大学学报》 * |
| 柴争义: "《基于免疫优化的认知无线网络频谱决策与资源分配》", 《中国博士学位论文全文数据库》 * |
| 谢丽萍: "《A Hybrid Vector Artificial Physics Optimization Algorithm》", 《IEEE》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106257849A (en) * | 2016-09-13 | 2016-12-28 | 哈尔滨工程大学 | Frequency spectrum sensing method based on multi-target quantum Lampyridea search mechanisms |
| CN106257849B (en) * | 2016-09-13 | 2019-05-17 | 哈尔滨工程大学 | Frequency spectrum sensing method based on multi-target quantum firefly search mechanisms |
| CN106254008A (en) * | 2016-10-09 | 2016-12-21 | 哈尔滨工程大学 | A kind of frequency spectrum sensing method of energy acquisition green cognitive radio |
| CN106254008B (en) * | 2016-10-09 | 2019-06-11 | 哈尔滨工程大学 | A kind of frequency spectrum sensing method of energy acquisition green cognitive radio |
| CN107634811A (en) * | 2017-09-27 | 2018-01-26 | 天津工业大学 | A kind of cognition Internet of Things frequency spectrum detecting method based on mimicry physics multiple-objection optimization |
| CN107634811B (en) * | 2017-09-27 | 2021-03-09 | 天津工业大学 | Simulated physical multi-objective optimization-based cognitive Internet of things spectrum detection method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103096415B (en) | Route optimizing device and method catering to perceive wireless mesh network | |
| CN109905918A (en) | A kind of NOMA honeycomb car networking dynamic resource scheduling method based on efficiency | |
| CN103096485B (en) | A kind of method of multi-user's multiple-input, multiple-output frequency selection scheduling in LTE system | |
| CN103916355B (en) | Distribution method for sub carriers in cognitive OFDM network | |
| CN104168592A (en) | Recognition parameter adjustment method based on multi-target artificial physics optimization | |
| CN115174396B (en) | Low-carbon energy management and control communication network service management method based on digital twinning | |
| CN103781115B (en) | Distributed base station buffer replacing method based on transmission cost in a kind of cellular network | |
| CN111182511B (en) | AGA-based NOMA resource allocation method in mMTC scene | |
| CN101820671A (en) | Particle swarm algorithm-based distributed power distributing method used for OFDMA system | |
| CN110213766A (en) | Based on the online Poewr control method of Lyapunov frame in collection of energy secure transmission system | |
| CN103781166B (en) | Mobile terminal power distribution method in heterogeneous wireless network cooperative communication system | |
| CN110418360A (en) | Wirelessly taking can network multi-user subcarrier bit combined distributing method | |
| CN111314960B (en) | Social awareness-based collaborative caching method in fog wireless access network | |
| CN111212438A (en) | Resource allocation method of wireless energy-carrying communication technology | |
| Chiumento et al. | Impact of CSI feedback strategies on LTE downlink and reinforcement learning solutions for optimal allocation | |
| CN111556511B (en) | Partial opportunistic interference alignment method based on intelligent edge cache | |
| CN107682934A (en) | A kind of adaptive resource improves allocative decision in OFDM multi-user systems | |
| CN106559491B (en) | A kind of cloud business uplink dispatch method based on the optimization of improved Liapunov | |
| CN104168572A (en) | Artificial physics optimization cognitive radio network spectrum distribution method | |
| CN104768161A (en) | Wireless spectrum distributing method based on chaos search quantum genetic algorithm | |
| CN106850643B (en) | A kind of radio transmitting method of the scalable video real time flow medium of high energy efficiency | |
| CN116341679A (en) | Design method of federal edge learning scheduling strategy with high aging | |
| CN116249162A (en) | Collaborative caching method based on deep reinforcement learning in vehicle-mounted edge network | |
| CN106714262A (en) | Adaptive fuzzy neural topology control algorithm for wireless sensor networks | |
| CN114615296A (en) | Joint optimization method for access control and transmission scheduling of nodes in Internet of vehicles |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20141126 |
|
| WD01 | Invention patent application deemed withdrawn after publication |