CN109743735A - A kind of dynamic channel assignment method based on depth enhancing study in satellite communication system - Google Patents
A kind of dynamic channel assignment method based on depth enhancing study in satellite communication system Download PDFInfo
- Publication number
- CN109743735A CN109743735A CN201811550296.9A CN201811550296A CN109743735A CN 109743735 A CN109743735 A CN 109743735A CN 201811550296 A CN201811550296 A CN 201811550296A CN 109743735 A CN109743735 A CN 109743735A
- Authority
- CN
- China
- Prior art keywords
- channel
- user
- signal
- satellite
- value
- 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
- Radio Relay Systems (AREA)
Abstract
The invention discloses the dynamic channel assignment methods based on depth enhancing study in a kind of satellite communication system, are related to satellite communication field.First against GEO multibeam satellite system, constructing includes satellite multi-beam antenna, cell, the scene of channel and user；After certain user's k initiating business request, calculating the user one by one on each channel with Shannon capacity is the service quality C characterizedk.Then judge whether user k meets C in the Shannon capacity that each channel providesk≥CTh, if so, the channel of the distribution is available, carry out business normal transmission；Otherwise, then the channel is unavailable, and service request is by call drop or obstruction and terminates.From meeting Ck≥CThAvailable channel list in dynamically distribute out a channel and give user k, when user k completes specific transactions transmission, abandon occupied channel；When waiting new user request next time, then enter the channel allocation of a new round.System congestion rate is effectively reduced in the present invention, promotes channel utilization, improves satellite communication system load-carrying amount and spectrum efficiency.
Description
Technical field
Dynamic the present invention relates to satellite communication field, based on depth enhancing study in specifically a kind of satellite communication system
Method for channel allocation.
Background technique
The high quality provided to satellite with people and the demand of low rate service are further urgent, and satellite communication system passes through
Satellite-borne multi-beam configuration obtains extensive concern because spectral efficient and terminal size miniaturization can be achieved.Current high throughput is defended
Star is also or high capacity satellite further improves multi-beam and defend all using load flexible on star and multi-beam allocation plan
The availability of frequency spectrum of star system, therefore dynamic channel allocation becomes research hotspot.
Under multibeam satellite system scene, traditional channel assignment scheme is static assignment, i.e., by satellite system
What middle available channel resources were fixed distributes to each wave beam, and cochannel between wave beam is avoided to interfere.But the static assignment can make
It obtains system channel resource to be difficult to efficiently make full use of, this portfolio between wave beam shows even more serious when uneven.Compared to quiet
State channel assignment scheme, dynamic channel allocation schemes can then be adjusted according to beam service volume dynamic, can promote channel resource utilization
Rate.
Existing dynamic channel allocation schemes consider current each beam service volume and user distribution more, and are new service request
The time domain associate feature of dynamic channel allocation between each wave beam is ignored when distributing available channel, i.e. current time is to a new business
The channel resource of distribution can request follow-up business to have an impact when distribution channel resource.This is mainly due to same channel moneys
Source distribution can generate between different beams to be interfered with each other, and then deteriorates the channel quality of service.
Therefore, how to consider that time domain associate feature carries out dynamic channel allocation is current urgent problem to be solved.
Summary of the invention
The present invention proposes a kind of satellite communication system aiming at the problem that above-mentioned time domain associate feature carries out dynamic channel allocation
Dynamic channel assignment method based on depth enhancing study in system, for solving the dynamic channel allocation in multi-beam satellite scene
Problem.
The dynamic channel assignment method based on depth enhancing study, the specific steps are as follows:
Step 1: being directed to GEO multibeam satellite system, building includes satellite multi-beam antenna, cell, channel and user
Scene；
In this scene, multibeam satellite system is configured by satellite multi-beam antenna, coverage area is divided into multiple only
Vertical cell, and distribute available channel and provide service for the accessing user of each cell.
User's collection of all services in satellite system is combined into { 1,2 ... k..., K }；Set of available channels be 1,2 ...
m...,M}；
Step 2: calculating the user one by one on each channel using Shannon capacity as table after certain user's k initiating business request
The service quality C of signk；
Specific step is as follows:
Step 201 calculates receiving end signal y of the user k on each available channelk；
Receiving end signal ykAre as follows:
Wherein,Indicate the useful signal of user k,It indicates to remove outside user k, remaining
The common-channel interference of all users；σkIndicate the thermal noise that user k receiving antenna introduces；
hk,kFor the signal for being sent to user k, after being received by user k, the channel coefficients of the useful signal of formation, wkIt indicates
System is provided the channel result that service distributes by user k, is the vector of 1*M；Distribute channel then element wkIt is 1, is otherwise 0.
OperatorIndicate Hadamard product；skIt is sent to the signal of user k for satellite, is the column vector of M*1.hk,iTo be sent to user i
Signal, the interference channel coefficient that is formed after being received by user k, wiExpression system is provided the channel that service distributes by user i
As a result, being the column vector of M*1；Distribute channel then element wiIt is 1, is otherwise 0.siThe signal of user i is sent to for satellite, i's takes
Value is 1~k.
The channel that all users are distributed forms channel occupancy matrix W=[w1,w2,…,wK]；
Step 202 is directed to user k, according to the receiving end signal y of each available channelkIn useful signal and channel is dry altogether
Signal is disturbed, the corresponding available signal power D of each channel is calculated separatelykAnd common-channel interference signal power Ik,
User k received available signal power D on certain channelkCalculation formula is as follows:
Dk=| hk,k|2·diag{wk}·[diag{wk}]H
Common-channel interference signal power IkCalculation formula is as follows:
gk=[hK, 1,hk,2,…,hk,K]\hk,k=0, it is the vector of interference channel coefficient, characterizes satellite and be sent respectively to
The signal of 1~K of user, the interference channel coefficient formed after being received by user k.
Allocation vector vm=[vm,1,vm,2,…,vm,K]TIt indicates to occupying all user emission powers of channel m.
Step 203, according to common-channel interference signal power Ik, in conjunction with the sum of noise calculation interference signal and noise power Uk；
Uk=Ik+|σk|2·EM
EMIt is M rank unit matrix.
Step 204, for user k, according to receiving end available signal power Dk, the sum of interference signal and noise power Uk,
It is the service quality C characterized that user k, which is calculated, with Shannon capacityk；
Ck=Bc·det[log2(EM+Γk)]
Wherein,Indicate that user k receives SINR (signal and the interference plus noise of signal on every channel
Than)；BcExpression system is fixed as the bandwidth value of each channel setting.
Step 3: judging whether user k meets C in the Shannon capacity that each channel providesk≥CTh, if so, the distribution
Channel is available, carries out business normal transmission；Otherwise, then the channel is unavailable, and service request is by call drop or obstruction and terminates.
Transmission performance of the user k on certain channel meets the transmission rate request of user k, which is available channel.
CThIt is related with transmission services type and receiver noiseproof feature for the lowest capacity threshold value of setting.
Step 4: from C is metk≥CThAvailable channel list in dynamically distribute out a channel and give user k.
Specific step is as follows:
Quantity of state in step 401, construction markov decision process.
Dynamic channel allocation problem is modeled as markov decision process MDP, need to establish corresponding state s, movement a and
Income r.Quantity of state is defined as follows: st=(Ut,Wt,ut)
stFor the state of t moment, UtFor user being served set, W is system channel allocation matrix, utIt is to be allocated
The user of channel, i.e. user k.
The quantity of state is input in trained deep neural network by step 402, is obtained the output of network, that is, is acted
Value.
The function of deep neural network is the state s that will be inputtedtIt is mapped as M real number value, referred to as value of the movements.
Value of the movements network is denoted as Q (s；θ):s→Q(s|a；θ)；θ is the parameter of neural network；
Q(s|a；It θ) indicates when parameter is the neural network of θ, current state is under s, and selection acts the corresponding Q value of a, i.e.,
Prospective earnings value.
Trained process is as follows:
Firstly, according to the selection of value of the movements network there is the channel of maximum mapping real value to be divided when business reaches every time
Match, and records last state st-1, last time act at-1, income r immediatelyt, this next state stIt is saved in caching；
Then, the data training action value network of batch, exact value y used in training are randomly choosed from cachingj
Definition rule is as follows:
γ is discount factor parameter；J is moment variable.
Step 403, there is maximum actuation to be worth corresponding channel for selection from available channel list, distribute to user k.
Step 404 is updated deep neural network based on error back propagation principle, with lifting system dynamic channel
The performance of distribution.
Step 5: abandoning occupied channel when user k completes specific transactions transmission；New user asks next time for waiting
When asking, then enter the channel allocation of a new round.
Advantages of the present invention and bring beneficial effect are:
1, it is a kind of based on depth enhancing study dynamic channel assignment method, by with traditional static channel assignment scheme,
The comparison of conventional dynamic channel assignment scheme can effectively reduce system congestion rate.
2, a kind of dynamic channel assignment method based on depth enhancing study, by the way that satellite is modeled as intelligent body, user
Service request is modeled as external environment, and the dynamic channel allocation in multi-beam satellite scene is minimized asking for service blocking rate
Topic is modeled as maximizing the process of reachable income during intelligent body and environmental interaction, and is calculated in turn using depth enhancing study
Method solves.
3, a kind of dynamic channel assignment method based on depth enhancing study, more effectively promotion channel utilization, reduce
Service blocking rate.
4, a kind of dynamic channel assignment method based on depth enhancing study, it is contemplated that time domain is closed between dynamic channel allocation
System congestion rate can be effectively reduced in connection property, improves satellite communication system load-carrying amount and spectrum efficiency.
Detailed description of the invention
Fig. 1 is the scene of the dynamic channel assignment method based on depth enhancing study in a kind of satellite communication system of the present invention
Schematic diagram；
Fig. 2 is the process of the dynamic channel assignment method based on depth enhancing study in a kind of satellite communication system of the present invention
Figure；
Fig. 3 is the service blocking rate comparison diagram of the present invention with traditional static, dynamic channel allocation schemes.
Specific embodiment
Below in conjunction with drawings and examples, the present invention is described in further detail.
Dynamic channel assignment method (Deep based on depth enhancing study (DRL) in a kind of satellite communication system of the present invention
reinforcement learning based dynamic channel allocation method in satellite
Communication systems), as shown in Figure 2, the specific steps are as follows:
Step 1: being directed to GEO multibeam satellite system, building includes satellite multi-beam antenna, cell, channel and user
Scene；
GEO multibeam satellite system is as shown in Figure 1, the present embodiment uses 37 wave beams.In this scene, multi-beam satellite system
System is configured by satellite multi-beam antenna, and coverage area is divided into multiple independent cells, and it is each small for distributing available channel
The accessing user in area provides service.The service quality that user is obtained by the channel resource that system is distributed, depends in the channel
The information rate that can be transmitted in resource, and the service quality usually has a minimum information rate thresholding.
In the service range of satellite multi-beam antenna, user is had according to the traffic model of different geographical and is constantly initiated
Service request, then system can seek available channel in corresponding wave beam and distribute to the user, and may not be used when in system
When with channel, which will be blocked.When the user serviced is when completing specific transactions transmission, can abandon to dividing
The occupancy of allocating channel.
The user's collection serviced in satellite communication system is combined into { 1,2 ... k..., K }；System set of available channels be 1,
2,…m...,M}；The system provides each user and services distributed channel, passes through a channel occupancy matrix W=[w1,
w2,…,wK] characterize.
Step 2: calculating the user one by one on each channel using Shannon capacity as table after certain user's k initiating business request
The service quality C of signk；
Specific step is as follows:
Step 201, note transmitting terminal useful signal are s, and signal is by channel matter during satellite transmitter to receiver user
Amount is H.At this point, calculating receiving end signal y of the user k on each available channelk；
Receiving end signal ykAre as follows:
Wherein,Indicate the useful signal of user k,It indicates to remove outside user k, remaining
The common-channel interference of all users；σkIt indicates noise, the thermal noise that user's k receiving antenna introduces mainly is considered under this scene；
hk,kFor the signal for being sent to user k, after being received by user k, the channel coefficients of the useful signal of formation, wkIt indicates
System is provided the channel result that service distributes by user k, is the vector of 1*M；Distribute channel then element wkIt is 1, is otherwise 0.
OperatorIndicate Hadamard product；skIt is sent to the signal of user k for satellite, is the column vector of M*1.hk,iTo be sent to user i
Signal, the interference channel coefficient that is formed after being received by user k, wiExpression system is provided the channel that service distributes by user i
As a result, being the column vector of M*1；Distribute channel then element wiIt is 1, is otherwise 0.siThe signal of user i is sent to for satellite, i's takes
Value is 1~k.
Step 202 is directed to user k, according to the receiving end signal y of each available channelkIn useful signal and channel is dry altogether
Signal is disturbed, the corresponding available signal power D of each channel is calculated separatelykAnd common-channel interference signal power Ik,
User k received available signal power D on certain channelkCalculation formula is as follows:
Dk=| hk,k|2·diag{wk}·[diag{wk}]H (2)
Common-channel interference signal power IkCalculation formula is as follows:
gk=[hk,1,hk,2,…,hk,K]\hk,k=0, it is the vector of interference channel coefficient, characterizes satellite and be sent respectively to
The signal of 1~K of user, the interference channel coefficient formed after being received by user k.
Allocation vector vm=[vm,1,vm,2,…,vm,K] T indicates to occupying all user emission powers of channel m.
Step 203, according to common-channel interference signal power Ik, in conjunction with the sum of noise calculation interference signal and noise power Uk；
Uk=Ik+|σk|2·EM (4)
EMIt is M rank unit matrix.
Step 204, for user k, according to receiving end available signal power Dk, the sum of interference signal and noise power Uk,
It is the service quality C characterized that user k, which is calculated, with Shannon capacityk；
Ck=Bc·det[log2(EM+Γk)] (5)
Wherein,Indicate that user k receives SINR (signal and the interference plus noise of signal on every channel
Than)；BcExpression system is fixed as the bandwidth value of each channel setting.
Step 3: judging whether user k meets C in the Shannon capacity that each channel providesk≥CTh, if so, the distribution
Channel is available, carries out business normal transmission；Otherwise, then the channel is unavailable, and service request is by call drop or obstruction and terminates.
Transmission performance of the user k on certain channel meets the transmission rate request of user k, which is available channel.For
Guarantee that the business of user k obtains satisfactory service, should at least guarantee that the Shannon capacity that distributed channel resource provides is not less than setting
Lowest capacity threshold value CTh, the threshold value is related with transmission services type and receiver noiseproof feature.
Step 4: from C is metk≥CThAvailable channel list in dynamically distribute out a channel and give user k.
Dynamic channel allocation problem under satellite communication system scene, can regard Sequence Decision problem as.From the angle
For, multibeam satellite system is modeled as discrete event control system system, is asked wherein the business successively reached one by one can be divided into
It asks.In each service request Time To Event t, user u is rememberedtFor the initiation user of the service request, b is rememberedtFor the service request
Service corresponding wave beam.When new service request reach when, satellite system checks currently whether there is available channel resources, if having from
It is based on allocation strategy in available channel resources, selects allocation vectorTo user so that it is serviced；Otherwise, which will
It is blocked.Define performance indicator ΦtTo indicate to be serviced in the service request of moment t generation or be to be blocked, such as formula (6)
It is shown:
The purpose of dynamic channel allocation is intended to minimize system congestion rate, that is, minimizes the number for the business that is blocked.
Dynamic channel allocation problem is a Sequence Decision problem in satellite communication system, thus by being modeled as one
A Markovian decision process, and go to solve by the model based on depth enhancing study in turn.
Specific step is as follows:
Quantity of state in step 401, construction markov decision process.
Markov decision process MDP is one group of Sequence Decision process with Markov attribute.By dynamic channel point
It is modeled as markov decision process with problem, corresponding state s, movement a and income r need to be established.For this purpose, state be defined as be
Present channel of uniting distributes state, new service request information, and action definition is the channel that is distributed the service request, and income defines
Whether to block lower system benefit.Based on this, the MDP that can be modeled is characterized as MDP={ s, a, r }
Quantity of state is defined as follows: st=(Ut,Wt,ut)
stFor the state of t moment, UtFor user being served set, W is system channel allocation matrix, utIt is to be allocated
The user of channel, i.e. user k.
The quantity of state is input in trained deep neural network by step 402, is obtained the output of network, that is, is acted
Value.
Based on the state and movement in the MDP established, action of configuration value network.The value of the movements network is using nerve
Network (also including deep neural network), input is state value st, output is the corresponding M real number value of mapping of each movement.
This value of the movements network is denoted as Q (s；θ):s→Q(s|a；θ).It is interpreted as taking after the movement (after distributing the channel), is
Unite obtainable prospective earnings value Q (m；S_t), i.e., in state stUnder, distribute prospective earnings when channel m.
θ is the parameter of neural network；Q(s|a；It θ) indicates when parameter is the neural network of θ, current state is choosing under s
Select the corresponding Q value of movement a, i.e. prospective earnings value.
The training of the network is intended to so that the value of the movements network performance is optimal, can be selected according to current state optimal
Channel be allocated, to minimize system congestion rate.Trained process is implemented referring to following steps:
A) when business reaches every time, according to the selection of value of the movements network there is the channel of maximum mapping real value to be allocated,
And record last state st-1, last time act at-1, income r immediatelyt, this next state st；
B) by one group of data (s of recordt-1,at-1,rt,st) be saved in caching；
C) the data training action value network of batch, exact value y used in training are randomly choosed from cachingjDefinition
Rule is as follows:
γ is discount factor parameter, and general value is 0.95；J is moment variable.
Step 403, there is maximum actuation to be worth corresponding channel for selection from available channel list, distribute to user k.
Step 404 is updated deep neural network based on error back propagation principle, with lifting system dynamic channel
The performance of distribution.
Step 5: abandoning occupied channel when user k completes specific transactions transmission；New user asks next time for waiting
When asking, then enter the channel allocation of a new round.
The dynamic channel assignment method of the depth enhancing study used in artificial intelligence proposed by the invention, by with
The comparison of traditional static channel assignment scheme, conventional dynamic channel assignment scheme, performance evaluation is as shown in figure 3, side of the invention
Case can effectively reduce system congestion rate.
Claims (3)
1. the dynamic channel assignment method based on depth enhancing study in a kind of satellite communication system, which is characterized in that specific step
It is rapid as follows:
Step 1: being directed to GEO multibeam satellite system, constructing includes satellite multi-beam antenna, cell, the field of channel and user
Scape；
In this scene, multibeam satellite system is configured by satellite multi-beam antenna, coverage area is divided into multiple independent
Cell, and distribute available channel and provide service for the accessing user of each cell；
User's collection of all services in satellite system is combined into { 1,2 ... k..., K }；Set of available channels be 1,2 ... m...,
M}；
Step 2: calculating the user one by one on each channel with Shannon capacity is characterization after certain user's k initiating business request
Service quality Ck；
Specific step is as follows:
Step 201 calculates receiving end signal y of the user k on each available channelk；
Receiving end signal ykAre as follows:
Wherein, hk,k·wk⊙skIndicate the useful signal of user k,It indicates to remove outside user k, remaining is all
The common-channel interference of user；σkIndicate the thermal noise that user k receiving antenna introduces；
hk,kFor the signal for being sent to user k, after being received by user k, the channel coefficients of the useful signal of formation, wkExpression system is
User k, which is provided, services distributed channel result, is the vector of 1*M；Distribute channel then element wkIt is 1, is otherwise 0；Operator
⊙ indicates Hadamard product；skIt is sent to the signal of user k for satellite, is the column vector of M*1；hk,iFor the letter for being sent to user i
Number, the interference channel coefficient formed after being received by user k, wiExpression system is provided the channel result that service distributes by user i,
For the column vector of M*1；Distribute channel then element wiIt is 1, is otherwise 0；siThe signal of user i is sent to for satellite, the value of i is 1
~k；
The channel that all users are distributed forms channel occupancy matrix W=[w1,w2,…,wK]；
Step 202 is directed to user k, according to the receiving end signal y of each available channelkIn useful signal and common-channel interference letter
Number, calculate separately the corresponding available signal power D of each channelkAnd common-channel interference signal power Ik,
User k received available signal power D on certain channelkCalculation formula is as follows:
Dk=| hk,k|2·diag{wk}·[diag{wk}]H
Common-channel interference signal power IkCalculation formula is as follows:
gk=[hk,1,hk,2,…,hk,K]\hk,k=0, it is the vector of interference channel coefficient, characterizes satellite and be sent respectively to user 1
The signal of~K, the interference channel coefficient formed after being received by user k；
Allocation vector vm=[vm,1,vm,2,…,vm,K]TIt indicates to occupying all user emission powers of channel m；
Step 203, according to common-channel interference signal power Ik, in conjunction with the sum of noise calculation interference signal and noise power Uk；
Uk=Ik+|σk|2·EM
EMIt is M rank unit matrix；
Step 204, for user k, according to receiving end available signal power Dk, the sum of interference signal and noise power Uk, calculate
User k is the service quality C characterized with Shannon capacityk；
Ck=Bc·det[log2(EM+Γk)]
Wherein,Indicate that user k receives the SINR (Signal to Interference plus Noise Ratio) of signal on every channel；Bc
Expression system is fixed as the bandwidth value of each channel setting；
Step 3: judging whether user k meets C in the Shannon capacity that each channel providesk≥CTh, if so, the channel of the distribution
It can use, carry out business normal transmission；Otherwise, then the channel is unavailable, and service request is by call drop or obstruction and terminates；
Transmission performance of the user k on certain channel meets the transmission rate request of user k, which is available channel；
CThIt is related with transmission services type and receiver noiseproof feature for the lowest capacity threshold value of setting；
Step 4: from C is metk≥CThAvailable channel list in dynamically distribute out a channel and give user k；
Step 5: abandoning occupied channel when user k completes specific transactions transmission；When waiting new user request next time,
Then enter the channel allocation of a new round.
2. the dynamic channel assignment method based on depth enhancing study in a kind of satellite communication system as described in claim 1,
It is characterized in that, the step four is specific as follows:
Quantity of state in step 401, construction markov decision process；
Dynamic channel allocation problem is modeled as markov decision process MDP, corresponding state s, movement a and income need to be established
r；Quantity of state is defined as follows: st=(Ut,Wt,ut)；
stFor the state of t moment, UtFor user being served set, W is system channel allocation matrix, utFor channel to be allocated
User, i.e. user k；
The quantity of state is input in trained deep neural network by step 402, obtains the output of network, i.e. value of the movements；
The function of deep neural network is the state s that will be inputtedtIt is mapped as M real number value, referred to as value of the movements；
Value of the movements network is denoted as Q (s；θ):s→Q(s|a；θ)；θ is the parameter of neural network；
Q(s|a；It θ) indicates when parameter is the neural network of θ, current state is under s, and selection acts the corresponding Q value of a, i.e., expected
Financial value；
Step 403, there is maximum actuation to be worth corresponding channel for selection from available channel list, distribute to user k；
Step 404 is updated deep neural network based on error back propagation principle, with lifting system dynamic channel allocation
Performance.
3. the dynamic channel assignment method based on depth enhancing study in a kind of satellite communication system as claimed in claim 2,
It is characterized in that, the process of training is as follows in the step 402:
Firstly, according to the selection of value of the movements network there is the channel of maximum mapping real value to be allocated when business reaches every time, and
Record last state st-1, last time act at-1, income r immediatelyt, this next state stIt is saved in caching；
Then, the data training action value network of batch, exact value y used in training are randomly choosed from cachingjDefinition rule
It is then as follows:
γ is discount factor parameter；J is moment variable.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811550296.9A CN109743735A (en) | 2018-12-18 | 2018-12-18 | A kind of dynamic channel assignment method based on depth enhancing study in satellite communication system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811550296.9A CN109743735A (en) | 2018-12-18 | 2018-12-18 | A kind of dynamic channel assignment method based on depth enhancing study in satellite communication system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109743735A true CN109743735A (en) | 2019-05-10 |
Family
ID=66360539
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811550296.9A Pending CN109743735A (en) | 2018-12-18 | 2018-12-18 | A kind of dynamic channel assignment method based on depth enhancing study in satellite communication system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109743735A (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110233941A (en) * | 2019-06-28 | 2019-09-13 | 厦门大学嘉庚学院 | A kind of hot line traffic channel prediction technique and system based on Markov model |
| CN110234167A (en) * | 2019-05-14 | 2019-09-13 | 深圳信息职业技术学院 | A kind of method for channel allocation, channel dividing arrangement and electronic equipment |
| CN111211831A (en) * | 2020-01-13 | 2020-05-29 | 东方红卫星移动通信有限公司 | Multi-beam low-orbit satellite intelligent dynamic channel resource allocation method |
| CN111835407A (en) * | 2020-07-09 | 2020-10-27 | 北京邮电大学 | Low-earth-orbit satellite network flow optimization method and device |
| CN111970047A (en) * | 2020-08-25 | 2020-11-20 | 桂林电子科技大学 | LEO satellite channel allocation method based on reinforcement learning |
| WO2021040592A1 (en) * | 2019-08-30 | 2021-03-04 | Telefonaktiebolaget Lm Ericsson (Publ) | User plane selection using reinforcement learning |
| CN112769510A (en) * | 2019-10-21 | 2021-05-07 | 三星电子株式会社 | Apparatus and method for estimating interference based on machine learning |
| CN113258988A (en) * | 2021-05-13 | 2021-08-13 | 重庆邮电大学 | DQN-based multi-service low-orbit satellite resource allocation method |
| CN113271677A (en) * | 2021-05-13 | 2021-08-17 | 北京邮电大学 | Satellite channel allocation method and device |
| CN113765553A (en) * | 2021-09-09 | 2021-12-07 | 东南大学 | Multi-beam satellite communication system robust precoding method based on machine learning |
| CN114710200A (en) * | 2022-04-07 | 2022-07-05 | 中国科学院计算机网络信息中心 | Satellite network resource arrangement method and system based on reinforcement learning |
| CN115483960A (en) * | 2022-08-23 | 2022-12-16 | 爱浦路网络技术(南京)有限公司 | Beam hopping scheduling method, system, device and storage medium for low-earth-orbit satellite |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108966352A (en) * | 2018-07-06 | 2018-12-07 | 北京邮电大学 | Dynamic beam dispatching method based on depth enhancing study |
-
2018
- 2018-12-18 CN CN201811550296.9A patent/CN109743735A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108966352A (en) * | 2018-07-06 | 2018-12-07 | 北京邮电大学 | Dynamic beam dispatching method based on depth enhancing study |
Non-Patent Citations (1)
| Title |
|---|
| SHUAIJUN LIU等: "Deep Reinforcement Learning Based Dynamic Channel Allocation Algorithm in Multibeam Satellite Systems", 《IEEE ACCESS》 * |
Cited By (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110234167A (en) * | 2019-05-14 | 2019-09-13 | 深圳信息职业技术学院 | A kind of method for channel allocation, channel dividing arrangement and electronic equipment |
| CN110234167B (en) * | 2019-05-14 | 2023-04-07 | 深圳信息职业技术学院 | Channel allocation method, channel allocation device and electronic equipment |
| CN110233941A (en) * | 2019-06-28 | 2019-09-13 | 厦门大学嘉庚学院 | A kind of hot line traffic channel prediction technique and system based on Markov model |
| WO2021040592A1 (en) * | 2019-08-30 | 2021-03-04 | Telefonaktiebolaget Lm Ericsson (Publ) | User plane selection using reinforcement learning |
| CN112769510A (en) * | 2019-10-21 | 2021-05-07 | 三星电子株式会社 | Apparatus and method for estimating interference based on machine learning |
| CN112769510B (en) * | 2019-10-21 | 2024-04-05 | 三星电子株式会社 | Apparatus and method for estimating interference based on machine learning |
| CN111211831A (en) * | 2020-01-13 | 2020-05-29 | 东方红卫星移动通信有限公司 | Multi-beam low-orbit satellite intelligent dynamic channel resource allocation method |
| CN111835407A (en) * | 2020-07-09 | 2020-10-27 | 北京邮电大学 | Low-earth-orbit satellite network flow optimization method and device |
| CN111970047B (en) * | 2020-08-25 | 2022-03-25 | 桂林电子科技大学 | LEO satellite channel allocation method based on reinforcement learning |
| CN111970047A (en) * | 2020-08-25 | 2020-11-20 | 桂林电子科技大学 | LEO satellite channel allocation method based on reinforcement learning |
| CN113271677A (en) * | 2021-05-13 | 2021-08-17 | 北京邮电大学 | Satellite channel allocation method and device |
| CN113258988B (en) * | 2021-05-13 | 2022-05-20 | 重庆邮电大学 | DQN-based multi-service low-orbit satellite resource allocation method |
| CN113271677B (en) * | 2021-05-13 | 2022-10-14 | 北京邮电大学 | Satellite channel allocation method and device |
| CN113258988A (en) * | 2021-05-13 | 2021-08-13 | 重庆邮电大学 | DQN-based multi-service low-orbit satellite resource allocation method |
| CN113765553A (en) * | 2021-09-09 | 2021-12-07 | 东南大学 | Multi-beam satellite communication system robust precoding method based on machine learning |
| CN114710200A (en) * | 2022-04-07 | 2022-07-05 | 中国科学院计算机网络信息中心 | Satellite network resource arrangement method and system based on reinforcement learning |
| CN115483960A (en) * | 2022-08-23 | 2022-12-16 | 爱浦路网络技术(南京)有限公司 | Beam hopping scheduling method, system, device and storage medium for low-earth-orbit satellite |
| CN115483960B (en) * | 2022-08-23 | 2023-08-29 | 爱浦路网络技术(南京)有限公司 | Wave beam jumping scheduling method, system and device for low orbit satellite and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109743735A (en) | A kind of dynamic channel assignment method based on depth enhancing study in satellite communication system | |
| CN109947545B (en) | Task unloading and migration decision method based on user mobility | |
| TWI474728B (en) | A three layer cascade adaptive neural fuzzy inference system (anfis) based cognitive engine scheme and device | |
| US8929886B2 (en) | Apparatus and method for allocating resource in multi-beam satellite communication | |
| US20060126552A1 (en) | Apparatus and method for resource allocation in MF-TDMA broadband satellite system | |
| JP2002533987A (en) | Method in a communication system | |
| CN112583453A (en) | Downlink NOMA power distribution method of multi-beam LEO satellite communication system | |
| EP1788830A1 (en) | System and method for allocating resource and user terminal using the method | |
| CN115021799B (en) | Low-orbit satellite switching method based on multi-agent cooperation | |
| CN112566261A (en) | Deep reinforcement learning-based uplink NOMA resource allocation method | |
| CN113098584B (en) | Multi-beam satellite communication power control method for large-scale radiation array | |
| CN114885420A (en) | User grouping and resource allocation method and device in NOMA-MEC system | |
| CN114071528A (en) | Service demand prediction-based multi-beam satellite beam resource adaptation method | |
| CN113596785A (en) | D2D-NOMA communication system resource allocation method based on deep Q network | |
| CN111431646A (en) | Dynamic resource allocation method in millimeter wave system | |
| CN101883364B (en) | CMR based on multichannel machine structure and frequency spectrum distribution method | |
| CN115189721A (en) | Multi-beam satellite bandwidth power meter joint optimization distribution method and application | |
| US6947407B2 (en) | Power control based admission methods for maximum throughput in DS-CDMA networks with multi-media traffic | |
| CN117833997B (en) | Multidimensional resource allocation method of NOMA multi-beam satellite communication system based on reinforcement learning | |
| CN115942460A (en) | Low-orbit satellite wireless resource scheduling method and device based on resource map and countermeasure learning | |
| CN113541768B (en) | NOMA-based LEO satellite communication system frequency point distribution method | |
| CN115103409A (en) | Resource allocation method for multi-beam unmanned aerial vehicle cooperative communication | |
| US7974631B2 (en) | Method and system for resource allocation based on distributed scheduling in a wireless communication system including heterogeneous mobile stations | |
| CN113873658A (en) | Hopping beam resource allocation method taking user service weight gain as objective function | |
| CN116505998A (en) | Multi-beam satellite communication resource distribution system and method based on deep reinforcement learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190510 |
|
| RJ01 | Rejection of invention patent application after publication |