WO2020181695A1 - 基于贝叶斯分类器的能量采集中继系统自适应调制方法 - Google Patents

基于贝叶斯分类器的能量采集中继系统自适应调制方法 Download PDF

Info

Publication number
WO2020181695A1
WO2020181695A1 PCT/CN2019/094701 CN2019094701W WO2020181695A1 WO 2020181695 A1 WO2020181695 A1 WO 2020181695A1 CN 2019094701 W CN2019094701 W CN 2019094701W WO 2020181695 A1 WO2020181695 A1 WO 2020181695A1
Authority
WO
WIPO (PCT)
Prior art keywords
relay
modulation
energy
time slot
channel
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.)
Ceased
Application number
PCT/CN2019/094701
Other languages
English (en)
French (fr)
Inventor
朱琦
刘康
朱洪波
杨龙祥
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Nanjing University of Posts and Telecommunications
Original Assignee
Nanjing University of Posts and Telecommunications
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Nanjing University of Posts and Telecommunications filed Critical Nanjing University of Posts and Telecommunications
Priority to US17/266,428 priority Critical patent/US11323192B2/en
Publication of WO2020181695A1 publication Critical patent/WO2020181695A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B17/00—Monitoring; Testing
    • H04B17/30—Monitoring; Testing of propagation channels
    • H04B17/391—Modelling the propagation channel
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W52/00—Power management, e.g. Transmission Power Control [TPC] or power classes
    • H04W52/04—Transmission power control [TPC]
    • H04W52/38—TPC being performed in particular situations
    • H04W52/46—TPC being performed in particular situations in multi-hop networks, e.g. wireless relay networks
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
    • G06F18/24155—Bayesian classification
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B17/00—Monitoring; Testing
    • H04B17/30—Monitoring; Testing of propagation channels
    • H04B17/373—Predicting channel quality or other radio frequency [RF] parameters

Definitions

  • the invention belongs to the field of communication technology, and specifically relates to an adaptive modulation method for an energy harvesting relay system based on a naive Bayes classifier.
  • Energy harvesting technology means that a node collects energy from the surrounding environment and uses it to send its own information as transmission power. Since energy is obtained from the environment, energy harvesting technology is not only environmentally friendly, but also extends the life of energy self-sufficient devices. At the same time, as the communication range expands, the communication quality of edge users decreases due to the increase in channel fading.
  • Cooperative relay technology refers to the addition of relay nodes between the sender and the receiver, and the signal quality is improved through relay forwarding, thereby effectively improving the coverage of the network, avoiding the degradation of communication quality due to channel fading, and providing certain Communication gain. Energy harvesting relay has become an important research direction in the next generation of mobile communications.
  • M-ary Quadrature Amplitude Modulation is one of the most commonly used modulation techniques.
  • MQAM M-ary Quadrature Amplitude Modulation
  • the system includes source nodes and relays, both of which need to send data, so how the two determine their own modulation methods will directly determine the throughput that the system can provide. Since the relay uses energy harvesting technology and the available transmit power is limited, the modulation method is limited by the harvested energy, and the modulation method of the source node mainly depends on the amount of traffic allowed by the system.
  • the relay since the relay is allowed to store energy, the relay can choose whether to store energy according to changes in the channel status. When the channel status is better, a higher modulation method can be used to achieve greater system throughput.
  • the algorithm research of using a reasonable modulation method to improve system throughput is still lacking.
  • the present invention provides an adaptive modulation method for an energy harvesting relay system based on a naive Bayes classifier, which includes the following steps:
  • the system is composed of a sending end S, a receiving end D and an energy harvesting relay R.
  • the relay includes a data buffer and an energy buffer. If the storage capacity of the data buffer in time slot t-1 is Q t-1 , according to Determine the modulation mode of the transmitting end of the time slot t Where Q MAX is the maximum storage capacity of the relay buffer;
  • the bit error rate requirement at the receiving end Is the modulation order of the modulation method used by the relay, N 0 is the noise power, W is the received signal bandwidth, and h rd represents the channel state from the relay to the source node. Calculate the remaining energy B t of the relay, if B t ⁇ 0, modify the modulation order Indicates the maximum modulation order that can be selected at this moment.
  • step S2 includes the following steps:
  • adjusting the modulation mode of the time slot t in the step S2-3 specifically includes the following steps:
  • step B According to the new stored energy state, use the method in step A to obtain the time slot t 1 > t, the maximum modulation order that can be used for the relay
  • the basic idea of the present invention is to use the naive Bayes classification algorithm in the machine learning algorithm to provide an on-line energy harvesting relay adaptive modulation method when the channel information is unknown in the future.
  • First determine the modulation method used by the sender according to the status of the relay data buffer, so that the sent data does not exceed the storage capacity of the relay, and then the relay according to the data sent by the sender, and the channel status and collection from itself to the receiver in the past period of time
  • the off-line algorithm is used to determine the modulation method used during this period of time, in order to obtain a higher total system throughput.
  • use the modulation method and channel state information obtained by the offline algorithm as training data and use the naive Bayes classifier method to obtain the classification model.
  • the relay can further adaptively select the modulation according to the channel state information of the current time slot According to the energy state, the classification result is dynamically adjusted, and finally the modulation method adopted by all time slot relays is obtained.
  • the present invention has the following advantages:
  • the selection strategy of the modulation mode under the constraints of energy storage and data storage is given.
  • the appropriate modulation mode can be selected under different receiving end bit error rate requirements, thereby increasing the total system Throughput.
  • the relay node can adaptively select a suitable modulation method according to the existing classification model, without relying on the system information of the subsequent time slot, which is more in line with the actual communication system model.
  • the classification model does not need to dynamically change according to the energy storage state and the channel state, which greatly reduces the loss of the relay in order to maintain the classification model.
  • Figure 1 is a schematic flow diagram of the method of the present invention
  • Figure 2 is a comparison diagram of the throughput change of the relay system introduced by the wireless communication system
  • Figure 3 is a comparison chart of the total system throughput with the user error rate requirements
  • Figure 4 is a comparison chart of the total system throughput with the change of energy arrival rate
  • Figure 5 is a schematic diagram of the composition of the system of the present invention.
  • the method of the present invention considers that in the single-relay wireless communication scenario, the nodes all use the MQAM modulation method to send signals, and the nodes need to adaptively determine the currently adopted modulation method according to the current time slot system information and use the machine learning method to improve the system throughput The amount of goals.
  • the sender determines the modulation method to be used according to the state of the relay data buffer so that the sent data does not exceed the relay storage capacity, and then relays the data sent by the sender, and a period of time Channel state and energy collection state information, determine the optimal modulation method used by the relay during this period to achieve the goal of increasing the total system throughput, and then use the modulation method, channel state information and energy information as the training data, according to Pu Su Bei
  • the Yees classifier algorithm obtains the classification model.
  • the relay can adaptively select the modulation mode under the condition of only knowing the current channel information and energy information, so as to achieve the goal of increasing the total system throughput and improving the energy utilization efficiency.
  • this embodiment provides an adaptive modulation method for an energy harvesting relay system based on a naive Bayes classifier, as shown in FIG. 1, which includes the following steps:
  • the system consists of a sending end S, a receiving end D and an energy harvesting relay R.
  • the relay includes a data buffer and an energy buffer.
  • the storage capacity of the time slot t-1 data buffer is Q t-1 , according to the formula ( 1) Determine the modulation method of time slot t
  • Q max represents the maximum storage capacity of the data buffer
  • ⁇ represents the half-slot length
  • ⁇ s represents the transmission symbol period
  • the channel gain relayed to the destination node at time slot t is Then the corresponding normalized channel gain is among them Represents the average channel gain of the channel from the relay to the destination node.
  • the channel is discretized according to the value of k t
  • h d is the discretized value of the channel state.
  • the energy storage state of the time slot t relay is Obtain the discretized value of the energy storage state:
  • the relay selects the modulation order with the largest probability as the modulation method of the current time slot
  • the transmission power of the relay is determined according to the modulation order:
  • BER is the bit error rate requirement of the receiving end
  • W is the received signal bandwidth
  • h rd represents the channel state from the relay to the source node, and then calculates the remaining energy B t according to the energy collected by the relay:
  • E t represents the initial energy state of the energy buffer at time slot t.
  • the determination of the probability distribution matrix in the above step S2 includes the following steps S2-1 to S2-5:
  • the new energy state is:
  • step B According to the new stored energy state, use the method in step A to obtain the time slot t 1 > t, then the maximum modulation order that can be used for the relay which is
  • the posterior probability can be rewritten as:
  • the denominator of the posterior probability formula is the same, so the posterior probability is only determined by the numerator P(h d /M r )P(E d /M r )P(M r ). Assuming that for all modulation orders, P(M r ) is equal, the posterior probability is only obtained by the product of conditional probabilities P(h d /M r )P(E d /M r ), and the posterior probability of all orders is calculated , Get the final probability distribution matrix.
  • the specific process of the method of the present invention can be summarized as follows: First, according to the state of the relay data buffer, it is necessary to determine the modulation mode adopted by the sending end; the relay then according to the data sent by the sending end and the past For a period of time, the channel state and collected energy from itself to the receiving end are used to determine the modulation method used during this period using the offline algorithm; then, the modulation method and channel state information obtained by the offline algorithm are used as training data and calculated according to the naive Bayes formula The probability distribution matrix of the energy storage state and channel state corresponding to different modulation methods is used to obtain the classification model; according to the classification model, the relay can further adaptively select the modulation method according to the channel state information and energy storage state information of the current time slot, and finally Obtain the modulation mode used by all timeslot relays.
  • Figure 2 shows the change in system throughput after the wireless system introduces relay. It can be seen that after using the method of the present invention, the system throughput is significantly improved after the relay is introduced. This is because for users, especially users at the edge of the base station, due to the large-scale fading of the channel, the channel fading is relatively large, and the transmission power reaching the receiving end attenuates significantly.
  • the use of relay technology can effectively improve user coverage. Through the forwarding of the relay node, the originally fading signal can be enhanced, so that the signal strength of the receiving end can be maintained at a higher level, and the signal-to-noise ratio of the receiving end signal is improved. System throughput has been improved.
  • Fig. 3 is a comparison diagram of the system throughput of the online adaptive modulation algorithm adopted by the present invention and the algorithm in which the relay directly adopts the maximum modulation mode. Obviously, the system throughput has been significantly increased by adaptively adjusting the modulation mode of the relay. Because by adaptively adjusting the modulation mode of the relay, the appropriate modulation mode can be selected according to the channel state, rather than just based on the collected energy.
  • the channel gain of time slot t is The channel gain of time slot t 1 is Suppose that the energy collected in time slot t is significantly more than that in time slot t 1 , but the channel But worse than If the modulation method is not adjusted, the relay will use the collected energy as much as possible in the time slot t, but will obtain a lower system throughput due to the poor channel, but at t 1 it will also obtain the same due to lack of energy. Lower system throughput.
  • the relay adjusts its own modulation mode according to the channel state, and stores energy until time t 1 , so that although part of the throughput at time slot t is discarded, more system throughput is obtained in time slot t 1 Therefore, the total throughput of the system is improved. As the bit error rate requirements at the receiving end increase, the total system throughput also increases.
  • Figure 4 is a comparison diagram of the average throughput of the system in the online adaptive modulation algorithm adopted by the present invention and the algorithm in which the relay directly adopts the maximum modulation mode. It can be seen that the average system throughput of the method of the present invention is higher than that of the relay directly.
  • the maximum modulation method significantly improves the total system throughput, and as the energy arrival rate gradually increases, the system's achievable capacity also increases. These are better than the relay method that directly uses the maximum modulation.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Signal Processing (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Electromagnetism (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Quality & Reliability (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Probability & Statistics with Applications (AREA)
  • Radio Relay Systems (AREA)
  • Mobile Radio Communication Systems (AREA)

Abstract

本发明公开了一种基于朴素贝叶斯分类器的能量采集中继系统自适应调制方法，首先发送端根据中继数据缓存器的状态，确定采用的调制方式，使得发送数据不超过中继存储能力，接着中继根据发送端发送的数据，以及一段时间内信道状态和能量采集状态信息，确定这段时间内中继采用的最优调制方式，再将调制方式以及信道状态信息和能量信息作为训练数据，根据朴素贝叶斯分类器算法得到分类模型，中继可以在只知道当前时刻信道信息和能量信息的条件下，自适应的选择调制方式。本发明考虑了节点的具体调制方式，给出了在能量存储和数据存储限制条件下的调制方式选择策略，从而增大了系统总的吞吐量，提高了能量利用效率。

Description

基于贝叶斯分类器的能量采集中继系统自适应调制方法 技术领域
本发明属于通信技术领域，具体涉及一种基于朴素贝叶斯分类器的能量采集中继系统自适应调制方法。
背景技术
近年来，随着无线通信技术的发展，无线业务的应用越来越广泛，能源消耗越来越大，所以绿色通信已成为关注的焦点。能量采集技术是指节点从周围环境采集能量，并用于自身信息的发送，作为发送功率。由于从环境获取能量，能量采集技术不仅绿色环保，同时延长了能量自给设备的寿命，同时，随着通信范围的扩大，边缘用户的通信质量由于信道衰落的增加而降低。协作中继技术是指在发送端和接收端之间增加中继节点，通过中继转发改善信号质量，从而有效提升网络的覆盖率，避免因为信道衰落造成通信质量的降低，并能提供一定的通信增益。能量采集中继已成为下一代移动通信中一个重要的研究方向。
对于能量采集系统中功率分配的研究已有很多，但是对于实际的通信系统，通常需要考虑具体的调制方式。数字通信系统中，M进制正交振幅调制(MQAM)是最常用的调制技术之一。对于能量采集中继系统，考虑系统中节点如何选择调制方式，将面临两个主要挑战。首先，系统包含源节点和中继，都需要发送数据，所以两者分别如何确定自己的调制方式，将直接决定系统所能提供的吞吐量。由于中继采用能量采集技术，可用发送功率受限，所以调制方式受采集能量的限制，而源节点的调制方式则主要取决于系统所能允许的业务量。因此，需要跟据不同的限制条件确定两者各自的调制阶数。其次，由于允许中继存储能量，中继可以根据信道状态的变化情况选择是否对能量进行存储，当信道状态更好时，可以采用更高的调制方式，达到更大的系统吞吐量。然而对于实际的通信系统，由于存在小尺度衰落，节点很难预测将来时刻的信道状态，难以提前做出判断是否对能量进行存储，系统吞吐量难以提高。所以，对于在缺乏信道信息的前提下，采用合理的调制方式，提升系统吞吐量的算法研究目前还比较欠缺。
发明内容
发明目的：为了克服现有技术中存在的不足，提供一种基于朴素贝叶斯分类器的能量采集中继系统自适应调制方法，其针对实际场景节点只能得到当前时刻的信道信息，而无法预测未来信道状态的问题，构建了一种利用朴素贝叶斯分类算法确定节点调制方式的方案，能够有效的增大系统吞吐量。
技术方案：为实现上述目的，本发明提供一种基于朴素贝叶斯分类器的能量采集中继系统自适应调制方法，包括如下步骤：
S1：根据中继数据缓存器的状态，确定发送端采用的调制方式：
系统由发送端S，接收端D和一个能量采集中继R组成，中继包含一个数据缓存器和一个能量缓存器，若时隙t-1数据缓存器存储能力为Q t-1，根据
Figure PCTCN2019094701-appb-000001
确定时隙t发送端的调制方式
Figure PCTCN2019094701-appb-000002
其中Q MAX为中继缓存器的最大存储量；
S2：根据概率分布矩阵确定中继采用的调制方式：
时隙t时中继到目的节点的信道增益为
Figure PCTCN2019094701-appb-000003
则对应的归一化信道增益为
Figure PCTCN2019094701-appb-000004
其中
Figure PCTCN2019094701-appb-000005
表示中继到目的节点信道的平均信道增益，根据k t的值将信道离散化，h d为信道状态离散化后的值，时隙t中继的能量存储状态为
Figure PCTCN2019094701-appb-000006
得到能量存储状态离散化的值
Figure PCTCN2019094701-appb-000007
根据概率分布矩阵，分别对所有的调制阶数M r＝1,4,16,64,256计算条件概率P(h d/M r)和P(E d/M r)，并得到后验概率为
Figure PCTCN2019094701-appb-000008
中继通过比较不同调制阶数后验概率的大小，选择概率最大的调制阶数为当前时隙的调制方式
Figure PCTCN2019094701-appb-000009
S3：确定中继的发送功率：
根据调制阶数确定中继的发送功率
Figure PCTCN2019094701-appb-000010
其中，BER是
接收端的误码率要求，
Figure PCTCN2019094701-appb-000011
是中继所采用调制方式的调制阶数，N 0是噪声功率，W是接收信号带宽，h rd表示中继到源节点的信道状态，根据
Figure PCTCN2019094701-appb-000012
计算中继的剩余能量B t，如果B t＜0，则修改调制阶数
Figure PCTCN2019094701-appb-000013
表示此时刻可以选择的最大调制阶数。
进一步地，所述步骤S2中概率分布矩阵的确定包括如下步骤：
S2-1：在时隙t，中继根据采集到的能量，确定可以采用的最大调制阶数
Figure PCTCN2019094701-appb-000014
其中
Figure PCTCN2019094701-appb-000015
表示从集合M＝{1,4,16,64,256}选择出最大的但不大于x的元素，并根据
Figure PCTCN2019094701-appb-000016
计算总的系统吞吐量
Figure PCTCN2019094701-appb-000017
其中T表示总时间长度；
S2-2：根据调制阶数确定中继的发送功率
Figure PCTCN2019094701-appb-000018
确定中继的剩余能量B t；
S2-3：调整时隙t的调制方式；
S2-4：重复以上步骤直到所有T时刻中继都已选择合适的调制方式；
S2-5：按中继所采用的调制阶数，根据朴素贝叶斯分类算法得到概率分布矩阵，将训练数据中时隙对应的信道状态信息和能量存储状态信息离散化，分别统计同一调制方式下不同信道状态和能量状态的个数，计算出两种属性对应的条件概率P(h d/M r)和P(E d/M r)，根据贝叶斯公式得到后验概率P(M r/h d×E d)，统计所有阶数的后验概率，得到最终概率分布矩阵。
进一步地，所述步骤S2-3中调整时隙t的调制方式具体包括如下步骤：
A)随机选择
Figure PCTCN2019094701-appb-000019
满足
Figure PCTCN2019094701-appb-000020
计算新的中继端发送功率
Figure PCTCN2019094701-appb-000021
得到新的存储能量
Figure PCTCN2019094701-appb-000022
B)根据新的存储能量状态，利用步骤A中方法得到时隙t 1＞t，中继的可采用的最大调制阶数
Figure PCTCN2019094701-appb-000023
C)根据新的调制阶数
Figure PCTCN2019094701-appb-000024
确定新的系统总吞吐量V′，如果V′＞V，更新t 0时刻调制阶数
Figure PCTCN2019094701-appb-000025
否则调制阶数不变；
D)重复此步骤直到t 1时刻所有可选择的调制阶数都被选择。
本发明的基本思路是利用机器学习算法中的朴素贝叶斯分类算法，在将来信道信息未知的情况下，给出一种在线的能量采集中继自适应的调制方法。首先根据中继数据缓存器的状态，确定发送端采用的调制方式，使得发送数据不超过中继存储能力，接着中继根据发送端发送的数据，以及过去一段时间自身到接收端的信道状态和采集能量，利用离线算法确定这段时间内所采用的调制方式，目的是获得更高的系统总吞吐量。再将离线算法得到的调制方式以及信道状态信息作为训练数据，利用朴素贝叶斯分类器方法得到分类模型，根据分类模型，中继可进一步的根据当前时隙的信道状态信息自适应的选择调制方式，并根据能量状态，动态调整分类结果，最终得到所有时隙中继所采用的调制方式。
有益效果：本发明与现有技术相比，具有以下优点：
1、考虑了节点的具体调制方式，给出了在能量存储和数据存储限制条件下的调制方式选择策略，可以在不同接收端误码率要求下选择合适的调制方式，从而增大系统总的吞吐量。
2、中继节点在采用能量采集的前提下，可以根据已有的分类模型自适应的选择合适的调制方式，不需要依赖之后时隙的系统信息，更加符合实际的通信系统模型。且分类模型不需要根据能量存储器状态和信道状态动态变化，大大减少了中继为了维护分类模型而产生的损耗。
3、通过对存储能量的合理分配，使得中继采集能量利用率提高，避免了由于信道状况不佳造成的能量浪费，提升系统总的吞吐量性能。
附图说明
图1为本发明方法的流程示意图；
图2为无线通信系统引入中继系统吞吐量变化对比图；
图3为系统总吞吐量随用户误码率要求变化对比图；
图4为系统总吞吐量随能量到达率的变化对比图；
图5为本发明系统的组成示意图。
具体实施方式
下面结合附图和具体实施例，进一步阐明本发明。
本发明方法考虑在单中继无线通信场景中，节点均采用MQAM调制方式发送信号，节点需要根据当前时隙系统信息，利用机器学习方法自适应的确定当前采用的调制方式，以达到提升系统吞吐量的目标。参照图5，其具体过程为：发送端根据中继数据缓存器的状态，确定采用的调制方式，使得发送数据不超过中继存储能力，接着中继根据发送端发送的数据，以及一段时间内信道状态和能量采集状态信息，确定这段时间内中继采用的最优调制方式，以达到提升系统总吞吐量的目标，再将调制方式以及信道状态信息和能量信息作为训练数据，根据朴素贝叶斯分类器算法得到分类模型，中继可以在只知道当前时刻信道信息和能量信息的条件下，自适应的选择调制方式，以达到提升系统总吞吐量，提高能量利用效率的目标。
根据上述设计原理和工作过程，本实施例提供一种基于朴素贝叶斯分类器的能量采集中继系统自适应调制方法，如图1所示，其包括如下步骤：
S1：根据中继数据缓存器的状态，确定发送端采用的调制方式：
系统由发送端S，接收端D和一个能量采集中继R组成，中继包含一个数据缓存器和一个能量缓存器，时隙t-1数据缓存器存储能力为Q t-1，根据公式(1)确定时隙t的 调制方式
Figure PCTCN2019094701-appb-000026
Figure PCTCN2019094701-appb-000027
其中Q max表示数据缓存器的最大存储量，τ表示半时隙长度，τ s表示发送符号周期，
Figure PCTCN2019094701-appb-000028
的取值属于集合M＝{1,4,16,64,256}。
S2：根据概率分布矩阵确定中继采用的调制方式：
时隙t时中继到目的节点的信道增益为
Figure PCTCN2019094701-appb-000029
则对应的归一化信道增益为
Figure PCTCN2019094701-appb-000030
其中
Figure PCTCN2019094701-appb-000031
表示中继到目的节点信道的平均信道增益，根据k t的值将信道离散化，h d为信道状态离散化后的值，将信道状态分为四个等级，即h d＝1,2,3,4。具体如下表达式(2)：
Figure PCTCN2019094701-appb-000032
时隙t中继的能量存储状态为
Figure PCTCN2019094701-appb-000033
得到能量存储状态离散化的值：
Figure PCTCN2019094701-appb-000034
根据概率分布矩阵，分别对所有的调制阶数M r＝1,4,16,64,256计算条件概率P(h d/M r)和P(E d/M r)，并得到后验概率为：
Figure PCTCN2019094701-appb-000035
中继通过比较不同调制阶数后验概率的大小，选择概率最大的调制阶数为当前时隙的调制方式
Figure PCTCN2019094701-appb-000036
S3：确定中继的发送功率：
本实施例中为了充分利用能量，假设中继确定调制方式后，均采用最小发射功率发送数据，则根据调制阶数确定中继的发送功率:
Figure PCTCN2019094701-appb-000037
其中，BER是接收端的误码率要求，
Figure PCTCN2019094701-appb-000038
是中继所采用调制方式的调制阶数，N 0是噪声功率，W是接收信号带宽，h rd表示中继到源节点的信道状态，再根据中继采集的能量计算剩余能量B t：
Figure PCTCN2019094701-appb-000039
其中，E t表示时隙t能量缓存器的初始能量状态。
如果
B t＜0(7)
则修改选择的调制阶数为：
Figure PCTCN2019094701-appb-000040
上述步骤S2中概率分布矩阵的确定包括如下步骤S2-1～S2-5：
S2-1：在时隙t，中继根据采集到的能量，确定可以采用的最大调制阶数
Figure PCTCN2019094701-appb-000041
其中
Figure PCTCN2019094701-appb-000042
表示从集合M＝{1,4,16,64,256}选择出最大的但不大于x的元素，并根据
Figure PCTCN2019094701-appb-000043
计算总的系统吞吐量
Figure PCTCN2019094701-appb-000044
其中T表示总时间长度。
S2-2：根据公式(2)确定中继的发送功率
Figure PCTCN2019094701-appb-000045
再根据公式(3)确定中继的剩余能量B t。
S2-3：利用机器学习的方法调整时隙t的调制方式，具体包括如下步骤A～D：
A)随机选择
Figure PCTCN2019094701-appb-000046
满足
Figure PCTCN2019094701-appb-000047
则中继可多存储的能量为：
Figure PCTCN2019094701-appb-000048
即新的能量状态为：
Figure PCTCN2019094701-appb-000049
B)根据新的存储能量状态，利用步骤A中方法得到时隙t 1＞t，则中继的可采用的最大调制阶数
Figure PCTCN2019094701-appb-000050
即
Figure PCTCN2019094701-appb-000051
C)根据新的调制阶数
Figure PCTCN2019094701-appb-000052
确定新的系统总吞吐量V′，如果V'＞V，更新t时刻调制 阶数
Figure PCTCN2019094701-appb-000053
否则调制阶数不变；
D)重复步骤A～C直到t 1时刻所有可选择的调制阶数都被选择。
S2-4：重复以上步骤直到所有T时刻中继都已选择合适的调制方式。
S2-5：根据公式(2)和(3)得到离散信道状态h d和离散能量存储器状态E d，分别统计同一调制方式下不同信道状态和能量状态的个数，计算出两种属性对应的条件概率P(h d/M r)和P(E d/M r)。假设P(M r)为不同调制阶数的初始概率分布，根据贝叶斯公式得到后验概率：
Figure PCTCN2019094701-appb-000054
根据属性条件独立性的假设，后验概率可以改写为：
Figure PCTCN2019094701-appb-000055
对于不同调制阶数，后验概率公式的分母均相同，因此后验概率仅由分子P(h d/M r)P(E d/M r)P(M r)决定。假设对于所有调制阶数，P(M r)相等，后验概率仅通过条件概率的乘积P(h d/M r)P(E d/M r)求得，统计所有阶数的后验概率，得到最终概率分布矩阵。
根据上述具体方法步骤，可将本发明方法的具体过程进行如下归纳：首先根据中继数据缓存器的状态，需要先确定发送端采用的调制方式；中继再根据发送端发送的数据，以及过去一段时间自身到接收端的信道状态和采集能量，利用离线算法确定这段时间内所采用的调制方式；之后，将离线算法得到的调制方式以及信道状态信息作为训练数据，根据朴素贝叶斯公式计算不同调制方式对应的能量存储状态和信道状态的概率分布矩阵，得到分类模型；根据分类模型，中继可进一步的根据当前时隙的信道状态信息和能量存储状态信息自适应的选择调制方式，最终得到所有时隙中继所采用的调制方式。
本实施例中为了验证本发明方法的实际效果进行仿真实验，具体如图2～图4所示：
如图2所示为无线系统引入中继后系统吞吐量的变化。可见运用本发明的方法后，引入中继后系统吞吐量有明显的提升。这是由于对于用户，尤其是处于基站边缘覆盖的用户，由于信道的大尺度衰落影响，信道衰落比较大，发送功率到达接收端衰减明显。而采用中继技术可以有效改善用户的覆盖，通过中继节点的转发，可以增强原本已经衰落的信号，使得接收端信号强度维持在较高的水平，提升了接收端信号的信噪比，因此系统吞吐量得到提升。当中继位置越靠近发送端与接收端中心时，其对于信号衰落的改善效果最好，而中继位于两端时对于信号基本没有改善，因此图2系统吞吐量随中继的 位置呈现先增大后减小的趋势。
如图3所示为本发明采用的在线自适应调制算法与中继直接采用最大调制方式的算法系统吞吐量的对比图。显然通过自适应调整中继的调制方式，系统吞吐量有了明显的升高。因为通过自适应的调整中继的调制方式，可以根据信道状态选择合适的调制方式，而不是仅仅根据采集到的能量。比如时隙t的信道增益为
Figure PCTCN2019094701-appb-000056
时隙t 1的信道增益为
Figure PCTCN2019094701-appb-000057
假设时隙t采集的能量明显多于时隙t 1，但是信道
Figure PCTCN2019094701-appb-000058
却差于
Figure PCTCN2019094701-appb-000059
如果不调整调制方式，中继会将采集到的能量尽可能的多用于时隙t，而却因为信道较差而获得较低的系统吞吐量，而在t 1时刻却因为缺少能量而同样获得较低的系统吞吐量。本发明方法中中继会根据信道状态调整自身调制方式，而将能量存储至t 1时刻，这样虽然放弃了时隙t时刻的部分吞吐量，但是在时隙t 1获得了更多的系统吞吐量，系统的总吞吐量因此得到提升。随着接收端误码率要求的提高，系统总吞吐量也随之提高。
如图4所示为本发明采用的在线自适应调制算法与中继直接采用最大调制方式的算法在系统平均吞吐量的对比图，可见本发明方法的系统平均吞吐量要高于中继直接采用最大调制的方式，显著提升了系统总吞吐量，且随着能量到达率逐渐提高，系统能达到的容量也随之提高，这些都优于中继直接采用最大调制的方式。

Claims (4)

  1. 基于朴素贝叶斯分类器的能量采集中继系统自适应调制方法，其特征在于：包括如下步骤：
    S1：根据中继数据缓存器的状态，确定发送端采用的调制方式：
    系统由发送端S，接收端D和一个能量采集中继R组成，中继包含一个数据缓存器和一个能量缓存器，若时隙t-1数据缓存器存储能力为Q t-1，根据
    Figure PCTCN2019094701-appb-100001
    确定时隙t发送端的调制方式
    Figure PCTCN2019094701-appb-100002
    其中Q MAX为中继缓存器的最大存储量；
    S2：根据概率分布矩阵确定中继采用的调制方式：
    时隙t时中继到目的节点的信道增益为
    Figure PCTCN2019094701-appb-100003
    则对应的归一化信道增益为
    Figure PCTCN2019094701-appb-100004
    其中
    Figure PCTCN2019094701-appb-100005
    表示中继到目的节点信道的平均信道增益，根据k t的值将信道离散化，h d为信道状态离散化后的值，时隙t中继的能量存储状态为
    Figure PCTCN2019094701-appb-100006
    得到能量存储状态离散化的值
    Figure PCTCN2019094701-appb-100007
    根据概率分布矩阵，分别对所有的调制阶数计算条件概率P(h d/M r)和P(E d/M r)，并得到后验概率为
    Figure PCTCN2019094701-appb-100008
    中继通过比较不同调制阶数后验概率的大小，选择概率最大的调制阶数为当前时隙的调制方式
    Figure PCTCN2019094701-appb-100009
    S3：确定中继的发送功率：
    根据调制阶数确定中继的发送功率
    Figure PCTCN2019094701-appb-100010
    其中，BER是接收端的误码率要求，
    Figure PCTCN2019094701-appb-100011
    是中继所采用调制方式的调制阶数，N 0是噪声功率，W是接收信号带宽，h rd表示中继到源节点的信道状态，根据
    Figure PCTCN2019094701-appb-100012
    计算中继的剩余能量B t，如果B t＜0，则修改调制阶数
    Figure PCTCN2019094701-appb-100013
    表示此时刻可以选择的最大调制阶数。
  2. 根据权利要求1所述的基于朴素贝叶斯分类器的能量采集中继系统自适应调制方法，其特征在于：所述步骤S2中概率分布矩阵的确定包括如下步骤：
    S2-1：在时隙t，中继根据采集到的能量，确定可以采用的最大调制阶数
    Figure PCTCN2019094701-appb-100014
    其中
    Figure PCTCN2019094701-appb-100015
    表示从集合M＝{1,4,16,64,256}选择出最大的但不大于x的元素，并根据
    Figure PCTCN2019094701-appb-100016
    计算总的系统吞吐量
    Figure PCTCN2019094701-appb-100017
    其中T表示总时间长度；
    S2-2：根据调制阶数确定中继的发送功率
    Figure PCTCN2019094701-appb-100018
    确定中继的剩余能量B t；
    S2-3：调整时隙t的调制方式；
    S2-4：重复以上步骤直到所有T时刻中继都已选择合适的调制方式；
    S2-5：按中继所采用的调制阶数，根据朴素贝叶斯分类算法得到概率分布矩阵，将训练数据中时隙对应的信道状态信息和能量存储状态信息离散化，分别统计同一调制方式下不同信道状态和能量状态的个数，计算出两种属性对应的条件概率P(h d/M r)和P(E d/M r)，根据贝叶斯公式得到后验概率P(M r/h d×E d)，统计所有阶数的后验概率，得到最终概率分布矩阵。
  3. 根据权利要求2所述的基于朴素贝叶斯分类器的能量采集中继系统自适应调制方法，其特征在于：所述步骤S2-3中调整时隙t的调制方式具体包括如下步骤：
    A)随机选择
    Figure PCTCN2019094701-appb-100019
    满足
    Figure PCTCN2019094701-appb-100020
    计算新的中继端发送功率
    Figure PCTCN2019094701-appb-100021
    得到新的存储能量
    Figure PCTCN2019094701-appb-100022
    B)根据新的存储能量状态，利用步骤A中方法得到时隙t 1＞t，中继的可采用的最大调制阶数
    Figure PCTCN2019094701-appb-100023
    C)根据新的调制阶数
    Figure PCTCN2019094701-appb-100024
    确定新的系统总吞吐量V′，如果V'＞V，更新t 0时刻调制阶数
    Figure PCTCN2019094701-appb-100025
    否则调制阶数不变；
    D)重复此步骤直到t 1时刻所有可选择的调制阶数都被选择。
  4. 根据权利要求1所述的基于朴素贝叶斯分类器的能量采集中继系统自适应调制方法，其特征在于：所述步骤2中调制阶数M r＝1,4,16,64,256。
PCT/CN2019/094701 2019-03-12 2019-07-04 基于贝叶斯分类器的能量采集中继系统自适应调制方法 Ceased WO2020181695A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US17/266,428 US11323192B2 (en) 2019-03-12 2019-07-04 Adaptive modulation method for Bayes classifier-based energy harvesting relay system

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910183808.0A CN109951239B (zh) 2019-03-12 2019-03-12 基于贝叶斯分类器的能量采集中继系统自适应调制方法
CN201910183808.0 2019-03-12

Publications (1)

Publication Number Publication Date
WO2020181695A1 true WO2020181695A1 (zh) 2020-09-17

Family

ID=67009507

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/094701 Ceased WO2020181695A1 (zh) 2019-03-12 2019-07-04 基于贝叶斯分类器的能量采集中继系统自适应调制方法

Country Status (3)

Country Link
US (1) US11323192B2 (zh)
CN (1) CN109951239B (zh)
WO (1) WO2020181695A1 (zh)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109951239B (zh) 2019-03-12 2020-05-19 南京邮电大学 基于贝叶斯分类器的能量采集中继系统自适应调制方法
FI129428B (en) * 2019-09-18 2022-02-15 Wirepas Oy A decentralized synchronization solution for wireless communication networks
CN111800194B (zh) * 2020-06-22 2021-06-18 北京理工大学 针对少模多芯oam光纤传输概率分布的非线性补偿方法
CN119835658B (zh) * 2023-10-13 2025-12-30 中国移动通信有限公司研究院 自适应调制编码方法、网络设备、网元及存储介质

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108551383A (zh) * 2018-03-26 2018-09-18 南京邮电大学 一种基于机器学习的能量采集中继系统自适应调制方法
CN109286426A (zh) * 2018-09-25 2019-01-29 中国计量大学 一种无线携能协作预编码空间调制系统的传输方法
CN109951239A (zh) * 2019-03-12 2019-06-28 南京邮电大学 基于贝叶斯分类器的能量采集中继系统自适应调制方法

Family Cites Families (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR100996080B1 (ko) * 2003-11-19 2010-11-22 삼성전자주식회사 직교 주파수 분할 다중 방식을 사용하는 통신 시스템에서적응적 변조 및 코딩 제어 장치 및 방법
WO2007036031A1 (en) * 2005-09-30 2007-04-05 Nortel Networks Limited Adaptive power control data transmission systems and methods
CN102436740A (zh) * 2011-09-29 2012-05-02 东南大学 一种高速公路交通事件自动检测方法
CN102647391A (zh) * 2012-03-30 2012-08-22 南京邮电大学 一种基于决策层数据融合的协作调制信号识别方法
CN104038993B (zh) * 2014-05-28 2017-08-11 南京邮电大学 一种盲源分离的无线传感器网络节点选择方法
US10264525B2 (en) * 2014-11-17 2019-04-16 University Of Notre Dame Du Lac Energy efficient communications
DE112015006151B4 (de) * 2015-02-13 2025-11-13 Nec Corporation Verfahren zum Betrieb eines mobilen Kommunikationsnetzwerks
CN104994564B (zh) * 2015-06-25 2019-01-25 苏州大学 存在反馈延迟的放大转发中继选择的优化方法
ES2899137T3 (es) * 2017-11-09 2022-03-10 Samsung Electronics Co Ltd Procedimiento y aparato para comunicación inalámbrica en un sistema de comunicación inalámbrica
CN109067488B (zh) * 2018-07-13 2021-03-09 吉林大学 基于能量累积的信息与能量双向传输系统性能的优化方法

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108551383A (zh) * 2018-03-26 2018-09-18 南京邮电大学 一种基于机器学习的能量采集中继系统自适应调制方法
CN109286426A (zh) * 2018-09-25 2019-01-29 中国计量大学 一种无线携能协作预编码空间调制系统的传输方法
CN109951239A (zh) * 2019-03-12 2019-06-28 南京邮电大学 基于贝叶斯分类器的能量采集中继系统自适应调制方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
CHIKHA, W.B.; ATTIA, R.: "On the Performance Evaluation of Bayesian Network Classifiers in Modulation Identification for Cooperative MIMO Systems", 2015 23RD INTERNATIONAL CONFERENCE ON SOFTWARE, TELECOMMUNICATIONS AND COMPUTER NETWORKS (SOFTCOM), 2 November 2015 (2015-11-02), XP032803789, DOI: 20191023103133A *

Also Published As

Publication number Publication date
US20210399815A1 (en) 2021-12-23
CN109951239A (zh) 2019-06-28
CN109951239B (zh) 2020-05-19
US11323192B2 (en) 2022-05-03

Similar Documents

Publication Publication Date Title
CN106993320B (zh) 基于多中继多跳的无线传感器网络协作传输路由方法
CN103052129B (zh) 一种无线多跳中继网络中节能路由及功率分配方法
CN109951239B (zh) 基于贝叶斯分类器的能量采集中继系统自适应调制方法
CN108541001B (zh) 一种用于能量可收集双向协作通信的中断率优化方法
CN111491358A (zh) 基于能量采集的自适应调制和功率控制系统与优化方法
CN110267294A (zh) 基于能量协作的随机中继选择方法
CN114520989B (zh) 多载波数能同传noma网络能效最大化方法
CN112788629B (zh) 一种能量收集通信系统功率和调制方式在线联合控制方法
Zhao et al. MLRS-RL: An energy-efficient multilevel routing strategy based on reinforcement learning in multimodal UWSNs
CN109714806A (zh) 一种非正交多址接入的无线供电中继网络优化方法
CN107135469B (zh) 一种分布式用户接入方法
CN106102117B (zh) 两跳有缓存网络中基于能量收割的安全传输方法
CN103781166B (zh) 异构无线网络协作通信系统中的移动终端功率分配方法
US20050122912A1 (en) Link adaptation method using feedback information for wireless communication system
CN112073094A (zh) 基于ts策略的包含直接链路swipt中继系统中吞吐量优化方法
CN112333809A (zh) 一种基于Lyapunov框架的能量收集中继安全传输在线功率控制方法
CN108551383B (zh) 一种基于机器学习的能量采集中继系统自适应调制方法
CN117956586A (zh) 一种基于强化学习的时延敏感网络发射功率自适应分配方法
CN109041009B (zh) 一种车联网上行功率分配方法及装置
CN106301501B (zh) 一种联合编码调制的即时数据传输优化方法
Wei et al. A low-overhead energy-efficient ARQ protocol for wireless sensor networks
CN115226128A (zh) 一种恢复5g通信多模应急通信方法及系统
CN107872270B (zh) 一种基于最优阈值传输调度的中继节点选择方法
CN106454899A (zh) 一种无线mesh网络生存时间优化方法
CN113810933A (zh) 一种基于能量采集与用户移动性的缓存方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19919209

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19919209

Country of ref document: EP

Kind code of ref document: A1