WO2022236916A1 - 基于序列最小二乘的一致性时钟同步频率偏移估计方法 - Google Patents
基于序列最小二乘的一致性时钟同步频率偏移估计方法 Download PDFInfo
- Publication number
- WO2022236916A1 WO2022236916A1 PCT/CN2021/099979 CN2021099979W WO2022236916A1 WO 2022236916 A1 WO2022236916 A1 WO 2022236916A1 CN 2021099979 W CN2021099979 W CN 2021099979W WO 2022236916 A1 WO2022236916 A1 WO 2022236916A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- clock
- synchronization
- relative
- frequency offset
- nodes
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims abstract description 81
- 238000004891 communication Methods 0.000 claims abstract description 36
- 238000009826 distribution Methods 0.000 claims abstract description 12
- 230000008569 process Effects 0.000 claims description 24
- 239000011159 matrix material Substances 0.000 claims description 17
- 230000001360 synchronised effect Effects 0.000 claims description 9
- 230000000694 effects Effects 0.000 claims description 5
- 230000014509 gene expression Effects 0.000 claims description 5
- 238000003860 storage Methods 0.000 abstract description 9
- 230000001934 delay Effects 0.000 abstract description 4
- 230000006870 function Effects 0.000 description 20
- 230000001427 coherent effect Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 230000003993 interaction Effects 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 230000000737 periodic effect Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 231100001261 hazardous Toxicity 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 238000009776 industrial production Methods 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W56/00—Synchronisation arrangements
- H04W56/0035—Synchronisation arrangements detecting errors in frequency or phase
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W56/00—Synchronisation arrangements
- H04W56/004—Synchronisation arrangements compensating for timing error of reception due to propagation delay
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W84/00—Network topologies
- H04W84/18—Self-organising networks, e.g. ad-hoc networks or sensor networks
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Definitions
- the invention belongs to the technical field of wireless sensor networks, and relates to a method for estimating frequency offset of consistent clock synchronization based on sequence least squares.
- Wireless sensor network is a typical distributed wireless communication network. Due to its functions of information collection, data processing and wireless communication, it has very important theoretical significance and applications in the fields of environmental monitoring, medical health, industrial production and hazardous environments. value. Time synchronization is an important prerequisite for the effective application of wireless sensor networks. Various applications such as protocol operation, TDMA scheduling, energy management, and target positioning need to run on the basis of node time synchronization in the network. As the research core of multi-agent system cooperative control, consensus synchronization protocol is an effective method to solve the problem of distributed network cooperative control. Introducing consistency theory into wireless sensor network clock synchronization can improve the robustness and scalability of the synchronization method. Therefore, it has a good development prospect to study the clock synchronization method based on consistency to solve the distributed synchronization problem.
- random communication delays during clock synchronization are usually unavoidable.
- random communication delay can be modeled as Gaussian distribution, exponential distribution, gamma distribution, etc.
- the earlier proposed consistent clock synchronization method ignoring the influence of delay cannot effectively guarantee the convergence of synchronization between network nodes.
- the relative frequency offset estimation plays an important role in the coherent clock synchronization method, because the estimated value will be directly used in the compensation of the logic clock parameters, which will affect the synchronization accuracy and convergence performance of the coherent clock synchronization algorithm.
- the purpose of the present invention is to provide a method for estimating the frequency offset of consistent clock synchronization based on sequence least squares, aiming at the problem that consistent time synchronization cannot converge under any random communication delay and the frequency offset estimation process does not.
- the problem of making full use of clock information focusing on efficient and practical data processing optimization methods, considering the needs of high-precision synchronization and low-storage nodes, adopts the sequential least squares method to estimate the relative frequency offset, and applies the estimated results to clock synchronization based on consistency
- the logical clock parameter compensation is carried out in the method, which effectively ensures the convergence performance of the consistent time synchronization method in the presence of delay, and at the same time achieves the effect of adapting to different types of delay scenarios and improving the clock synchronization accuracy of the entire network.
- the present invention provides the following technical solutions:
- a method for estimating the frequency offset of consistent clock synchronization based on sequence least squares which is oriented to bounded random communication delay scenarios that obey arbitrary distribution, establishes a relationship model between clock information and delay between nodes, and fully considers the information received by nodes from All the clock information of the neighbors, build a clock parameter estimation model and a cost function based on the least squares principle, use the sequential least squares method to iteratively estimate the relative clock frequency offset between nodes, and use the consistent clock synchronization method to update the node's Logical clock parameters enable all nodes in the network to achieve global clock consistency in a fully distributed manner.
- the method specifically includes the following steps:
- S4 Use sequential least squares to iteratively estimate the relative clock frequency offset between nodes, and then use the consistent clock synchronization method to update the logical clock parameters of the nodes. After estimating the relative frequency offset between nodes, periodically repeat the relative frequency offset Estimation and logical clock parameter update operations until the logical clocks of the entire network nodes achieve the effect of synchronization.
- the established inter-node clock information and delay relationship model specifically includes: assuming that any sensor node i in the network periodically broadcasts the local clock at an interval T And related synchronous clock information, its neighbor node j receives and records its own local clock at the receiving time Get the communication delay relationship between nodes on the real time scale:
- step S2 the relative relationship of any set of local clock information is established according to the communication delay relationship, which specifically includes: after neighbor node j receives n+1 pieces of synchronous clock information from node i, it will obtain n+1 sets of clock information Local Clock Observations Establish the relative relationship of any set of local clock information according to the communication delay relationship:
- ⁇ ij and ⁇ ij represent the relative frequency offset and relative phase offset of node i relative to node j, respectively, and ⁇ j represents the local clock frequency offset of node j.
- a clock parameter estimation model and a cost function based on the least squares principle are constructed, specifically including: According to the relative relationship of clock information, the communication delay item is processed to reduce the impact of delay on parameter estimation and clock synchronization , consider the delay part in the relative relationship of local clock information as an error function:
- Relative phase offset estimates and relative frequency offset estimates are obtained by minimizing a cost function.
- the relative frequency offset estimation based on the principle of least squares is the value obtained by minimizing the cost function J( ⁇ ij , ⁇ ij ), the cost function includes the clock information of all received records, synchronized with time The clock information to be stored by the process node more and more. For sensor nodes with limited storage capacity, it does not have enough memory space to save all clock information.
- the sequential least square method is used to iteratively estimate the relative clock frequency offset between nodes to reduce the storage overhead of sensor nodes; specifically, The following steps:
- node j receives and stores two sets of clock information and Directly use the standard least squares method to obtain estimates during the first round of synchronization and
- the covariance matrix during the first round of synchronization is set as:
- K(m) represents the gain matrix of the m-th round
- ⁇ (m) represents the covariance matrix of the m-th round parameter estimation
- I2 is the second-order identity matrix
- the logic clock parameters include: logic frequency offset compensation and logic phase offset compensation.
- the present invention considers the bounded communication time delay obeying any distribution type, utilizes the time delay relationship between nodes to establish a clock information model, and adopts the sequence least squares method applicable to any distribution to estimate the relative frequency offset, Improved the robustness of coherent clock synchronization to multiple types of delays.
- the present invention directly processes all clock information based on the principle of least squares, constructs a cost function through the relationship between each pair of clock information, and obtains a higher-precision relative frequency by minimizing the cost function including all clock information
- the offset estimation value effectively guarantees the convergence performance of the consistent clock synchronization.
- the present invention focuses on using the sequential least squares method to deal with the scene where the storage capacity of the node is limited and cannot store all the recorded clock information.
- the node does not need to store all the clock information received, but only needs to store the clock information received in the current synchronization round, the relevant estimated value in the previous round of synchronization process, and the covariance calculated in the last round Matrix, under the premise of obtaining high-precision clock parameter estimation, can also effectively reduce the storage and computing costs of nodes.
- Figure 1 is a distributed wireless sensor network communication topology
- Figure 2 is a schematic diagram of clock synchronization information interaction between nodes under delay
- FIG. 3 is a flow chart of a consistent synchronization method based on a sequence least squares frequency offset estimation method in this embodiment.
- Fig. 1 is the distributed wireless sensor network communication topological figure that the present invention considers, as shown in Fig. 1, each sensor node randomly distributed in the wireless sensor network periodically broadcasts local clock information, simultaneously It also receives clock information broadcast by nodes within its communication range.
- the communication topology of a wireless sensor network can be represented as a strongly connected directed graph in Represents the set of sensor nodes in the network, Represents a collection of reliable communication links. Additionally, use to represent the set of all neighbor nodes within the communication range of node i, where (i,j) ⁇ means that node j can successfully receive the clock synchronization information from node i.
- FIG. 2 shows the process of clock synchronization information interaction between nodes under time delay.
- the schematic diagram takes the synchronous interaction between two adjacent nodes i and j on the local time scale and the real time scale as an example.
- Node i transmits periodically at an interval T local clock information
- its neighbor node j receives the clock information from node i, it saves the information and records its own current local clock
- the communication delay between nodes i and j is modeled as the sending moment of node i on the real time scale and the receiving moment of node j on the real time scale difference:
- neighbor node j establishes a clock information pair according to the delay model Relative relationship:
- ⁇ ij and ⁇ ij represent the relative frequency offset and relative phase offset of node i relative to node j, respectively, and ⁇ j represents the local clock frequency offset of node j.
- the time delay part in the clock relational expression is regarded as an error function
- the cost function J( ⁇ ij , ⁇ ij ) can be obtained by extending the error function to the recorded n+1 pairs of clock information, and using the standard least square method to process the errors corresponding to all clock information at the same time:
- the relative frequency offset estimate obtained It has relatively high precision, so that it can effectively guarantee the convergence performance of the algorithm in the presence of communication delay when it is applied to the consistent clock synchronization algorithm.
- node j needs to record two clock information and Then use the above-mentioned standard least squares method to process the obtained and Expressions that compute the first-round parameter estimates:
- K(m) is the gain matrix of the m-th round
- ⁇ (m) represents the covariance matrix of the m-th round parameter estimation
- I2 is the second-order identity matrix.
- FIG. 3 is a flow chart of a method for estimating frequency offset of consistent clock synchronization based on sequence least squares in this embodiment.
- This embodiment provides a relative frequency offset estimation method based on sequence least squares for average consistency synchronization, as shown in FIG. 3 , which specifically includes the following steps:
- m2 ⁇ m4 Initialize clock synchronization parameters including relative clock frequency offset and relative clock phase offset, set the synchronous broadcast period, and judge whether the node meets the condition of periodic broadcast, if so, broadcast its own clock synchronization message, otherwise wait until the broadcast condition is met.
- m5 ⁇ m6 Neighbor nodes within the communication range receive synchronous clock information and record their own local clock values at the time of reception, and use this round of clock information to establish relative clock relationships between nodes under communication delay.
- Neighbor nodes process the delay part in the clock relationship, and construct error function and cost function by minimizing the influence of delay on relative frequency offset estimation.
- the node estimates the relative frequency offset and relative phase offset in the first round of synchronization process according to the cost function and the recorded clock information, and calculates the covariance matrix in the first round of synchronization iteration process ; If it is the m>1st round of synchronization process, the node uses the sequence least square method to iteratively estimate the relative frequency offset in this round of synchronization according to the clock information received in this round and the estimated information stored in the previous round of synchronization.
- m9 Update logic clock parameter compensation using a synchronization protocol based on average consistency, mainly including logic frequency offset compensation and logic phase offset compensation.
- m10 ⁇ m11 Determine whether the logical clocks of all nodes in the network meet the synchronization end conditions, including whether the logical frequency offset error is consistent, whether the logical phase offset error is maintained within a low range, and whether the logical clock error is within an acceptable range If the desired synchronization effect is achieved, it will end; otherwise, continue to monitor the clock synchronization information, update relative frequency offset estimation and logic clock parameter compensation and other synchronization processes until the synchronization end condition is met.
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Synchronisation In Digital Transmission Systems (AREA)
Abstract
一种基于序列最小二乘的一致性时钟同步频率偏移估计方法，属于无线传感器网络技术领域。该方法面向服从任意分布的有界随机通信时延场景，建立节点间时钟信息和时延的关系模型，充分考虑节点接收到的来自邻居的所有时钟信息，构建时钟参数估计模型和基于最小二乘原则的成本函数，采用序列最小二乘方法迭代估计节点间的相对时钟频率偏移，并利用一致性的时钟同步方法来更新节点的逻辑时钟参数，使得网络中全部节点以完全分布式的方式实现全局时钟的一致。该方法提高了相对频率偏移估计的精度，有效降低了节点的存储开销，提高了一致性同步算法对通信时延的鲁棒性。
Description
本发明属于无线传感器网络技术领域，涉及一种基于序列最小二乘的一致性时钟同步频率偏移估计方法。
无线传感器网络是一种典型的分布式无线通信网络，由于具备信息采集、数据处理和无线通信等功能，它在环境监测、医疗健康、工业生产以及危险环境等领域有非常重要的理论意义和应用价值。时间同步是无线传感器网络得以有效应用的重要前提，协议运行、TDMA调度、能量管理、目标定位等多种应用都需要在网络中节点时间保持同步的基础上运行。一致性同步协议作为多代理系统协同控制的研究核心，是解决分布式网络协同控制问题的有效方法。将一致性理论引入无线传感器网络时钟同步能提高同步方法的鲁棒性和扩展性，因此，研究基于一致性的时钟同步方法以解决分布式同步问题具有良好的发展前景。
在实际的无线传感器网络场景中，时钟同步过程中的随机通信时延通常是无法避免的。针对不同的应用场合，随机通信时延可被建模为高斯分布、指数分布、伽马分布等。在考虑通信时延存在的情况下，早期提出的忽略了时延影响的一致性时钟同步方法不能够有效保证网络节点间同步的收敛性。相对频偏估计在一致性时钟同步方法中起着重要的作用，因其估计值会被直接用于逻辑时钟参数补偿中，进而影响一致性时钟同步算法的同步精度和收敛性能。近年来，部分一致性时钟同步算法通过改进相对频偏估计方法，来抑制通信时延对时钟参数估计以及共识同步的影响，进而有效地解决了时延存在情况下同步无法收敛的问题。但是，这些一致性时钟同步方法有两方面的局限性：一是受限于具体的随机时延分布类型，无法很好地应用到时延多变的实际网络环境中；二是使用的频偏估计方法相对简单，无法充分利用节点所接收到的时钟信息来估计更精确的频偏，时间同步的性能也会受到限制。
因此，亟需一种新的能够解决无线传感器网络中一致性时钟同步问题的相对频偏估计方法，使其在任意有界的随机通信时延场景下，具有较低存储需求和较高估计精度。
发明内容
有鉴于此，本发明的目的在于提供一种基于序列最小二乘的一致性时钟同步频率偏移估计方法，针对存在任意随机通信时延下一致性时间同步无法收敛问题以及频偏估计过程中未充分利用时钟信息的问题，围绕高效实用的数据处理优化方法，考虑同步高精度和节点低存 储的需求，采用序列最小二乘方法来估计相对频偏，并应用估计结果到基于一致性的时钟同步方法中进行逻辑时钟参数补偿，有效地保证一致性时间同步方法在时延存在情况下的收敛性能，同时达到适应不同类型时延场景、提高全网时钟同步精度的效果。
为达到上述目的，本发明提供如下技术方案：
一种基于序列最小二乘的一致性时钟同步频率偏移估计方法，面向服从任意分布的有界随机通信时延场景，建立节点间时钟信息和时延的关系模型，充分考虑节点接收到的来自邻居的所有时钟信息，构建时钟参数估计模型和基于最小二乘原则的成本函数，采用序列最小二乘方法迭代估计节点间的相对时钟频率偏移，并利用一致性的时钟同步方法来更新节点的逻辑时钟参数，使得网络中全部节点以完全分布式的方式实现全局时钟的一致。该方法具体包括以下步骤：
S1：面向服从任意分布的有界随机通信时延场景，建立节点间时钟信息和时延关系模型；
S2：根据通信时延关系建立任意一组本地时钟信息的相对关系；
S3：根据时钟信息的相对关系，构建时钟参数估计模型和基于最小二乘原则的成本函数；
S4：采用序列最小二乘法迭代估计节点间的相对时钟频率偏移，然后利用一致性的时钟同步法来更新节点的逻辑时钟参数，估计出节点间的相对频偏后，周期性重复相对频偏估计以及逻辑时钟参数更新操作，直到整个网络节点的逻辑时钟达到同步的效果。
进一步，步骤S1中，建立的节点间时钟信息和时延关系模型，具体包括：假设网络中的任意传感器节点i都以间隔T周期性广播本地时钟
以及相关同步时钟信息，其邻居节点j接收并且记录接收时刻自身的本地时钟
得到在真实时间刻度下节点间的通信时延关系：
其中，
表示服从任意分布的、非负的、上界为常数D
d的随机通信时延。
进一步，步骤S2中，根据通信时延关系建立任意一组本地时钟信息的相对关系，具体包括：邻居节点j收到n+1个来自节点i的同步时钟信息后，它将获得n+1组本地时钟观测值
根据通信时延关系建立任意一组本地时钟信息的相对关系：
其中，α
ij和β
ij分别表示节点i相对于节点j的相对频率偏移和相对相位偏移，α
j表示节点j的本地时钟频率偏移。
进一步，步骤S3中，构建时钟参数估计模型和基于最小二乘原则的成本函数，具体包括： 根据时钟信息的相对关系，对通信时延项进行处理，降低时延对参数估计和时钟同步的影响，将本地时钟信息的相对关系式中时延部分考虑为一个误差函数：
扩展误差函数到所有本地时钟观测，然后应用最小二乘原理来处理误差，得到以下包含时钟参数α
ij和β
ij的成本函数J(α
ij,β
ij)：
通过最小化成本函数来获取相对相位偏移估计和相对频率偏移估计。
进一步，步骤S4中，基于最小二乘原理的相对频偏估计为最小化成本函数J(α
ij,β
ij)所求得的值，成本函数中包含了所有接收记录的时钟信息，随时间同步过程的进行节点需存储的时钟信息
越来越多。对于存储容量受限的传感器节点，它没有足够的内存空间来保存所有时钟信息，在这种场景下采用序列最小二乘法迭代估计节点间的相对时钟频率偏移，降低传感器节点存储开销；具体包括以下步骤：
S41：在第一轮同步中，节点j接收并存储两组时钟信息
和
直接使用标准最小二乘方法来获取第一轮同步过程中的估计值
和
此外，为了开始时钟参数迭代估计过程，第一轮同步过程中的协方差矩阵被设定为：
Σ(1)＝(H
T(1)H(1))
-1
其中，H(1)＝[h(0) h(1)]
T，
S42：对于第m(m＝2,3,4,...,n)轮同步过程，节点j只需要存储当前轮接收到的时钟信息
前一轮计算得到的时钟参数估计值
和
以及前一轮迭代产生的协方差矩阵Σ(m-1)，然后采用序列最小二乘方法，迭代估计此轮同步过程中的相对时钟频偏和相对时钟相偏；
估计更新：
增益更新：
协方差更新：Σ(m)＝(I
2-K(m)h
T(m))Σ(m-1)
其中，
表示第m轮时钟参数估计值，即
K(m)表示第m轮的增益矩阵，Σ(m)表示第m轮参数估计的协方差矩阵，I
2是二阶单位矩阵。由此可见，在基于序列最小二乘的频率偏移估计方法下，节点仅需存储当前同步轮次的时钟信息和前一轮所获得的估计信息。
进一步，步骤S4中，所述的逻辑时钟参数包括：逻辑频偏补偿和逻辑相偏补偿。
本发明的有益效果在于：
1)本发明考虑了服从任意分布类型的有界通信时延，利用节点间的时延关系建立了时钟信息模型，并采用适用于任何分布的序列最小二乘方法对相对频率偏移进行估计，提高了一致性时钟同步对多类型时延的鲁棒性。
2)本发明采用基于最小二乘原则，直接对所有的时钟信息进行处理，通过每对时钟信息的关系构造出成本函数，通过最小化包含所有时钟信息的成本函数获得了更高精度的相对频率偏移估计值，有效地保证了一致性时钟同步的收敛性能。
3)本发明重点考虑了使用序列最小二乘法来应对节点存储容量受限，无法存储所记录的所有时钟信息的场景，通过利用节点的相对时钟关系和标准最小二乘法可迭代计算的特点，使得节点在每一轮同步估计中无需存储接收到的所有时钟信息，而只需要存储当前同步轮次收到的时钟信息、上一轮同步过程中的相关估计值以及上一轮计算得到的协方差矩阵，在获取到高精度的时钟参数估计前提下，还能有效降低节点的存储和计算开销。
本发明的其他优点、目标和特征在某种程度上将在随后的说明书中进行阐述，并且在某种程度上，基于对下文的考察研究对本领域技术人员而言将是显而易见的，或者可以从本发明的实践中得到教导。本发明的目标和其他优点可以通过下面的说明书来实现和获得。
为了使本发明的目的、技术方案和优点更加清楚，下面将结合附图对本发明作优选的详细描述，其中：
图1为分布式的无线传感器网络通信拓扑图；
图2为时延下节点间时钟同步信息交互示意图；
图3为本实施例基于序列最小二乘频偏估计方法下的一致性同步方法流程图。
以下通过特定的具体实例说明本发明的实施方式，本领域技术人员可由本说明书所揭露的内容轻易地了解本发明的其他优点与功效。本发明还可以通过另外不同的具体实施方式加 以实施或应用，本说明书中的各项细节也可以基于不同观点与应用，在没有背离本发明的精神下进行各种修饰或改变。需要说明的是，以下实施例中所提供的图示仅以示意方式说明本发明的基本构想，在不冲突的情况下，以下实施例及实施例中的特征可以相互组合。
请参阅图1～图3，图1为本发明考虑的分布式的无线传感器网络通信拓扑图，如图1所示，无线传感器网络中随机分布的每个传感器节点周期性广播本地时钟信息，同时又接收其通信范围内的节点广播的时钟信息。无线传感器网络的通信拓扑可以被表示为一个强连通的有向图
其中
表示网络中的传感器节点集合，
代表可靠的通信链路集合。此外，使用
来表示节点i的通信范围内的所有邻居节点集合，其中(i,j)∈ε表示节点j能成功接收来自节点i的时钟同步信息。
图2表示时延下节点间时钟同步信息交互示过程，示意图以两个相邻节点i和j在本地时间尺度上和真实时间尺度上的同步交互情况为例，节点i以间隔T周期性传递本地时钟信息
其邻居节点j接收到来自节点i的时钟信息就保存该信息并记录下自身的当前本地时钟
对于第k轮传递时钟信息的过程，将节点i和j之间的通信时延建模为节点i在真实时间尺度上的发送时刻
和节点j在真实时间尺度上的接收时刻
的差距：
其中，
被考虑为一个正的、上界为固定值D
d的随机变量，通常将此上界设置为可测量的最大的往返时间。于是，邻居节点j根据时延模型建立时钟信息对
的相对关系：
其中，α
ij和β
ij分别表示节点i相对于节点j的相对频偏和相对相偏，α
j表示节点j的本地时钟频偏。为减小时延对相对频偏估计的影响，将时钟关系式中的时延部分看作一个误差函数
扩展误差函数到记录的n+1对时钟信息，利用标准最小二乘法来同时处理所有的时钟信息对应的误差即可获取成本函数J(α
ij,β
ij)：
接着是对成本函数进行最小化处理来获取参数估计值的过程，先求取J(α
ij,β
ij)相对于未知 参数α
ij和β
ij的偏导，令求取的偏导表达式结果等于零，联立两个偏导等式和周期广播时刻
进行求解得到估计
和
因为考虑到了所有记录过的时钟信息，因此获取的相对频偏估计
具有相对较高的精度，使得其应用于一致性时钟同步算法中能够有效地保证算法在通信时延存在下的收敛性能。
从标准最小二乘法推导得出的相对频偏估计表达可以看到，节点需要存储所记录的全部时钟信息用于参数估计，且需要的计算量相对繁杂。显然，随着同步轮次的增加，接收到的同步时钟信息数量会随之增多，对于传感器节点存储容量有限和计算能力受限、不能够存储所记录的全部时钟信息的场景，需要特别考虑降低存储开销，设计一种完全等效的序列最小二乘估计方法。由于在成本函数中存在两个参数需要估计，因此至少需要两个时钟信息作为估计的初始化点，同时为方便在序列最小二乘估计过程中使用将待估计参数记为矢量形式θ＝[α
ij β
ij]
T。在第一轮同步过程中，节点j需要记录两个时钟信息
和
然后利用上述标准最小二乘法处理得到的
和
表达式，可以计算第一轮参数估计值：
其中，
表示第一轮时钟参数估计值。而在第m(m＝2,3,4,...,n)轮同步过程中，节点j只需要存储当前轮接收到的时钟信息
以及前一轮计算得到的估计值
以及用于迭代估计过程中的协方差矩阵，根据序列最小二乘方法迭代估计此轮的相对时钟频偏和相对时钟相偏：
估计更新：
增益：
协方差更新：Σ(n)＝(I
2-K(n)h
T(n))Σ(n-1)
其中，
表示第m轮时钟参数估计值，
K(m)是 第m轮的增益矩阵,Σ(m)表示第m轮参数估计的协方差矩阵，I
2是二阶单位矩阵。此外，为了开始迭代地进行时钟参数估计，第一轮的协方差矩阵给定为Σ(1)＝(H
T(1)H(1))
-1，其中H(1)＝[h(0) h(1)]
T。由此可以明显看到，使用序列最小二乘法来估计相对频偏时，节点仅需要存储少量的时钟信息和相关的迭代信息。
实施例1：
图3为本实施例基于序列最小二乘的一致性时钟同步频率偏移估计方法流程图。本实施例提供了一种用于平均一致性同步的基于序列最小二乘的相对频偏估计方法，如图3所示，具体包括以下步骤：
m1：时钟同步过程开始。
m2～m4：初始化时钟同步参数包括相对时钟频偏和相对时钟相偏，设置同步广播周期，节点判断是否满足周期性广播的条件，若是则广播自身的时钟同步消息，否则等待直到广播条件满足。
m5～m6：通信范围内的邻居节点接收同步时钟信息并记录接收时刻自己的本地时钟值，利用此轮时钟信息建立通信时延下节点间的相对时钟关系。
m7：邻居节点对时钟关系中的时延部分进行处理，通过最小化时延对相对频偏估计的影响构造误差函数和成本函数。
m8：如果是第一轮同步过程，节点根据成本函数和记录的时钟信息估计第一轮同步过程中的相对频率偏移和相对相位偏移，并计算第一轮同步迭代过程中的协方差矩阵；如果是第m＞1轮同步过程，节点根据此轮接收到的时钟信息和上一轮同步存储的估计信息，采用序列最小二乘方法迭代估计此轮同步下的相对频率偏移。
m9：采用基于平均一致性的同步协议更新逻辑时钟参数补偿，主要包括逻辑频偏补偿和逻辑相偏补偿。
m10～m11：判断网络中所有节点的逻辑时钟是否满足同步结束条件，包括逻辑频偏误差是否一致、逻辑相偏误差是否维持在一个较低的范围内、逻辑时钟误差是否在一个可接受的范围内，若达到期望的同步效果则结束，否则继续监听时钟同步信息、更新相对频偏估计和逻辑时钟参数补偿等同步过程，直到满足同步结束的条件。
在估计出节点间的相对频偏后，将其应用到一致性同步方法中补偿节点的逻辑时钟参数，比如基于平均一致性协议的频偏补偿和相偏补偿。周期性重复相对频偏估计、逻辑频偏补偿和逻辑相偏补偿过程，直到分布式网络中的所有节点的逻辑时钟达到同步。
最后说明的是，以上实施例仅用以说明本发明的技术方案而非限制，尽管参照较佳实施 例对本发明进行了详细说明，本领域的普通技术人员应当理解，可以对本发明的技术方案进行修改或者等同替换，而不脱离本技术方案的宗旨和范围，其均应涵盖在本发明的权利要求范围当中。
Claims (6)
- 一种基于序列最小二乘的一致性时钟同步频率偏移估计方法，其特征在于，该方法具体包括以下步骤：S1：面向服从任意分布的有界随机通信时延场景，建立节点间时钟信息和时延关系模型；S2：根据通信时延关系建立任意一组本地时钟信息的相对关系；S3：根据时钟信息的相对关系，构建时钟参数估计模型和基于最小二乘原则的成本函数；S4：采用序列最小二乘法迭代估计节点间的相对时钟频率偏移，然后利用一致性的时钟同步法来更新节点的逻辑时钟参数，估计出节点间的相对频偏后，周期性重复相对频偏估计以及逻辑时钟参数更新操作，直到整个网络节点的逻辑时钟达到同步的效果。
- 根据权利要求1所述的一致性时钟同步频率偏移估计方法，其特征在于，步骤S1中，建立的节点间时钟信息和时延关系模型，具体包括：假设网络中的任意传感器节点i都以间隔T周期性广播本地时钟以及相关同步时钟信息，其邻居节点j接收并且记录接收时刻自身的本地时钟
得到在真实时间刻度下节点间的通信时延关系：
其中，表示服从任意分布的、非负的、上界为常数D d的随机通信时延。
- 根据权利要求2所述的一致性时钟同步频率偏移估计方法，其特征在于，步骤S2中，根据通信时延关系建立任意一组本地时钟信息的相对关系，具体包括：邻居节点j收到n+1个来自节点i的同步时钟信息后，它将获得n+1组本地时钟观测值根据通信时延关系建立任意一组本地时钟信息的相对关系：
其中，α ij和β ij分别表示节点i相对于节点j的相对频率偏移和相对相位偏移，α j表示节点j的本地时钟频率偏移。 - 根据权利要求3所述的一致性时钟同步频率偏移估计方法，其特征在于，步骤S3中，构建时钟参数估计模型和基于最小二乘原则的成本函数，具体包括：根据时钟信息的相对关系，对通信时延项进行处理，将本地时钟信息的相对关系式中时延部分考虑为一个误差函数：
扩展误差函数到所有本地时钟观测，然后应用最小二乘原理来处理误差，得到以下包含时钟参数α ij和β ij的成本函数J(α ij,β ij)：
通过最小化成本函数来获取相对相位偏移估计和相对频率偏移估计。 - 根据权利要求4所述的一致性时钟同步频率偏移估计方法，其特征在于，步骤S4中，采用序列最小二乘法迭代估计节点间的相对时钟频率偏移，具体包括以下步骤：S41：在第一轮同步中，节点j接收并存储两组时钟信息和
直接使用标准最小二乘方法来获取第一轮同步过程中的估计值
和
此外，为了开始时钟参数迭代估计过程，第一轮同步过程中的协方差矩阵被设定为：Σ(1)＝(H T(1)H(1)) -1其中，H(1)＝[h(0)h(1)] T，
S42：对于第m(m＝2,3,4,...,n)轮同步过程，节点j只需要存储当前轮接收到的时钟信息前一轮计算得到的时钟参数估计值
和
以及前一轮迭代产生的协方差矩阵Σ(m-1)，然后采用序列最小二乘方法，迭代估计此轮同步过程中的相对时钟频偏和相对时钟相偏；
估计更新：
增益更新：
协方差更新：Σ(m)＝(I 2-K(m)h T(m))Σ(m-1)其中，表示第m轮时钟参数估计值，即
K(m)表示第m轮的增益矩阵，Σ(m)表示第m轮参数估计的协方差矩阵，I 2是二阶单位矩阵。
- 根据权利要求1所述的一致性时钟同步频率偏移估计方法，其特征在于，步骤S4中，所述的逻辑时钟参数包括：逻辑频偏补偿和逻辑相偏补偿。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US17/775,298 US20240163821A1 (en) | 2021-05-10 | 2021-06-15 | Consistent clock synchronization frequency offset estimation method based on sequential least squares |
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110506223.5 | 2021-05-10 | ||
| CN202110506223.5A CN113207167B (zh) | 2021-05-10 | 2021-05-10 | 基于序列最小二乘的一致性时钟同步频率偏移估计方法 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2022236916A1 true WO2022236916A1 (zh) | 2022-11-17 |
Family
ID=77030512
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2021/099979 WO2022236916A1 (zh) | 2021-05-10 | 2021-06-15 | 基于序列最小二乘的一致性时钟同步频率偏移估计方法 |
Country Status (3)
| Country | Link |
|---|---|
| US (1) | US20240163821A1 (zh) |
| CN (1) | CN113207167B (zh) |
| WO (1) | WO2022236916A1 (zh) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114423073B (zh) * | 2022-01-17 | 2023-05-23 | 重庆邮电大学 | 基于事件触发的平均一致性时间同步方法 |
| CN115103432B (zh) * | 2022-06-06 | 2023-05-23 | 重庆邮电大学 | 随机移动无线网络下基于事件触发的时间同步方法 |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150092640A1 (en) * | 2013-09-30 | 2015-04-02 | Honeywell International Inc. | Wireless sensor time synchronization |
| US20180041329A1 (en) * | 2015-04-29 | 2018-02-08 | Chongqing University Of Posts And Telecommunications | Clock synchronization frequency offset estimation method adapted to wia-pa network |
| CN110505683A (zh) * | 2019-09-23 | 2019-11-26 | 重庆邮电大学 | 一种面向平均一致性时钟同步的频率偏移估计方法 |
| CN111416785A (zh) * | 2020-02-25 | 2020-07-14 | 重庆邮电大学 | 基于加权中值的一致性时钟同步相对频偏估计方法 |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102480350B (zh) * | 2010-11-29 | 2014-10-15 | 北京中和威软件有限公司 | 一种基于频率偏差估计的时间同步和维持的方法及装置 |
| US9967112B2 (en) * | 2012-05-03 | 2018-05-08 | Apple Inc. | Apparatus and methods for adaptively resuming radio channel estimation |
| CN103095638B (zh) * | 2012-12-19 | 2016-06-08 | 西安电子科技大学 | 一种多径衰落信道下ofdm系统的采样频率偏移盲估算方法 |
| CN108449791B (zh) * | 2018-03-19 | 2020-12-11 | 重庆邮电大学 | 一种基于温度补偿的自适应时间同步方法 |
-
2021
- 2021-05-10 CN CN202110506223.5A patent/CN113207167B/zh active Active
- 2021-06-15 WO PCT/CN2021/099979 patent/WO2022236916A1/zh active Application Filing
- 2021-06-15 US US17/775,298 patent/US20240163821A1/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150092640A1 (en) * | 2013-09-30 | 2015-04-02 | Honeywell International Inc. | Wireless sensor time synchronization |
| US20180041329A1 (en) * | 2015-04-29 | 2018-02-08 | Chongqing University Of Posts And Telecommunications | Clock synchronization frequency offset estimation method adapted to wia-pa network |
| CN110505683A (zh) * | 2019-09-23 | 2019-11-26 | 重庆邮电大学 | 一种面向平均一致性时钟同步的频率偏移估计方法 |
| CN111416785A (zh) * | 2020-02-25 | 2020-07-14 | 重庆邮电大学 | 基于加权中值的一致性时钟同步相对频偏估计方法 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113207167B (zh) | 2022-03-08 |
| CN113207167A (zh) | 2021-08-03 |
| US20240163821A1 (en) | 2024-05-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2022236916A1 (zh) | 基于序列最小二乘的一致性时钟同步频率偏移估计方法 | |
| CN111416785B (zh) | 基于加权中值的一致性时钟同步相对频偏估计方法 | |
| CN110505683B (zh) | 一种面向平均一致性时钟同步的频率偏移估计方法 | |
| Luo et al. | Distributed clock parameters tracking in wireless sensor network | |
| CN110460553B (zh) | 一种免时间戳交互的隐含节点时钟频率偏移估计方法 | |
| Djenouri et al. | Fast distributed multi-hop relative time synchronization protocol and estimators for wireless sensor networks | |
| Huang et al. | Long term and large scale time synchronization in wireless sensor networks | |
| Chaudhari | A simple and robust clock synchronization scheme | |
| Luo et al. | Fully distributed clock synchronization in wireless sensor networks under exponential delays | |
| Wang et al. | Clock synchronization with partial timestamp information for wireless sensor networks | |
| Swain et al. | A weighted average–based external clock synchronisation protocol for wireless sensor networks | |
| Upadhyay et al. | Maximum probable clock offset estimation (MPCOE) to reduce time synchronization problems in wireless sensor networks | |
| Ting et al. | Clock synchronization in wireless sensor networks: a new model and analysis approach based on networked control perspective | |
| Giridhar et al. | The spatial smoothing method of clock synchronization in wireless networks | |
| Wang et al. | Distributed parameter estimation in unreliable WSNs: Quantized communication and asynchronous intermittent observation | |
| Yildirim et al. | Drift estimation using pairwise slope with minimum variance in wireless sensor networks | |
| CN111385825B (zh) | 一种无线传感器网络快速收敛的一致性时间同步方法 | |
| He et al. | Clock synchronization for random mobile sensor networks | |
| Niu et al. | Consensus tracking-based clock synchronization for the Internet of Things | |
| Yang | Time synchronization in wireless sensor networks: a survey | |
| Sun et al. | Long-term clock synchronization in wireless sensor networks with arbitrary delay distributions | |
| Gou et al. | Low energy consumption and high-precision time synchronization algorithm based on improved tpsn in wireless sensor networks | |
| Bagherizadeh et al. | A high performance and secure way to time synchronization in wireless sensor network | |
| Liu et al. | An on-demand global time synchronization based on data analysis for wireless sensor networks | |
| Chen et al. | Clock synchronization for distributed multi-hop wireless networks using Markov Random field |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| WWE | Wipo information: entry into national phase |
Ref document number: 17775298 Country of ref document: US |
|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 21941485 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: 21941485 Country of ref document: EP Kind code of ref document: A1 |