KR20140062592A - Distributed frequency synchronization method and apparatus in wireless mesh network - Google Patents

Abstract

The present invention provides a distributed frequency synchronization method and a device thereof in a wireless mesh network are provided. According to the present invention, a distributed frequency synchronization method in a wireless mesh network includes: a step of receiving a preamble from at least one neighbor node which are existing within a predetermined radius at the current frequency update time and collecting neighbor carrier frequency offset information; a step of determining a carrier frequency offset compensation value of transmission data in the next frequency update time using the collected neighbor carrier frequency offset information; and a step of repetitively performing the steps until the neighbor carrier frequency offset information becomes less than a predetermined threshold value.

Description

[0001] DISTRIBUTED FREQUENCY SYNCHRONIZATION METHOD AND APPARATUS IN WIRELESS MESH NETWORK [0002]

The present invention relates to a distributed frequency synchronization method and apparatus in a wireless mesh network, and more particularly, to a method and apparatus for frequency synchronization in a mesh network in which a plurality of nodes having different carrier frequencies exist.

In OFDMA-based wireless mesh networks (WMN), carrier frequency offsets are generated by modulation frequency error of an oscillator, mobility of a node, estimation of a frequency offset by a channel noise, Frequency Offset (CFO), which is a major problem that degrades overall system performance. In a wireless mesh network, transmission power is adjusted through a ranging process as in a conventional cellular communication system in order to minimize inter-node interference. However, in the case of a network requiring wide coverage, the NFP (Near-Far Problem) due to the power difference between nodes can not be completely overcome because the transmission / reception time is different for each node by the mesh selection algorithm and the distributed scheduling Therefore, the effect of the carrier frequency offset is more serious.

In order to solve such frequency asynchronism problem, researches on frequency synchronization technique for minimizing the influence of carrier frequency offset occurring between transceivers have been carried out for a long time in academia. However, in the conventional method, it is difficult to fundamentally solve the problem of multiple carrier frequency offsets (multiple CFO) of OFDMA-based mesh networks in which signals having different carrier frequency offsets from a plurality of transmitting terminals are combined to arrive at a receiver.

In order to examine the effect of carrier frequency offset in an OFDMA-based wireless mesh network, a simple mesh network composed of four nodes is assumed as shown in FIG.

은 노드 간 발생하는 normalized CFO를 의미하고,

의 아래 첨자는 네트워크 엔트리 과정을 통해 할당된 각 노드의 노드 고유 번호를 의미한다. 즉,

은 노드 1에서 측정된 노드 2와의 상대적인 normalized CFO를 의미한다. 또한, 노드 2에서 측정된

는 그 크기가 같고, 부호는 반대가 된다. It is assumed that the nodes present in the wireless mesh network of FIG. 1 have different oscillator modulation frequencies. here

Is the normalized CFO occurring between nodes,

The subscript of the node means the node unique number of each node allocated through the network entry process. In other words,

Means the normalized CFO relative to node 2 measured at node 1. In addition,

Are the same in size, and the sign is opposite.

In the network of FIG. 1, node 1 is capable of receiving data simultaneously through resource allocation from its one hop node, nodes 2, 3, and 4. In this case, however, ICI (Inter-Carrier Interference) is generated due to the carrier frequency offset of the remaining nodes even if the carrier frequency offset of one node is compensated for Which can lead to serious deterioration of system performance.

FIG. 2 shows an example of orthogonal resource allocation received at node 1 of FIG. Here, it is assumed that the carrier frequency offset estimation for the adjacent node at node 1 is completely performed. For example, if node 1 compensates for a carrier frequency offset corresponding to node 2 to a signal received simultaneously from neighboring nodes, the message at node 2 can be received without affecting the carrier frequency offset. However, since the resource blocks of the other adjacent nodes, nodes 3 and 4, are still influenced by the carrier frequency offset, the ICI occurs, and the messages of the two nodes can not be completely received.

In addition, the node 2 to which the frequency compensation is performed may also suffer from a bit error rate (BER) performance degradation due to Multi-User Interference (MUI) from neighboring resources occurring in the carrier frequency offset compensation process. The influence of the carrier frequency offset occurring between adjacent nodes in an OFDMA system can be expressed by the following equation.

번째 시간 샘플을 나타낸다. 여기서

은 전송된 노드의 수, 는

보호구간의 길이,

은 FFT의 크기,

는 수신 노드와

번 노드와의 반송파 주파수 옵셋,

은 AWGN(Additive White Gaussian Noise) 잡음을 의미한다. Equation (1) represents the signal received at a particular node from neighboring nodes having different carrier frequency offsets

Lt; th > time sample. here

Is the number of transmitted nodes,

The length of the protection zone,

The size of the FFT,

Lt; / RTI >

Carrier frequency offset with node #

Means additive white Gaussian noise (AWGN) noise.

번째 노드의 반송파 주파수 옵셋이 보상된 신호를 나타낸다. 마지막 식의 오른쪽 첫 번째 항은

번째 노드로부터 송신된 신호이고, 두 번째 항은 MUI, 세 번째 항은 잡음을 의미한다. 이로부터, 여러 노드로부터 송신된 신호를 하나의 주파수로 보상하여 수신하게 되면, MUI의 영향으로 인해 BER 성능의 열화가 발생하게 됨을 확인할 수 있다.Equation (2) also indicates that a signal received from several nodes

The carrier frequency offset of the ith node represents the compensated signal. The first term to the right of the last expression

The second term is the MUI, and the third term is the noise. From this, it can be confirmed that the BER performance deteriorates due to the influence of the MUI if the signals transmitted from the various nodes are compensated with one frequency and received.

In order to solve the interference problem due to the multi-carrier frequency offset, in the uplink of the OFDMA-based cellular system, the base station separately compensates the frequency offset corresponding to each resource block allocated after the DFT (Discrete Fourier Transform) (Choi-Lee-Jung-Lee) and HL (Huang-Letaief) techniques have been published. However, as the number of UEs increases with the number of UEs, these schemes are difficult to implement due to the computational complexity problem due to the increase in the number of convolution operations, and the influence of the carrier frequency offset still remaining after compensation increases. It is fundamentally difficult to solve.

Disclosure of Invention Technical Problem [8] In order to solve the above problems, the present invention proposes a distributed frequency synchronization method and apparatus in a wireless mesh network that can solve the interference problem due to multi-carrier frequency offsets.

According to a preferred embodiment of the present invention, there is provided a distributed frequency synchronization method in a wireless mesh network, comprising: (a) receiving a preamble from one or more neighboring nodes within a predetermined radius at a current frequency update time Collecting peripheral carrier frequency offset information; (b) determining a carrier frequency offset compensation value of transmission data at a next frequency update time using the collected peripheral carrier frequency offset information; And (c) repeating the steps (a) and (b) until the peripheral carrier frequency offset information is less than or equal to a preset threshold value.

The predetermined radius may be one of a one-hop distance and a maximum communicable link distance between nodes.

The step (b) may determine a carrier frequency offset compensation value of the transmission data using a fl oating algorithm.

Further comprising the step of determining a weight to be applied to carrier frequency offsets of the at least one neighboring node using the collected carrier frequency offset information prior to step (b), wherein the step (b) The carrier frequency offset compensation value of the transmission data may be determined using the frequency offset information and the weight value.

The carrier frequency offset compensation value of the transmission data may be determined by the following equation.

[Mathematical Expression]

은 노드 번호 n인 노드의 원 홉 노드 수,

은 k번째 갱신 시점(k번째 주파수 갱신 타임)에 n번 노드에서 추정된 인접 노드 i의 반송파 주파수 옵셋,

은 n번 노드의 다음 주파수 갱신 타임(k+1번째 주파수 갱신 타임)에서의 송신 데이터 반송파 주파수 옵셋 보상값,

는 네트워크의 주파수 수렴 특성을 고려하여 i번 노드의 반송파 주파수 옵셋에 적용되는 가중치임Where k is the number of frequency updates, n is the unique number of a node in the network,

The number of one-hop nodes of the node with the node number n,

Is the carrier frequency offset of the neighbor node i estimated at the nth node at the kth update time (kth frequency update time)

Is the transmission data carrier frequency offset compensation value at the next frequency update time (k + 1 < th > frequency update time) of the node n,

Is a weight applied to the carrier frequency offset of the node i in consideration of the frequency convergence characteristic of the network

Wherein the wireless mesh network includes two groups that converge at different frequencies and the weight determination step determines a bridge node that connects the two groups of the one or more peripheral nodes using the peripheral carrier frequency offset information ; And assigning different weights to the determined nodes other than the bridge node and the bridge node.

The weight of the bridge node may be determined as a ratio of the number of the at least one neighboring node.

And determining a moving direction of the frequency information around the bridge node.

According to another aspect of the present invention, there is provided a computer-readable recording medium on which a program for performing the above-described method is recorded.

According to another aspect of the present invention, there is provided a distributed frequency synchronization apparatus in a wireless mesh network, comprising: a frequency information collecting unit that receives a preamble from one or more neighboring nodes existing within a predetermined radius at a current frequency update time and collects peripheral carrier frequency offset information; part; And a compensation value determiner for determining a carrier frequency offset compensation value of transmission data at a next frequency update time using the collected peripheral carrier frequency offset information, wherein the peripheral carrier frequency offset information is less than or equal to a preset threshold value And the carrier frequency offset compensation value determination process is repeatedly performed until the carrier frequency offset offset information is collected.

According to the present invention, since the frequency of each node is adjusted by precompensating and transmitting the carrier frequency offset to the next data by an appropriate value during the data transmission, the carrier frequency of all the nodes existing in the network after a certain time has a predetermined value As shown in FIG.

1 illustrates a wireless mesh network model in which there are different carrier frequency offsets among a plurality of nodes.
2 is a resource block diagram that is compensated with one carrier frequency offset at node 1 of FIG.
3 is a flowchart of a distributed frequency synchronization process according to a preferred embodiment of the present invention.
Figure 4 illustrates an example of various mesh topologies composed of multiple nodes.
5 illustrates an example of a carrier frequency offset distribution estimated in two groups of bridge nodes when two groups are present;
FIG. 6 illustrates frequency convergence characteristics for each node in each topology of FIG. 4; FIG.
FIG. 7 is a view showing frequency convergence characteristics for each node according to a weight applying method (frequency movement direction) in each topology of FIG. 4; FIG.
8 shows frequency synchronization times before and after weight application.
9 is a diagram showing a change in frequency synchronization time for each topology;
10 is a block diagram of a node performing a distributed frequency synchronization process according to the present invention;

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. In order to facilitate a thorough understanding of the present invention, the same reference numerals are used for the same means regardless of the number of the drawings.

The present invention relates to frequency synchronization for a mesh network in which a plurality of nodes having different carrier frequencies exist, and a method for local frequency synchronization of nodes existing in the network using a flocking algorithm, We propose a distributed frequency synchronization algorithm that unifies the entire network to one specific frequency.

In addition, we propose an adaptive weighting method to overcome the problem of synchronization time difference per topology that occurs in the process of extending it to various topologies.

Hereinafter, the frequency synchronization process in the mesh network according to the present invention will be described in detail.

3 is a flowchart of a frequency synchronization process according to an exemplary embodiment of the present invention.

내에 존재하는 모든 노드들로부터 수신되는 프리앰블 신호를 이용하여 주변의 반송파 주파수 옵셋 정보를 수집한다(단계 300). FIG. 3 is a process performed by one node belonging to a wireless mesh network, and each node existing in the mesh network according to the present invention has its radius

The neighboring carrier frequency offset information is collected using the preamble signal received from all nodes existing in the neighboring node (step 300).

은 시스템에서 정의된 원 홉 거리 또는 신호의 크기 및 분산 네트워크 스케줄링 등을 고려한 노드 간 최대 통신 가능 링크거리일 수 있다. here,

May be a one-hop distance defined in the system, or a maximum communication link distance between nodes considering signal size and distributed network scheduling.

Each node can be allocated a region to transmit / receive its own preamble signal by a mesh selection algorithm. Therefore, it is possible to estimate a carrier frequency offset through a preamble with nodes at a one-hop distance of the node without collision between nodes.

It is assumed that accurate carrier frequency offset estimation has been performed on neighboring nodes through sufficient preamble repeat reception.

Then, each node determines a transmission data frequency offset compensation value at the next frequency update time using the collected carrier frequency offset information as described below (step 302).

Based on the Cucker-Smale model, which is one of the representative models of the flocking algorithm, we propose the frequency synchronization formula for the mesh network as follows.

은 노드 번호 n인 노드의 원 홉 노드 수,

은 k번째 갱신 시점(k번째 주파수 갱신 타임)에 n번 노드에서 추정된 인접 노드 i의 반송파 주파수 옵셋,

은 n번 노드의 다음 주파수 갱신 타임(k+1번째 주파수 갱신 타임)에서의 송신 데이터 반송파 주파수 옵셋 보상값,

는 네트워크의 주파수 수렴 특성을 고려하여 i번 노드의 반송파 주파수 옵셋에 적용되는 가중치를 나타낸다. Where k is the number of frequency updates, n is the unique number of a node in the network,

The number of one-hop nodes of the node with the node number n,

Is the carrier frequency offset of the neighbor node i estimated at the nth node at the kth update time (kth frequency update time)

Is the transmission data carrier frequency offset compensation value at the next frequency update time (k + 1 < th > frequency update time) of the node n,

Denotes the weight applied to the carrier frequency offset of node i in consideration of the frequency convergence characteristic of the network.

The weight of the algorithm according to the present invention should be appropriately selected in consideration of the fact that the convergence of the flocking algorithm is established when the weight function determining the force between individuals in the Cucker-Smale model is a decreasing function depending on the distance of the object, Will be described later.

) 이하가 된다면 노드 간 통신이 문제없이 이루어질 수 있다.Also, in an actual mesh network, even if the carrier frequency offset occurring between nodes is not necessarily 0, a certain threshold value

), The inter-node communication can be performed without any problem.

이하로 수렴되었는지 여부를 판단한다(단계 304). Therefore, in the frequency synchronization method according to the present invention, after each of the nodes performs steps 300 to 302, a carrier frequency offset occurring between all the nodes in the network

(Step 304). ≪ / RTI >

If the peripheral carrier frequency offset falls below the threshold value in step 304, it is determined that the frequency synchronization has been successfully performed.

If the carrier frequency offset does not converge below the threshold, steps 300 to 302 are repeated.

In a mesh network, a large number of nodes are distributed at random locations, and a topology of almost infinite type can be constructed.

Therefore, the proposed distributed frequency synchronization algorithm has a wide variety of convergence characteristics depending on the type of topology. In the case of a specific topology, a period in which frequency information between arbitrary nodes spreads very slowly occurs, It may take a long synchronization time to converge.

If the synchronization time of the network becomes long, a link that can not communicate due to the influence of the nodes that do not satisfy the system threshold value, which remains in the frequency convergence process, is generated.

Therefore, the synchronization scheme of the mesh network should be able to guarantee the convergence time as fast as possible regardless of the topology.

To this end, we will analyze the frequency convergence characteristics according to the topology type and explain the weighting method for improving the convergence speed of the topology with long synchronization time.

First, in order to confirm the frequency convergence characteristics according to the topology type, three types of topologies as shown in FIG. 4 are considered.

FIG. 4 (a) shows a mesh topology (topology 1) in which nodes having different frequencies are closely connected with each other.

In this topology, since the local carrier frequency change between adjacent nodes can propagate smoothly throughout the network, frequency synchronization can be achieved in a relatively short period of time through repetitive frequency update.

FIG. 4B shows a topology (topology 2) in which two groups converged at different frequencies are connected through one representative node existing in each group.

)' 라고 정의한다. Here, 'group' is defined for convenience of explanation, and the actual network is composed of a single network without group boundaries. In addition, a representative node, which is a link with another group at a boundary of each group, is called a 'bridge node

) '.

In this type of topology, as frequency convergence progresses, frequency information is shared gradually around the bridge node.

과

를 통해서만 상대 그룹의 주파수를 전달받게 된다. In other words, fast convergence among nodes in a group, in which a plurality of nodes are connected to each other and performing close information exchange, progresses between bridge nodes connecting each group,

and

The frequency of the other group is received.

This means that no matter how fast frequency convergence occurs in a group, if the frequency information can not be transmitted quickly between bridge nodes, then the whole network will eventually take a long convergence time for frequency synchronization.

FIG. 4 (c) shows a topology (topology 3) in which several groups are connected to each other via a bridge node.

In this topology, since the number of nodes forming each group is small, but the numbers are similar to each other, the movement of frequency information between arbitrary two groups does not appear quickly, and since a large number of groups are continuously connected, Directionality appears unclear.

For example, group 3 determines the convergence position of its frequency between the frequencies of group 2 and group 4, where groups 2 and 4 are again derived from the frequencies of group 1 and group 3 and the frequencies of group 3 and group 5 The convergence value is determined. Thus, this type of topology eventually becomes intertwined with each other to update the frequency, so a long convergence time is required for the entire network to be synchronized to a specific frequency.

The frequency convergence characteristics for each topology in Fig. 4 can be confirmed through simulation in the following section.

As described above, the frequency synchronization algorithm has various convergence characteristics according to the topology type of the mesh network. In particular, frequency convergence of a topology that requires a long synchronization time occurs in a situation where the movement of frequency information between groups is unclear mainly about the bridge node.

According to a preferred embodiment of the present invention, in order to solve the problem of the synchronization technique for a topology having a long convergence time, a bridge node is selected and a weight is given to a selected node, have.

To this end, any topology of two or more groups is considered. It is assumed that two arbitrary groups adjacent to each other in the topology are composed of nodes 1, 2, 3, 4 and 5 and nodes A, B, C and D, respectively, and bridge nodes of each group are node 1 and node A.

5 (a) shows a carrier frequency offset distribution of neighboring nodes estimated at node 1, which is a bridge node of two groups, after frequency synchronization progresses to some degree in this topology, and FIG. 5 (b) Frequency offset distribution.

At this time, carrier frequency offsets between adjacent two bridge nodes are equal in size and opposite in sign depending on which node is estimated.

)은 노드 1의 그룹 내에 존재하는 노드들의 반송파 주파수 옵셋에 비해 상대적으로 멀리 떨어진 곳에 분포하게 된다. 이는 동일 그룹 내 노드인 노드 1, 2, 3, 4, 5는 서로 긴밀하게 연결되어 있기 때문에 빠르게 주파수가 하나의 값으로 수렴되는 반면, 다른 그룹에 속해 있는 브릿지 노드 A는 자신이 속한 그룹의 주파수 수렴 값에 가깝게 위치하기 때문이다. As the frequency synchronization proceeds, the carrier frequency offset of the other group of bridge nodes (node A) estimated at node 1

) Is distributed relatively far away from the carrier frequency offset of the nodes in the group of node 1. Since the nodes 1, 2, 3, 4, and 5 in the same group are closely connected to each other, the frequency converges to one value quickly. On the other hand, the bridge node A belonging to another group has a frequency Because it is located close to the convergence value.

Therefore, once the synchronization algorithm is started and some degree of convergence has progressed, the nodes of the network can determine which node is the bridge node from the carrier frequency offset distribution of its neighbor node. Based on this fact, when the nodes update their frequencies, it is determined that the farthest value among the carrier frequency offset distributions of the collected neighboring nodes is a bridge node, and the weight is given to this value.

That is, each node can determine a bridge node using the collected carrier frequency offset information.

According to an embodiment of the present invention, when weight values are simultaneously given by using the largest value and the smallest value in two bridge nodes connecting two groups, the frequencies of the frequencies (in both directions around the bridge node) So that the frequency convergence characteristic is improved.

However, the frequency convergence characteristics can be improved even if weights are given only to one bridge node in both directions. In this case, frequency convergence will proceed in which one group follows the frequency of the other group unilaterally.

As can be seen from the simulation, we have investigated the convergence characteristics of 10,000 actual topologies, and it is found that convergence speeds are generally faster when frequency convergence through weighting is performed in one direction there was.

At this time, it is necessary to judge that a node having all the values that are not the largest or the smallest but the difference from the largest value satisfies the system threshold value or less as a bridge node.

This is because a plurality of bridge nodes may be connected to one node according to the topology.

If we assign a weight to only one of the farthest distances, the frequency shift may occur only in a specific group, which may be an obstacle to the optimum convergence speed. Considering these points, the weight values applied for fast frequency synchronization of the network are as follows.

The reason for using the ratio of the number of adjacent nodes as the weight is to equalize the frequency update proportion of the neighboring bridge node of all nodes existing in the network.

If this weight is not unified, one-way sharing of frequency is possible between specific bridge nodes even in the same network, but there may be a situation where information transfer between certain bridge nodes is still small.

This, in turn, makes the frequency convergence characteristic of the network worse or diverge.

는 가중치 비중 결정 상수이다. 만약, 이 값이 1이라면 브릿지 노드의 반송파 주파수 옵셋이 해당 노드의 다음 갱신 값의 약 50%의 비중을 차지하게 되고, 2가 되면 약 75%의 비중을 차지하게 된다. here

Is a weighting constant determining constant. If this value is 1, the carrier frequency offset of the bridge node occupies about 50% of the next update value of the node, and when it is 2, it occupies about 75% of the update value.

When the weight is determined, the network should determine the direction of the frequency information to have the optimal frequency convergence speed around the applied node of the selected weight.

To this end, it is possible to consider a weighting scheme in which the amount of movement of frequency information between groups is increased based on the bridge node, or one of them follows the carrier frequency of the other group. The change of the convergence characteristics according to the weighting method can be confirmed through simulation.

The performance of the frequency synchronization algorithm in the mesh network proposed by the present invention will be analyzed. First, in order to check the convergence characteristics of the frequency synchronization according to the topology type, the nodes are arranged in the distribution as shown in FIG. 4, and the convergence characteristics of the carrier frequency according to the number of updates for each node are confirmed.

Each node estimates a carrier frequency offset by receiving a preamble from its one-hop nodes and performs an operation of modifying its own frequency. In this case, it is assumed that the estimation of the carrier frequency offset is performed without error in each node in order to exclude the factors other than the topology which may affect the convergence speed.

FIG. 6 shows the frequency convergence characteristic of each topology of Equation (3) for each node when the weight is not applied.

Here, the y-axis represents the frequency of each node in terms of CFO, and the difference between the values converted between the two nodes means the carrier frequency offset actually occurring. Each node uses a TCXO with a center frequency of 2.59GHz with an error of within ± 1.0ppm and assumes a subcarrier interval defined by the system to be 10.94kHz and sets the carrier frequency offset value that can occur between nodes to an initial value.

In the case of topology 1 (FIG. 4 (a)) in which each node is closely connected to each other, the number of one hop nodes is evenly distributed, so that frequency update between nodes is uniformly performed.

Each node converges its entire carrier frequency to one frequency while quickly adapting its carrier frequency to the surrounding changes. On the other hand, in the case of topology 2 (FIG. 4B) and topology 3 (FIG. 4C) in which two different (or many) groups are connected to one (or a small number) Since the amount of frequency information to be moved between the two groups is small around the node, the width of the convergence of the frequency convergence is very small as a whole. At this time, it can be seen that the bridge node connecting each group leads the convergence of the network by being located at the midway point between the convergence value of the nodes belonging to the group and the frequency of the adjacent group.

7 shows the frequency convergence characteristics of the proposed algorithm for the topologies 2 and 3 under the same conditions. In order to confirm the difference in convergence speed according to the weighting method described in the previous chapter, we apply the weighting to both sides of the bridge node (Method 1) or restrict the movement of the frequency only in one direction (Method 2) Were simulated.

Figs. 7 (a) and 7 (b) show the simulation results of the method 1, the method (c), and the method (d)

First, by applying the weights, we can confirm that the frequency convergence time of the network decreases in both methods compared to before the weighting application. However, in the case of the method 1, the change of the frequency after the 13th frequency update point at which the weight is applied is observed. In the topology 2 (FIG. 4 (a)), The topology 3 (FIG. 4 (b)) has a problem in that a large number of groups are connected to each other to cancel the frequency shift between the groups, .

On the other hand, when the weight is applied using the method 2, as each group adjusts its frequency to one of the adjacent groups, a frequency shift having a certain direction among the groups occurs in the network, This quickly converges the entire network to a single frequency without much dependence on the topology.

Figs. 7 (c) and 7 (d) show frequency convergence characteristics of topologies 2 and 3 to which weights are applied using method 2, respectively.

In order to verify the performance of the frequency synchronization algorithm proposed in the present invention for arbitrary mesh topology, various types of topologies are configured by distributing various nodes in random space within a network space of a predetermined size to perform simulation.

Each node that enters the network estimates the carrier frequency offset using the preamble received from each node with the nodes existing within the communication distance defined between the nodes defined in the system.

At this time, IEEE 802.16m preamble is used for carrier frequency offset estimation. Based on this, it is determined that the frequency synchronization of the network succeeded when the carrier frequency offset occurring between all the nodes in the network converged below the threshold (= 0.01) defined by the system after the frequency synchronization algorithm was performed.

Figure 8 shows the average of the frequency synchronization times before and after weighting for 10,000 different mesh topologies. It can be seen that the network convergence speed when the weight is applied is greatly improved as compared with the case where the network convergence speed is not applied. As the one-hop radius defined in the system increases, the average number of one-hop nodes per node increases, so that the synchronization speed of the network increases. However, in the actual mesh network, problems such as deterioration of carrier frequency offset estimation performance due to inter-node interference and increase in frequency update time interval for collecting carrier frequency offset information of adjacent nodes occur due to one-hop distance increase, The system one hop distance considered should be determined. Therefore, the proposed frequency synchronization scheme does not depend on the influence of the one-hop distance, and generally has a merit that it can lead to fast synchronization time.

FIG. 9 shows a change in frequency convergence time of each topology before and after weighting. FIG. 9 (a) shows the number of frequency updates required from each node to before synchronization before weight application, and FIG. 9 (b) shows after applying the weight. As a result of checking the convergence time of the frequency synchronization algorithm for 10,000 topologies, the improvement of the synchronization performance after the application of the weight has been significant, and the convergence speed due to the increase of the average number of one hop nodes In the case of the above-mentioned example. As a result, the proposed synchronization scheme can lead to fast convergence speeds generally without being affected by the topology type or the number of one hop nodes in the network.

10 is a block diagram showing a detailed configuration of a node according to a preferred embodiment of the present invention.

10, the node according to the present invention may include a frequency offset information collection unit 1000, a compensation value determination unit 1002, and a weight determination unit 1004.

The frequency offset information collecting unit 1000 transmits and receives preamble to a neighboring node at a predetermined radius (for example, a one-hop distance) every frequency update time to collect peripheral carrier frequency offset information.

The compensation value determiner 1002 determines the carrier frequency offset compensation value of the transmission data at the next frequency update time using the carrier frequency offset information collected as described above.

According to the present invention, the compensation value determiner 1002 can determine the carrier frequency offset compensation value of the transmission data by using the collected carrier frequency offset information together with the weight determined by the weight determining unit 1004.

In more detail, when the wireless mesh network includes two groups (at least two groups) that converge at different frequencies, the weight determining unit 1004 connects the two groups using the peripheral carrier frequency offset information , And determine a weight to be applied to the determined bridge node.

A weight 1 can be set for a neighboring node that is not a bridge node.

The node according to the present invention can determine whether or not frequency synchronization is performed by determining whether the neighboring carrier frequency offset information is less than or equal to a preset threshold value. If the frequency offset is not reached below the threshold, the node collects frequency offset information, .

Embodiments of the present invention may be implemented in the form of program instructions that can be executed on various computer means and recorded on a computer readable medium. The computer-readable medium may include program instructions, data files, data structures, and the like, alone or in combination. The program instructions recorded on the medium may be those specially designed and constructed for the present invention or may be available to those skilled in the art of computer software. Examples of computer-readable media include magnetic media such as hard disks, floppy disks, and magnetic tape; optical media such as CD-ROMs and DVDs; magnetic media such as floppy disks; Examples of program instructions, such as magneto-optical and ROM, RAM, flash memory and the like, can be executed by a computer using an interpreter or the like, as well as machine code, Includes a high-level language code. The hardware devices described above may be configured to operate as one or more software modules to perform operations of one embodiment of the present invention, and vice versa.

It will be apparent to those skilled in the relevant art that various modifications, additions and substitutions are possible, without departing from the spirit and scope of the invention as defined by the appended claims. The appended claims are to be considered as falling within the scope of the following claims.

Claims (13)

A distributed frequency synchronization method in a wireless mesh network,
(a) receiving a preamble from at least one neighboring node existing within a predetermined radius at a current frequency update time and collecting peripheral carrier frequency offset information;
(b) determining a carrier frequency offset compensation value of transmission data at a next frequency update time using the collected peripheral carrier frequency offset information; And
(c) repeating the steps (a) and (b) until the peripheral carrier frequency offset information is less than or equal to a preset threshold value. The method according to claim 1,
Wherein the predetermined radius is one of a one-hop distance and a maximum communicable link distance between nodes. The method according to claim 1,
Wherein the step (b) comprises determining a carrier frequency offset compensation value of the transmission data by using a floating algorithm. The method according to claim 1,
Determining a weight to be applied to a carrier frequency offset of the at least one neighboring node using the collected carrier frequency offset information prior to step (b)
Wherein the step (b) comprises determining a carrier frequency offset compensation value of the transmission data using the collected peripheral carrier frequency offset information and the weight. 5. The method of claim 4,
Wherein the carrier frequency offset compensation value of the transmission data is determined by the following equation.
[Mathematical Expression]

Where k is the number of frequency updates, n is the unique number of a node in the network,

The number of one-hop nodes of the node with the node number n,

Is the carrier frequency offset of the neighbor node i estimated at the nth node at the kth update time (kth frequency update time)

Is the transmission data carrier frequency offset compensation value at the next frequency update time (k + 1 < th > frequency update time) of the node n,

Is a weight applied to the carrier frequency offset of the node i in consideration of the frequency convergence characteristic of the network 5. The method of claim 4,
Wherein the wireless mesh network comprises two groups converging at different frequencies,
The weight determination step may include:
Determining a bridge node connecting the two groups of the one or more peripheral nodes using the peripheral carrier frequency offset information; And
And assigning different weights to the determined nodes other than the bridge node and the bridge node. The method according to claim 6,
Wherein a weight of the bridge node is determined as a ratio of the number of the at least one neighboring node. The method according to claim 6,
Further comprising determining a direction of movement of the frequency information around the bridge node. 9. A computer-readable recording medium on which a program for performing the method according to any one of claims 1 to 8 is recorded. A distributed frequency synchronization apparatus in a wireless mesh network,
A frequency information collecting unit for receiving a preamble from at least one neighboring node existing within a predetermined radius at a current frequency update time and collecting peripheral carrier frequency offset information; And
And a compensation value determination unit for determining a carrier frequency offset compensation value of transmission data at a next frequency update time using the collected peripheral carrier frequency offset information,
Wherein the peripheral frequency offset information collection and the carrier frequency offset compensation value determination are repeatedly performed until the peripheral carrier frequency offset information becomes less than or equal to a preset threshold value. 11. The method of claim 10,
Wherein the predetermined radius is one of a one-hop distance and a maximum communicable link distance between nodes. 11. The method of claim 10,
Further comprising a weight determining unit for determining a weight to be applied to a carrier frequency offset of the at least one neighboring node using the collected carrier frequency offset information,
Wherein the compensation value determination unit determines the carrier frequency offset compensation value of the transmission data using the collected peripheral carrier frequency offset information and the weight. 13. The method of claim 12,
Wherein the weight determining unit includes:
Wherein the wireless mesh network comprises two groups converging at different frequencies,
Determining at least one node having a value that satisfies a predetermined threshold value or less as a difference from a largest value of the collected carrier frequency offset information as a bridge node and distributing the weighted value to the determined bridge node Frequency synchronization device.

Images