JP2015162884A - Temporal synchronization establishment method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To solve such a problem of wireless communication where a star network may be constituted of a large number of pairs of a receiver (FC) and a transmitter (node) that it lacks scalability when an intrinsic signal is used in order to establish timing synchronization of each transmitter, because the number of terminals capable of participating to the network is limited.SOLUTION: A temporal synchronization establishment method proposes a method for establishing temporal synchronization (timing offset compensation) which allows a plurality of nodes to use the same specific series. According to the temporal synchronization establishment method, a node can recognize the timing difference from other node, by informing all nodes of the arrival timing of a signal transmitted from a node. The transmission timing of the own node is random access based on a random number. Since a node can define the access timing autonomously, the number of nodes is not limited, and a temporal synchronization establishment method of high scalability can be provided.

Description

  The present invention relates to a method for establishing time synchronization in a wireless communication system.

  In a wireless communication system, there are many star-type network forms in which a large number of wireless terminals (nodes) access one base (base station: FC). Examples thereof include mobile communication systems such as mobile phones and smart phones, wireless LANs (Wi-Fi) in indoor environments, and sensor networks that aggregate sensing information. When establishing communication in a star network, time timing synchronization of each node is required. This is from both sides of a request from the application side that ensures time consistency of sensing results and a request from the wireless communication system side such as a radio access protocol and modulation / demodulation processing, and high timing synchronization accuracy is required.

  As a timing synchronization probability method in a star type network, closed loop type timing synchronization control is promising. In closed-loop timing synchronization control, the FC notifies a certain command and prompts a node that ensures synchronization to transmit a control signal. The node transmits a control signal at the time indicated in the command. As a result, when receiving various control signals, FC evaluates the arrival timing difference of the signals and notifies each node of the correction amount necessary for timing correction (Non-patent Document 1).

Edited by Takashi Shono, WiMAX textbook, Impress R & D 2008

  However, in the above method, since the signal of each node is separated, a signal sequence having excellent autocorrelation and cross-correlation is required. In this way, in order for each node to use a unique signal sequence, a sequence predetermined by the transmitter / receiver must be defined, which limits the expandability of the number of nodes. For this reason, it becomes difficult to apply to a dynamic wireless network in which the number of nodes becomes large or the number of nodes changes greatly.

  An object of the present invention is to provide an efficient time synchronization establishment method even when the number of nodes in a wireless communication system becomes large.

The time synchronization establishment method according to claim 1, which has been made to achieve the object, is a time synchronization establishment method in wireless communication between a transmitter and a receiver,
1. a control signal notification step in which a transmitter transmits a control signal at a predetermined transmission timing at a predetermined notification time; and 2. the receiver quantizes the control signal and sets the reception timing of the receiver to a frequency. A recording step for recording in the timing map as follows: 3. a timing map notification step for the receiver to notify the transmitter of the timing map; and 4. a timing from the timing map to which the transmitter is notified. A specifying step of specifying; and 5. a correcting step of correcting the transmission timing of the transmitter based on the timing of the transmitter specified by the transmitter.

  In addition, the time synchronization establishment method described in claim 2 of the claims is described in claim 1 of the claims, and is characterized by repeating steps 1 to 5.

  The time synchronization establishment method described in claim 3 of the claims is the method described in claim 1 or 2 of the claims, wherein the transmission timing is determined randomly by a transmitter. It is characterized by the timing.

  Further, the time synchronization establishment method described in claim 4 of the claims is described in any one of claims 1 to 3 of the claims, and in the specifying step, The transmitter estimates a recording position recorded as a frequency from its own transmission timing, and receives a band limited to the recording position.

  Further, the time synchronization establishment method described in claim 5 of the claim is the one described in any one of claims 1 to 4 of the claim, wherein the receiver notifies the notification. A notification time notification step of notifying the transmitter of a time request is further provided.

  Further, the time synchronization establishment method described in claim 6 of the claim is the one described in any one of claims 1 to 5 of the claim, wherein the control signal includes: The method further comprises a quantization interval determining step of detecting a maximum value of the timing difference from the control signal transmitted first, and determining the quantization interval by dividing the maximum value by a predetermined quantization number. .

  According to the present invention, it is possible to provide an efficient time synchronization establishment method even when the number of nodes is large in a wireless communication system.

It is a figure of the model of a wireless sensor network, and the flow of timing correction. It is an image figure of separation of a control signal and frequency mapping. It is an image figure of the timing recognition by a node. It is a flowchart showing the process of an example of the form of this invention. It is a graph showing the timing offset amount with respect to the number of stages. It is a graph showing the number of accumulation turns after the end of each stage. It is a graph showing the number of turns required to achieve a timing error of 1%. It is a graph showing the characteristic of the normalization timing error with respect to the number of stages.

  Hereinafter, modes for carrying out the present invention will be described.

  Figure 1 shows the model of the sensor network considered. In the present invention, a star-type wireless sensor network is assumed in which a plurality of sensing terminals (nodes) are present and a sensing result is notified to one receiver (FC) by wireless communication. Assume that all nodes share the same frequency band. Wireless access from the node to the FC and vice versa access are referred to as uplink and downlink. In order to simplify the modulation / demodulation process as much as possible, the node sets the frequency band of the signal modulation / demodulation process narrower than the modulation / demodulation process of the FC.

  The node is assumed to have a time clock that determines the time transmission timing. Using this time clock, the access timing in wireless communication is determined from the determination of sensing timing. The time clock has a difference (offset) for each node due to temperature, individual differences of terminals, and the like, and accordingly, an access timing error (timing offset) between nodes occurs.

  In the time synchronization method according to the present invention, in order to eliminate the timing offset, communication using a control signal is performed so that each node recognizes its own timing offset. Each node corrects the transmission timing based on the recognized own timing offset, thereby eliminating the timing offset between the nodes.

  In the communication using the control signal, each node needs to recognize a timing offset with another node, and therefore it is necessary to transmit the control signal at the same time. As an example of a method for realizing simultaneous transmission at the same time, for example, a method in which the FC transmits a request for a control signal to all nodes at the start of time synchronization processing can be cited. At this time, since the time at which the control signal is transmitted is described in the request signal, each node can transmit the control signal at the designated time with reference to its own time clock.

  As a result of performing the above request processing, a control signal arrives at the FC at a timing that spreads around the time axis with a predetermined time as the center. Since the timing offset in this case is irregular, a contention state in which all control signals are interfering is obtained. For this reason, the FC attempts to separate the signal from the received signal using the autocorrelation of the control signal. At the same time, the FC performs a mapping process for recording the timing of the separated signal.

  FIG. 2 shows an image diagram of the process of separating and mapping the arrival timing from the received signal. The received signal (a) and the control signal (b) generated locally in the FC are correlated at predetermined time intervals to determine the presence or absence of the signal. This method of recognizing the arrival of a signal while sliding along the time axis is called sliding window (SW) type timing detection.

  As a result of correlation detection for each time, if a correlation value equal to or greater than a threshold value is recognized, it is recognized that a signal has arrived. The presence / absence of recognition is expressed by the presence / absence of a frequency component (c) corresponding to a predetermined time. As a result, a timing map indicating whether or not a signal has arrived with respect to the time axis is formed.

  In order to form a map indicating whether or not there is an arrival signal for each time formed by FC, quantization for discretizing the time axis direction is required. In the time synchronization method according to the present invention, the quantization number is fixed and the quantization interval is dynamically switched to enable efficient time synchronization according to the initial timing offset amount.

  The quantization interval in the present invention is determined as follows. First, when the first control signal arrives, the FC detects the presence or absence of a signal by applying a correlator in a sufficiently wide time range that does not miss the arrival of the signal. Therefore, the timing difference of the incoming signal that maximizes the timing difference is evaluated, and this is divided by the number of quantizations to determine the quantization interval. After that, the FC notifies the node of the quantization level along with the notification of the timing map. As a result, the node can recognize the quantization interval and obtain the timing difference from the result of the discrete map. If notification of the quantization interval is difficult or complicated, the timing correction may be initially performed at a fixed interval.

  The quantization interval is proportional to the correctable amount at the time of timing correction. Therefore, in order to correct the timing below the quantization interval, it is effective to reduce the quantization interval as the time synchronization process proceeds. As a method of reducing the quantization interval, for example, timing correction in a stage system can be cited. In the time synchronization method according to the present invention, when the timing correction of all nodes is completed by the notification of the timing map, it can be determined that the timing offset is corrected to be equal to or less than the quantization interval at that time. Therefore, in FC, after the timing correction of all the nodes is completed, a new quantization interval is obtained by setting the quantization interval at that time as the maximum offset amount and dividing it by the quantization number. After that, by repeating the turn again, timing correction at shorter time intervals becomes possible.

  The FC notifies all nodes of the formed timing map. An example of the notification method in the present invention is, for example, subcarrier mapping based on OFDM. In this method, a digitized signal is notified based on the presence or absence of discrete subcarriers on the frequency axis by orthogonal frequency division multiplexing (OFDM) modulation. In the timing map formation in the present invention, the time number of the quantized arrival timing map is directly projected onto the OFDM subcarrier number. Thereby, the magnitude of the arrival timing is shown in a form that appears directly in the magnitude of the frequency.

  As a result of notification of the timing map from the FC, the node recognizes its own timing offset from the notified timing map, and uses this as a transmission timing correction amount. A node can recognize the relationship between itself and other nodes from the timing map, but cannot recognize its own timing. As an example of a method for recognizing its own timing, for example, there is a method in which a node arbitrarily changes the transmission timing of its own control signal and recognizes it by referring to the result. The transmission timing can be changed at random, for example.

  Figure 3 shows how a node uses random access to narrow down its own timing. Since the node recognizes its own transmission timing, it applies random transmission timing. The notification of the FC request signal, the transmission of the node control signal, and the notification of the FC timing map are set as one turn, and this is repeated for multiple turns.

  When a node transmits a control signal at each turn, it transmits at a random timing specific to the node. As a result, each node estimates its own transmission timing as follows. Shift the time by the unique timing set by the node in the map notified in the first turn. As a result, if the arrival of a signal is recognized at the same time as compared with the map notified on the second turn, it is left as a candidate. After the third turn, the result of giving a unique time shift based on the candidate for the first turn is compared with the map of each turn. By repeating the turn in this way, the uniqueness of the access timing can be used to narrow down to one candidate from the map of the first turn and the arrival timing of the signal can be recognized. The timing can be corrected by recognizing its own timing offset from the result finally narrowed down to one.

  With the above processing, time synchronization can be performed by the method according to the present invention. The process from transmission of the control signal to timing correction is made one stage, and this is repeated a plurality of stages to reduce the amount of timing offset, and finally the time synchronization processing is ended when the required timing correction is completed. FIG. 4 shows a flowchart showing the above processing protocol.

  In the time synchronization method according to the present invention, the node freely determines the access timing, and the FC does not need to know the access timing information. In other words, since the node and the FC do not need to share information about the access timing, the access timing options are greatly expanded, and thus high scalability can be realized. Since a node can freely set a highly unique sequence, the number of nodes that can simultaneously establish timing synchronization is large.

  In the time synchronization method according to the present invention, by utilizing the correlation of the transmission timing, in the timing map notification from the FC to the node, a wide band that can be notified by the FC can be secured even when the band on the node side is narrow. As a result, a wide selection of transmission timings and highly specific timings can be used, so that time synchronization with many nodes becomes possible. Also, if the quantization interval can be notified from the receiver to the transmitter, after the first control signal is transmitted from the node to the FC, the arrival time of all signals will be within the frequency map quantization number. By setting the quantization interval (time interval) to a large value, the arrival timing of all times can be notified even when the transmission bandwidth is narrow.

  In the time synchronization method according to the present invention, since the FC notifies the timing map to all the nodes, notification for each node is unnecessary, and the bandwidth utilization efficiency is extremely high.

  The effectiveness of the time synchronization method according to the present invention is evaluated by computer simulation. Table 1 shows the simulation specifications.

  In this embodiment, as a setting of the conventional method to be compared, the number of subcarriers corresponding to the detection bandwidth of the node is equal to the number of subcarriers that can be notified by the FC. The timing offset model is as follows.

  Here, τ ^ k is an unknown timing error component for the k-th node, and τ-k is a known timing error component. τ ^ k and τ-k were modeled with uniform random numbers in the range of ± N. α is a value that determines unknown and known timing components, and 0 ≦ α ≦ 1. Therefore, as α increases, the ratio of known timing components on the node side increases. Further, in order to make the comparison between the conventional method and this embodiment fair, the conventional method sets the maximum timing offset amount to half that of this embodiment. This corresponds to the case of α = 0.5.

  FIG. 5 shows the timing offset amount with respect to the number of stages. Here, the correlation α = 0.3 for timing synchronization was set. The timing error amount is normalized by an unknown timing error component τ ^ k. The number of nodes was set to 16. From the figure, it was confirmed that the proposed method achieved a lower timing error with a smaller number of stages than the conventional method. Moreover, it was recognized that the characteristic difference with respect to the conventional method increased as the number of stages increased. This is because the method according to the present embodiment can use 16 times more subcarriers than the conventional method, so the quantization interval of the timing error that can be notified from the FC to the node is reduced to 1/16, This is because the timing error can be rapidly reduced.

  FIG. 6 shows the cumulative number of turns after the end of each stage. From the figure, it is recognized that the present embodiment finishes each stage with a smaller number of turns. Specifically, when the number of stages is 6, the number of turns is suppressed to 1/4 as compared with the conventional method. This is because the number of subcarriers that can be used is large, so the range of random numbers that can be selected by the node is wide, the probability that a node will use the same subcarrier repeatedly is kept low, and node identification can be performed with a small number of turns. It is.

  FIG. 7 shows the characteristics of the number of turns necessary to achieve a normalized timing error of 1%. Similarly, the timing synchronization correlation α is set to 0.3. From the figure, it can be seen that the number of turns increases exponentially in the conventional method as the number of users increases. On the other hand, in the proposed method, the increase in the number of turns is slow, and even in 64 nodes, it is completed in about 20 turns, and even in an environment with many nodes, the required time timing synchronization accuracy is achieved by exchanging few control signals. It is recognized that it can be done.

  FIG. 8 shows the characteristics of the normalized timing error with respect to the number of stages. The result is shown when the number of nodes is 16 and the timing error correlation α is changed to 0−0.3. From the figure, a floor where the measurement error becomes constant as α becomes smaller than 0.3 is recognized. This is because, when the correlation between the upper and lower timing errors is below a certain level, a signal appears outside the signal detection range notified from the FC to the node, and detection is missed.

Claims (6)

A method for establishing time synchronization in wireless communication between a transmitter and a receiver, comprising: 1. a control signal notification step in which a transmitter transmits a control signal at a predetermined transmission timing at a predetermined notification time;
2.Recording step in which the receiver quantizes the control signal and records the reception timing of the receiver as a frequency in a timing map;
3. a timing map notification step in which the receiver notifies the transmitter of the timing map;
4. a specific step of identifying its own timing from the timing map notified by the transmitter;
5. A correction step of correcting the transmission timing of itself based on the timing of the transmitter specified by the transmitter;
A method of establishing time synchronization.   2. The time synchronization establishment method according to claim 1, wherein steps 1 to 5 are repeated.   3. The time synchronization establishment method according to claim 1, wherein the transmission timing is a timing randomly determined by a transmitter.   4. In the specifying step, the transmitter estimates a recording position recorded as a frequency from its own transmission timing, and receives the recording position with a band limited to the recording position. The time synchronization establishment method according to item 1.   5. The time synchronization establishment method according to claim 1, further comprising a notification time notification step in which the receiver notifies the transmitter of the request for the notification time.   A quantization interval determination step of detecting a maximum timing difference value from the control signal transmitted first among the control signals, and determining the quantization interval by dividing the maximum value by a predetermined quantization number; The time synchronization establishment method according to any one of claims 1 to 5, further comprising: